JP2011027756A - Gas concentration detection device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable the correction of the output of a sensor circuit with optional timings, and to improve accuracy in detecting gas concentration. <P>SOLUTION: A sensor control circuit 30 is connected to a sensor element 10 forming an A/F sensor. The sensor control circuit 30 measures an element current flowing when a voltage is applied to the sensor element 10, and generates an A/F output voltage AFO corresponding to a measured current. A microcomputer 20 temporarily sets the sensor control circuit 30 in the stoichiometric mixture detection state by opening a switch 35, and computes an offset error on the basis of the A/F output voltage AFO at that time. The A/F output voltage AFO is corrected on the basis of the offset error. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a gas concentration detection apparatus, and more particularly to a technique for eliminating an error in sensor output due to individual circuit differences.

  Conventionally, a limit current type air-fuel ratio sensor (so-called A / F sensor) that detects an oxygen concentration (air-fuel ratio) in an exhaust gas discharged from a vehicle engine is known. That is, the air-fuel ratio sensor has a sensor element made of a solid electrolyte body, and is configured to flow an element current corresponding to the oxygen concentration each time a voltage is applied to the sensor element. In this case, the element current flowing through the sensor element is measured by the sensor circuit, and the sensor circuit outputs an oxygen concentration signal proportional to the element current.

  Here, there is a characteristic variation in the air-fuel ratio sensor or sensor circuit, and a problem arises that the detection accuracy of the air-fuel ratio decreases due to the characteristic variation. To deal with this problem, for example, in Patent Document 1, it is determined that the air-fuel ratio sensor is in an inactive state, and an output error is detected from the output value of the sensor circuit and the reference output value when in the inactive state. The air-fuel ratio conversion map is calibrated based on this output error.

  However, since the output value of the sensor circuit measured in the sensor inactive state is an essential requirement in the technique of Patent Document 1, a circuit error can be detected only when the air-fuel ratio sensor is cold started. Therefore, there arises a disadvantage that the chance of detecting a circuit error is limited. It is also conceivable that a circuit error cannot be detected when the temperature of the sensor circuit rises, and the air-fuel ratio detection accuracy deteriorates. Since it is difficult to completely determine the inactive state of the air-fuel ratio sensor, there is a possibility that erroneous correction is performed.

  In vehicular gasoline engines, there is a high demand for detecting the vicinity of stoichiometric (theoretical air-fuel ratio) with high accuracy, and it is also considered to detect the air-fuel ratio precisely by enlarging only the vicinity of the stoichiometric area. Even in this situation, it cannot be said that the correction of the output error is sufficient with the existing technology, and there is still much room for improvement. Further, although circuit errors are reduced by trimming adjustment, problems still remain such that resistance errors cannot be completely adjusted.

Japanese Patent No. 3257319

  The main object of the present invention is to provide a gas concentration detection device capable of calculating an output error of a sensor circuit at an arbitrary timing, and thus improving the detection accuracy of the gas concentration.

  In the first configuration, the sensor circuit is normally set in the normal state, and the gas concentration is detected for the exhaust gas from the internal combustion engine. When the sensor circuit is temporarily switched to the reference output generation state, the sensor circuit attempts to generate a predetermined reference output regardless of the exhaust gas atmosphere. Then, after the sensor circuit is switched to the reference output generation state, the gas concentration output correction amount is calculated based on the difference between the gas concentration output generated at that time and the reference output.

  In this case, if there is a variation in the characteristics of the gas concentration sensor or sensor circuit, an error occurs in the output of the sensor circuit, and when the reference output is generated, the gas concentration output deviates from the reference output by the output error of the sensor circuit. . If the error of the gas concentration output generated in the reference output generation state is obtained, the gas concentration output correction amount for correcting the output of the sensor circuit to the reference output (normal value) can be calculated suitably and easily. Since the gas concentration output correction amount can be calculated in this way, the detection accuracy of the gas concentration can be improved. Also, in this configuration, the data for calculating the correction amount is obtained immediately by switching the state of the sensor circuit, so unlike the conventional technology that performs error detection on the condition that the sensor is inactive, the output error calculation timing is There are no restrictions. Accordingly, the output error of the sensor circuit can be calculated at an arbitrary timing, and a highly accurate gas concentration output can be obtained as desired.

  The stoichiometric (theoretical air-fuel ratio) is theoretically an air-fuel ratio at which fuel completely burns, and the exhaust gas after stoichiometric combustion has an oxygen concentration = 0 (element current = 0 mA). Therefore, in the stoichiometric detection state, a gas concentration output (stoichiometric reference output) originally corresponding to the element current = 0 mA is generated. In the second configuration, at the time of switching to the reference output generation state, the sensor circuit is set to a stoichiometric detection state in which the stoichiometric reference output is generated. Then, a gas concentration output correction amount is calculated based on the gas concentration output generated at that time and the stoichiometric reference output. In this case, a state in which the element current = 0 mA can be created relatively easily, and a desired effect can be obtained with relatively simple means.

  As specific means for switching the sensor circuit to the stoichiometric detection state, there are the following third to sixth configurations. In the third configuration, the switch circuit provided in the current path of the sensor circuit is opened, so that the sensor circuit is in the stoichiometric detection state. In this case, the device current is cut off by opening the switch means, and a stoichiometric detection state of device current = 0 mA can be created.

  In the fourth configuration, in the sensor circuit, the output circuit switch element of the amplifier circuit connected to the gas concentration sensor is forcibly turned off, so that the sensor circuit is in the stoichiometric detection state. In this case, by turning off the switch element, the element current is cut off, and a stoichiometric detection state where the element current = 0 mA can be created. The sensor circuit is provided with changing means (for example, a transistor as a switch element) for turning on or off the output stage switch element of the amplifier circuit, and the output stage switch element is turned off by this changing means. It is good also as composition to do.

  In the fifth configuration, the current detection resistor and the switch unit are respectively connected to the connection terminals on both the positive and negative sides of the gas concentration sensor, and the switch unit is opened, so that the sensor circuit is in the stoichiometric detection state. That is, a stoichiometric detection state with an element current = 0 mA can be created. Particularly in this case, since the switch means is connected to the side opposite to the gas concentration detection side (current detection resistor connection side), the gas concentration output fluctuates with the opening of the switch and is output as it is. Inconvenience can be suppressed.

  In the sixth configuration, the sensor circuit is in a stoichiometric detection state by setting the connection terminals on both the positive and negative sides connected to the gas concentration sensor to the same potential. That is, the sensor applied voltage = 0V, and as a result, a stoichiometric detection state with an element current = 0 mA can be created.

  As described above, the calculation timing of the output error is arbitrary, but at this time, specifically, as in the seventh configuration, switching to the reference output generation state by the state switching unit may be performed at a predetermined time period. Alternatively, as in the eighth configuration, the switching to the reference output generation state by the state switching unit may be performed in a state where the gas concentration output is not required in the control system of the internal combustion engine. In particular, in the eighth configuration, the gas concentration output correction amount can be calculated without affecting the execution of various controls in the control system. The state where the gas concentration output is not required in the control system includes, for example, a case where the fuel is being cut or a state before the sensor activation in the system that performs the air-fuel ratio feedback control based on the sensor output.

  Further, when the temperature of the sensor circuit changes, the output error amount changes due to a change in resistance value or the like. Therefore, as in the ninth configuration, the switching to the reference output generation state by the state switching unit may be performed when the temperature of the sensor circuit changes more than a predetermined value. That is, the temperature change of the sensor circuit is monitored, and when the temperature change exceeds a predetermined value, switching to the reference output generation state is performed to calculate the gas concentration output correction amount.

  Further, as in the tenth configuration, the switching to the reference output generation state by the state switching unit may be performed at least either immediately after starting the internal combustion engine or at the time of stopping. When the internal combustion engine is stopped, for example, switching by the state switching unit may be performed in the main relay control that is executed by temporarily delaying the power cutoff to the control device or the like.

  In the eleventh configuration, the gas concentration output correction amount calculated by the correction amount calculation means is sequentially updated and stored. At this time, the gas concentration output correction amount may be stored in the backup memory each time. Accordingly, for example, even after the internal combustion engine is started and before the gas concentration output correction amount is calculated, an appropriate gas concentration output can be obtained.

  Since the output error amount changes when the temperature of the sensor circuit changes, it is preferable to store the gas concentration output correction amount in association with the temperature of the sensor circuit as in the twelfth configuration. Thereby, even if the temperature of the sensor circuit changes in accordance with the operating state of the internal combustion engine, an appropriate gas concentration output can be obtained correspondingly.

  Further, the output error of the sensor circuit includes a substantially constant amount of offset error with respect to the reference output characteristic and a gain error that changes in accordance with the sensor current. In such a case, in order to perform error correction including gain error, it is preferable to perform error detection at a plurality of reference output points and calculate the gas concentration output correction amount from the result.

  In the thirteenth configuration, the gas concentration output correction amount is calculated as a first error amount. Further, it is determined that a predetermined reference gas atmosphere different from the gas atmosphere corresponding to the reference output generation state is obtained, and the gas concentration output generated in the sensor circuit when the reference gas atmosphere is reached and the original reference output at that time The second error amount is calculated based on the above. Then, the gas concentration output is corrected based on the first error amount and the second error amount. In this case, the error amount is detected at two points on the sensor output characteristics, so that an output error including not only an offset error but also a gain error can be obtained, and the output error of the sensor circuit can be reliably corrected. Become. Thereby, the detection accuracy of gas concentration can be improved.

  Here, the atmosphere is discharged as exhaust gas when the fuel of the internal combustion engine is cut. In this case, as in the fourteenth configuration, it is preferable to determine that the exhaust gas of the internal combustion engine becomes a gas atmosphere equivalent to the atmosphere, and calculate the second error amount at that time.

  In the fifteenth configuration, the first and second errors are determined based on the gas concentration output generated in the sensor circuit and the original reference output at that time in two predetermined gas atmospheres or sensor circuit states corresponding to the gas atmosphere. The amount is calculated, and the gas concentration output is corrected based on the first and second error amounts. In this case, the error amount is detected at two points on the sensor output characteristics, so that an output error including not only an offset error but also a gain error can be obtained, and the output error of the sensor circuit can be reliably corrected. Become. Thereby, the detection accuracy of gas concentration can be improved.

  Here, as in the sixteenth configuration, the first error amount is calculated based on the gas concentration output generated in the sensor circuit immediately after the start of the internal combustion engine or after the stop, and the exhaust gas of the internal combustion engine is equivalent to the atmosphere. The second error amount may be calculated based on the gas concentration output generated by the sensor circuit when the gas atmosphere is reached. In this case, there is a restriction that the calculation of the first error amount is performed immediately after starting the internal combustion engine in the sensor inactive state or after the stop. However, the first error amount is calculated when the internal combustion engine is started. Once calculated once, the gas concentration output can be appropriately corrected each time the second error amount is calculated.

  In the seventeenth configuration, the first error amount and the second error amount are sequentially updated and stored. At this time, the first error amount and the second error amount may be stored in the backup memory each time. As a result, for example, even after the internal combustion engine is started and before each error amount is calculated, the gas concentration output can be suitably corrected, and the gas concentration can be detected appropriately.

  Since the output error amount changes when the temperature of the sensor circuit changes, it is preferable to store the first error amount and the second error amount in association with the temperature of the sensor circuit as in the eighteenth configuration. Thereby, even if the temperature of the sensor circuit changes in accordance with the operating state of the internal combustion engine, an appropriate gas concentration output can be obtained correspondingly.

It is a circuit diagram which shows the electric constitution of a sensor control circuit. It is sectional drawing which shows the structure of a sensor element. It is a figure which shows the output characteristic of an A / F sensor. It is a figure which shows the output characteristic of a sensor control circuit. It is a flowchart which shows an A / F output voltage correction process. It is a circuit diagram which shows the electric constitution of a sensor control circuit. It is a figure which shows the output characteristic of a sensor control circuit. It is a flowchart which shows an A / F output voltage correction process. It is sectional drawing which shows the element structure of another A / F sensor. It is a circuit diagram which shows the electric constitution of a sensor control circuit. It is a circuit diagram which shows the electric constitution of a sensor control circuit. It is sectional drawing which shows the element structure of another A / F sensor. It is a circuit diagram which shows the electric constitution of a sensor control circuit. It is a circuit diagram which shows the electric constitution of a sensor control circuit. It is a circuit diagram which shows the electric constitution of a sensor control circuit. It is a figure which shows the exhaust gas purification characteristic of a gasoline engine. It is a figure which shows the relationship between O2 density | concentration and exhaust gas density | concentration of a diesel engine.

(First embodiment)
DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a gas concentration detection device according to the present invention will be described below with reference to the drawings. In the present embodiment, an air-fuel ratio detection device that detects an oxygen concentration (air-fuel ratio, hereinafter also referred to as A / F) in the same gas using exhaust gas (combustion gas) discharged from an on-vehicle engine as a detection gas is embodied. The air-fuel ratio detection result is used in an air-fuel ratio control system constituted by an engine ECU or the like. In the air-fuel ratio control system, stoichiometric combustion control in which the air-fuel ratio is feedback-controlled in the vicinity of the stoichiometry, lean combustion control in which the air-fuel ratio is feedback-controlled in a predetermined lean region, and the like are appropriately realized.

  First, the configuration of an A / F sensor as a gas concentration sensor will be described with reference to FIG. The A / F sensor has a sensor element 10 having a laminated structure, and FIG. 2 shows a cross-sectional configuration of the sensor element 10. Actually, the sensor element 10 has a long shape extending in the direction orthogonal to the paper surface of FIG. 2, and the entire element is accommodated in a housing or an element cover.

  The sensor element 10 includes a solid electrolyte 11, a diffusion resistance layer 12, a shielding layer 13, and an insulating layer 14, which are stacked on the top and bottom of the drawing. A protective layer (not shown) is provided around the element 10. The rectangular plate-shaped solid electrolyte 11 (solid electrolyte body) is a sheet made of partially stabilized zirconia, and a pair of upper and lower electrodes 15 and 16 are disposed opposite to each other with the solid electrolyte 11 interposed therebetween. The electrodes 15 and 16 are made of platinum Pt or the like. The diffusion resistance layer 12 is made of a porous sheet for introducing exhaust gas into the electrode 15, and the shielding layer 13 is made of a dense layer for suppressing permeation of the exhaust gas. Each of these layers 12 and 13 is formed by molding a ceramic such as alumina or zirconia by a sheet forming method or the like, but the gas permeability varies depending on the difference in the average pore diameter and porosity of the porosity.

  The insulating layer 14 is made of ceramics such as alumina or zirconia, and an air duct 17 is formed at a portion facing the electrode 16. In addition, a heater 18 made of platinum Pt or the like is embedded in the insulating layer 14. The heater 18 is a linear heating element that generates heat when energized by a battery power source, and heats the entire element by the generated heat. In the following description, in some cases, the electrode 15 is also referred to as a diffusion layer side electrode, and the electrode 16 is also referred to as an atmosphere side electrode. In the present embodiment, a terminal connected to the atmosphere side electrode 16 is a positive side terminal (+ terminal), and a terminal connected to the diffusion layer side electrode 15 is a negative side terminal (−terminal).

  In the sensor element 10, the surrounding exhaust gas is introduced from the side portion of the diffusion resistance layer 12 and reaches the diffusion layer side electrode 15. When the exhaust gas is lean, oxygen in the exhaust gas is decomposed by the diffusion layer side electrode 15 by applying a voltage between the electrodes 15 and 16, is ionized and passes through the solid electrolyte 11, and then is discharged from the atmosphere side electrode 16 to the atmosphere duct 17. Is done. At this time, a current (positive current) flows in the direction from the atmosphere side electrode 16 to the diffusion layer side electrode 15. On the other hand, when the exhaust gas is rich, oxygen in the atmosphere duct 17 is decomposed by the atmosphere side electrode 16, ionized and passes through the solid electrolyte 11, and then discharged from the diffusion layer side electrode 15. And it reacts with unburned components such as HC and CO in the exhaust gas. At this time, a current (negative current) flows in the direction from the diffusion layer side electrode 15 to the atmosphere side electrode 16.

  FIG. 3 is a drawing showing basic voltage-current characteristics (VI characteristics) of the A / F sensor. In FIG. 3, a flat portion parallel to the voltage axis (horizontal axis) is a limit current region for specifying the element current Ip (limit current) of the sensor element 10, and the increase / decrease in the element current Ip is the increase / decrease in the air / fuel ratio (that is, , Lean and rich). That is, the element current Ip increases as the air-fuel ratio becomes leaner, and the element current Ip decreases as the air-fuel ratio becomes richer.

  In this VI characteristic, the lower voltage side than the limit current region is a resistance dominant region, and the slope of the primary straight line portion in the resistance dominant region is specified by the DC internal resistance Ri of the sensor element 10. The DC internal resistance Ri changes according to the element temperature, and the DC internal resistance Ri increases as the element temperature decreases. That is, at this time, the slope of the primary straight line portion of the resistance dominating region becomes small (the straight line portion lies down). Further, when the element temperature rises, the DC internal resistance Ri decreases. That is, at this time, the slope of the primary straight line portion of the resistance dominating region becomes large (the straight line portion stands up). RG in the drawing represents an applied voltage characteristic (applied voltage line) for determining the applied voltage Vp to the sensor element 10.

  Next, the configuration of the sensor control system as the main part of the present invention will be described with reference to FIG. The sensor control system is provided with a microcomputer (hereinafter abbreviated as “microcomputer”) 20 and a sensor control circuit 30, thereby detecting A / F and sensor elements based on the detection results of the A / F sensor (sensor element 10). Ten impedances (element impedance Zac) are detected.

  In FIG. 1, a microcomputer 20 is composed of a well-known logic operation circuit including a CPU, various memories, an A / D converter, an I / O port, and the like. A current signal (detected by a sensor control circuit 30 described later) Analog signal) is taken in via an A / D converter, and an A / F value calculation and an element impedance Zac calculation are appropriately performed. The A / F value calculated by the microcomputer 20 is output to an engine ECU (not shown), for example, and used for air-fuel ratio feedback control and the like.

  In the sensor control circuit 30, a reference power source 33 is connected to the positive terminal connected to the atmosphere side electrode 16 of the sensor element 10 via an operational amplifier 31 and a current detection resistor 32 as shown in the figure. An applied voltage control circuit 36 is connected to the negative terminal connected to the diffusion layer side electrode 15 via an operational amplifier 34 and a switch 35. In this case, the point A at one end of the current detection resistor 32 is held at the same voltage as the reference voltage Vf (eg, 2.2 V). The element current Ip flows through the current detection resistor 32, and the voltage at the point B changes according to the element current Ip. If the exhaust gas is lean when the switch 35 is ON (closed), the element current Ip flows from the positive side terminal to the negative side terminal in the sensor element 10, so that the point B voltage increases and conversely rich. If so, the element current Ip flows from the negative side terminal to the positive side terminal in the sensor element 10, so that the voltage at the point B decreases.

  The applied voltage control circuit 36 monitors the point B voltage and determines the voltage to be applied to the sensor element 10 according to the voltage value (for example, determined based on the applied voltage characteristic RG in FIG. 3), and the operational amplifier 34 and the switch The D point voltage is controlled via 35. However, when A / F detection is performed only in the vicinity of the stoichiometry, the applied voltage can be fixed.

  Further, an operational amplifier 37 is connected to the reference power source 33, and an output of the operational amplifier 37 and a point B voltage are input to an operational amplifier (differential amplifier) 38 having a predetermined amplification factor. The operational amplifier 38 amplifies the voltage difference between the reference voltage Vf and the point B voltage, and outputs the result as an A / F output voltage AFO. In this case, as a configuration in which the voltage difference between the reference voltage Vf and the B point voltage is amplified in the operational amplifier 38, a configuration in which the A point voltage and the B point voltage are input to the operational amplifier 38 is also conceivable. An electric current flows through the current detection resistor 32, and an error may occur in air-fuel ratio detection. In contrast, in this configuration, since the output of the operational amplifier 37 and the point B voltage are input to the operational amplifier 38, the operational amplifier 37 functions as a feedback current absorption element, and an adverse effect on air-fuel ratio detection can be eliminated.

  A switch 40 and a capacitor 41 are provided on the path from the point B terminal of the current detection resistor 32 to the operational amplifier 38 as shown in the figure. In this case, the switch 40 is turned off (opened) at the time of impedance detection, which will be described later, and the voltage at the point B when the switch is turned off is stored and held in the capacitor 41. As a result, even when the voltage applied to the sensor element 10 changes in an alternating manner during impedance detection, it is possible to avoid the inconvenience that the output of the operational amplifier 38 changes inadvertently due to the influence and adversely affects air-fuel ratio detection. Even when impedance is detected, an appropriate air-fuel ratio output (actually a current signal immediately before the switch is turned off) can be obtained.

  The microcomputer 20 takes in the A / F output voltage AFO from the A / D port, and calculates the A / F value for each time based on the taken A / F output voltage AFO. This A / F value is appropriately used for air-fuel ratio feedback control or the like.

  Further, the microcomputer 20 instructs to temporarily change the applied voltage to the sensor element 10 in an alternating manner, and detects the element impedance Zac based on the current change amount at that time. More specifically, when the impedance is detected, the applied voltage control circuit 36 receives a command from the microcomputer 20, and the applied voltage (point D voltage in the figure) to the sensor element 10 has a predetermined width (for example, 0.2 V) on both positive and negative sides. To change. At this time, the microcomputer 20 measures the change in the B point voltage accompanying the change in the applied voltage, and based on the applied voltage change amount ΔV and the current change amount ΔI obtained by dividing the B point voltage change amount by the resistance value of the current detection resistor 32. The element impedance Zac is calculated (Zac = ΔV / ΔI). In the impedance detection, the current flowing through the sensor element 10 may be changed in an alternating manner, and the element impedance Zac may be calculated from the amount of change in current or voltage at that time.

  Impedance detection is performed at a predetermined cycle (that is, every predetermined time), and the execution timing is commanded from the microcomputer 20 to the applied voltage control circuit 36. Further, the microcomputer 20 controls energization to the heater 18 so that the element impedance Zac is maintained at a predetermined target value. As a result, the sensor element 10 is held in a predetermined active state.

  By the way, in the A / F sensor and the sensor control circuit 30, there are individual differences caused by characteristic variations of the sensors themselves, circuit characteristic variations, and the like, and the sensor output accuracy decreases due to the individual differences. FIG. 4 shows the output characteristic of the A / F output voltage AFO with respect to the element current Ip. In FIG. 4, the original reference output characteristic is indicated by a solid line, and one output characteristic having an offset error with respect to the reference output characteristic is shown. Shown with a chain line. In this case, according to the reference output characteristics, AFO = 2.2V in the stoichiometric state (Ip = 0 mA) and AFO = 4.1 V in the atmospheric state (Ip = 2.6 mA). On the other hand, when an offset error occurs, the A / F output voltage AFO increases by the offset error. In the present embodiment, correction for the offset error is performed in order to eliminate the decrease in accuracy of the sensor output due to the offset error.

  Since it is known that Ip = 0 mA in the stoichiometric state and the reference output is AFO = 2.2 V at that time, the sensor control circuit 30 is forcibly set to the stoichiometric detection state, and an offset error is obtained under this state. To do. Specifically, during normal engine operation, the switch 35 provided on the output side of the operational amplifier 34 is temporarily turned off (opened), and the A / F output voltage AFO at that time is measured. Then, an offset error K1 is calculated from the measured A / F output voltage AFO, and the offset error K1 is appropriately used as a “gas concentration output correction amount” to obtain an AFO output voltage AFO with no error. Yes. The state in which the switch 35 is turned on (closed) corresponds to the “normal state”, and the state in which the switch 35 is turned off (opened) corresponds to the “reference output generation state”.

  FIG. 5 is a flowchart showing the A / F output voltage correction process. This process is executed by the microcomputer 20 at a predetermined time period.

  In FIG. 5, first, in step S101, it is determined whether or not it is a learning timing. This learning timing is a timing at which the sensor control circuit 30 is temporarily set in the stoichiometric detection state to learn the offset error K1, and is determined to be a learning timing every time a predetermined time elapses, for example. For example, the predetermined time may be about 10 minutes, but is not limited thereto, and the predetermined time can be set as appropriate, such as several tens of minutes or 1 to several hours. Further, the learning timing may be set in a state where the A / F value is not used in air-fuel ratio control or the like. For example, the learning timing is set before the activation of the A / F sensor, during fuel cut, and at the time of main relay control after the ignition is turned off. Here, the process of updating and storing data is referred to as “learning”.

  If it is the learning timing, the process proceeds to step S102, a switching signal is output to the switch 35, and the switch 35 is turned off (opened) for a predetermined time (for example, about 5 msec). In the subsequent step S103, the A / F output voltage AFO at that time is captured. In step S104, an offset error K1 is calculated from the captured A / F output voltage AFO and the stoichiometric voltage Vst (K1 = AFO−Vst). The stoichiometric voltage Vst is an AFO reference output that should be output in the stoichiometric state. In this embodiment, Vst = 2.2V. In step S105, the offset error K1 is stored in the standby RAM.

  On the other hand, if step S101 is NO, the process proceeds to step S106, and the A / F output voltage AFO at that time is captured. In subsequent step S107, the A / F output voltage AFO is corrected by the offset error K1 (AFO = AFO−K1). Then, air-fuel ratio feedback control or the like is performed using the corrected A / F output voltage AFO thus calculated.

  According to the embodiment described above in detail, the following excellent effects can be obtained.

  Since the sensor control circuit 30 is temporarily in a stoichiometric detection state by opening the switch 35 and the offset error K1 (gas concentration output correction amount) is learned by the A / F output voltage AFO detected at that time, the offset error K1 is It can detect suitably and can improve the detection accuracy of A / F. Further, in this configuration, unlike the conventional technique in which error detection is performed on the condition that the sensor is inactive, there is no restriction on the learning timing of the offset error K1. Therefore, it becomes possible to learn the output error at an arbitrary timing, and a highly accurate A / F output can be obtained as desired.

  By improving the A / F detection accuracy, the air-fuel ratio stoichiometric control accuracy can be improved, and as a result, improvement of exhaust emission can be realized. In other words, as shown in FIG. 16, if the A / F is originally controlled by stoichiometry, the purification rate of the three components (HC, CO, NOx) in the exhaust gas can be increased. For example, the A / F output characteristic is In the case of A (before correction), the NOx purification rate is particularly deteriorated. On the other hand, if the stoichiometric control can be reliably performed by the above correction, the A / F output characteristic becomes B (after correction), and a high purification rate can be maintained for the three components (HC, CO, NOx) in the exhaust gas.

  Further, since the offset error K1 is sequentially learned (updated and stored), for example, the A / F output voltage AFO can be corrected even immediately after the engine is started, and an appropriate A / F output can be obtained. It will be obtained.

  In the circuit configuration of FIG. 1, the current detection resistor 32 is connected to the positive terminal of the sensor element 10 and the applied voltage control circuit 36 is connected to the negative terminal, but this configuration is changed. For example, the circuit configuration shown in FIG. The sensor control circuit 50 shown in FIG. 6 is basically obtained by replacing the positive side and negative side circuit configurations of the sensor element 10, and FIG. 6 shows only the main configuration. Hereinafter, the difference from FIG. 1 will be mainly described.

  A reference power supply 53 is connected to the negative terminal of the sensor element 10 via an operational amplifier 51 and a current detection resistor 52 as shown in the figure, and an applied voltage control circuit 56 is connected to the positive terminal via an operational amplifier 54 and a switch 55. Has been. In this case, the point A at one end of the current detection resistor 52 is held at the same voltage as the reference voltage (eg, 2.2 V) of the reference power supply 53. The element current Ip flows through the current detection resistor 52, and the voltage at the point B changes according to the element current Ip. When the exhaust gas is lean, the element current Ip flows through the sensor element 10 in the direction of + → −, so that the voltage at the point B decreases. Conversely, when the exhaust gas is rich, the sensor element 10 has the element current Ip in the direction of − → +. Flows, the voltage at point B rises.

  The applied voltage control circuit 56 monitors the point B voltage and performs applied voltage control so as to determine a voltage to be applied to the sensor element 10 according to the voltage value. An operational amplifier (differential amplifier) 57 is connected to points A and B in the figure, and an A / F output voltage AFO is output from the operational amplifier 57 to the microcomputer 20.

  Even in the circuit configuration of FIG. 6, learning of the offset error K1 and correction of the A / F output voltage AFO are performed by the A / F output voltage correction processing of FIG. That is, the sensor control circuit 50 is temporarily in a stoichiometric detection state by opening the switch 55, and the offset error K1 (gas concentration output correction amount) is learned by the A / F output voltage AFO detected at that time. Then, the A / F output voltage AFO detected at the normal time is corrected by the offset error K1.

(Second Embodiment)
In the first embodiment, the method for correcting the offset error of the A / F output voltage AFO has been described. However, in the actual sensor control circuit, in addition to the offset error, the gain error is caused by the gain adjustment deviation of the operational amplifier. Is considered to occur. This will be described with reference to FIG. The gain error is 0 in the stoichiometric state, and increases substantially proportionally as the A / F output voltage AFO increases. In this case, an offset error is included as an AFO output error in the stoichiometric state, and an offset error + gain error is included as an AFO output error in the atmospheric state. In the present embodiment, AFO output errors at two points of the stoichiometric state and the atmospheric state are obtained, and a gain error is calculated from the difference. Then, error correction is performed for both the offset error and the gain error.

  FIG. 8 is a flowchart showing the A / F output voltage correction process in the present embodiment. This process is executed by the microcomputer 20 at a predetermined time period. In carrying out this processing, for example, it is assumed that the offset error K1 is separately calculated according to FIG. 5 and is already stored in the standby RAM.

  In FIG. 8, in step S201, it is determined whether or not the fuel is currently being cut, and the process proceeds to step S202 on the condition that the fuel is being cut. At this time, the fact that the fuel is being cut corresponds to the fact that the learning condition for the atmospheric gain error K2 is satisfied. In step S202, the A / F output voltage AFO at that time is captured, and in step S203, the atmospheric error amount ΔV is calculated from the captured A / F output voltage AFO and the atmospheric specified voltage Vair (ΔV = AFO− Vair). The atmospheric regulation voltage Vair is an AFO reference output that should be output in the atmospheric state (fuel cut state), and in this embodiment, Vair = 4.1V.

  Thereafter, in step S204, the atmospheric gain error K2 is calculated by subtracting the offset error K1 from the atmospheric error amount ΔV (K2 = ΔV−K1). In step S205, the atmospheric gain error K2 is stored in the standby RAM. That is, the atmospheric gain error K2 is learned.

  On the other hand, if step S201 is NO, that is, if the fuel is not being cut and is in normal control, the process proceeds to step S206, and the A / F output voltage AFO at that time is taken in. In step S207, the offset error K1 and the atmospheric gain error K2 are read. In the subsequent step S208, the AFO correction value Kaf is calculated from these errors K1 and K2. At this time, the gain error is stoichiometric time = 0 and atmospheric time = K2, and the gain error K2a corresponding to the air-fuel ratio in each case is calculated by interpolation or the like. Then, the AFO correction value Kaf is calculated from the offset error K1 and the gain error K2a (Kaf = K1 + K2a).

  Thereafter, in step S209, the A / F output voltage AFO is corrected by the AFO correction value Kaf (AFO = AFO−Kaf). Then, air-fuel ratio feedback control or the like is performed using the corrected A / F output voltage AFO thus calculated.

  As described above, according to the second embodiment, the offset error K1 (first error amount) calculated with the sensor control circuit 30 in the stoichiometric detection state and the atmospheric gain error K2 calculated under the atmospheric condition at the time of fuel cut. Since the A / F output voltage AFO is corrected based on the (second error amount), the detection accuracy of the A / F can be improved.

  Further, since the offset error K1 and the atmospheric gain error K2 are sequentially learned, the A / F output voltage AFO can be corrected, for example, immediately after the engine is started. Output will be obtained.

(Third embodiment)
In the above embodiment, the A / F sensor having the sensor element structure of FIG. 2 has been described. However, the present invention can be applied to an A / F sensor having another sensor element structure. The sensor element 60 shown in FIG. 9 has two layers of solid electrolytes 61, 62. A pair of electrodes 63, 64 are disposed opposite to one solid electrolyte 61, and a pair of solid electrolytes 62 are disposed on the other solid electrolyte 62. Electrodes 65 and 66 are arranged to face each other. Note that the electrodes 63 to 65 are seen in two places on the left and right objects in the figure, but they are the same member connected at any part of the front and back of the page. In this sensor element 60, the solid electrolyte 61 and the electrodes 63 and 64 constitute a pump cell 71 as a “first cell”, and the solid electrolyte 62 and the electrodes 65 and 66 constitute a monitor cell 72 as a “second cell”. Yes. Each electrode 63 to 66 is connected to a sensor control circuit 80. The sensor element 60 has the laminated structure as in the sensor element 10 described above. In FIG. 9, reference numeral 67 is a gas introduction hole, reference numeral 68 is a porous diffusion layer, reference numeral 69 is an atmospheric duct, and reference numeral 70 is a heater. The monitor cell 72 is generally called an electromotive force cell or an oxygen concentration detection cell.

  In the A / F sensor having the above sensor element structure, the monitor cell 72 generates a binary (0 V or 0.9 V) electromotive force output depending on whether the exhaust gas is lean or rich with respect to the stoichiometry. For example, when it is lean, the electromotive force output of the monitor cell 72 is small, and conversely, when it is rich, the electromotive force output of the monitor cell 72 is large. In such a case, the applied voltage of the pump cell 71 is controlled so that the electromotive force output of the monitor cell 72 becomes the stoichiometric value (0.45 V).

  FIG. 10 is a circuit diagram showing a configuration of the sensor control circuit 80 for the A / F sensor having the sensor element structure of FIG. In FIG. 10, VM is a common terminal for the pump cell 71 and the monitor cell 72, and a reference voltage power supply 81 is connected to the common terminal VM. The reference voltage of the reference voltage power supply 81 is, for example, 2.5V. IP is a pump cell terminal connected to the electrode 63 of the pump cell 71, and UN is a monitor cell terminal connected to the electrode 66 of the monitor cell 72. A closed circuit having an operational amplifier 82 and a current detection resistor 83 is connected to each of the terminals IP and UN through the cells 71 and 72, and a non-inverting terminal (+ terminal) of the operational amplifier 82 has a reference voltage (3.0V). ) Is connected.

  When lean, a current flows through the current detection resistor 83 in the direction of B → A. Conversely, when rich, a current flows through the current detection resistor 83 in the direction of A → B. In such a case, the pump cell 71 is feedback-controlled so that the output voltage of the monitor cell 72 becomes a predetermined value (however, various feedback control circuits have already been disclosed, and illustration and detailed description thereof are omitted here. ).

  An operational amplifier 85 having a predetermined amplification factor is connected to both terminals A and B of the current detection resistor 83. The output of the operational amplifier 85 is an A / F output voltage AFO.

  A switch 87 is provided between the input terminals on both the positive and negative sides of the operational amplifier 82, a switch 88 is connected to the common terminal VM, and a switch 89 is connected to the monitor cell terminal UN. The switch 87 is a normally open switch, and the open / close state is controlled by the switching signal 1. The switches 88 and 89 are both normally closed switches, and the open / close state is controlled by the switching signal 2.

  In the above configuration, when the normal A / F is detected, the switch 87 is turned off (opened) and the switches 88 and 89 are turned on (closed), so that the A / F output voltage AFO corresponding to each A / F is obtained. It can be measured. At the learning timing, the switch 87 is temporarily turned on (closed) and the switches 88 and 89 are turned off (opened), so that the offset error of the sensor control circuit 80 can be calculated from the A / F output voltage AFO at that time. Although explanation is omitted, the calculation of the gain error can be performed in the same manner as described above.

  In the sensor control circuit 80 shown in FIG. 10, the point A voltage and the point B voltage at both terminals of the current detection resistor 83 are not fixed and fluctuate. However, in the sensor control circuit 90 shown in FIG. One terminal voltage can be fixed.

  In FIG. 11, a voltage (for example, 3 V) equivalent to the reference voltage Vf <b> 1 is applied to the common terminal of the pump cell 71 and the monitor cell 72 through the operational amplifier 93. That is, the point B voltage in the figure is fixed at 3V. Further, a closed circuit having a feedback circuit 91 and a current detection resistor 92 is configured through the monitor cell 72. The reference voltage Vf2 in the feedback circuit 91 is, for example, 2.55V.

  The operation of the sensor control circuit 90 will be described taking the rich time as an example. When rich, the C1 point in the figure rises to 3.45 V, for example, due to the electromotive force of the monitor cell 72, so the potential at the C2 point in the feedback circuit 91 decreases. Then, the output of the feedback circuit 91, that is, the point A voltage increases. That is, when rich, a current flows through the current detection resistor 92 in the direction of A → B. On the contrary, current flows through the current detection resistor 92 in the direction of B → A during lean. An operational amplifier 94 having a predetermined amplification factor is connected to both terminals A and B of the current detection resistor 92.

  A switch 96 is provided between the input terminals on both the positive and negative sides of the operational amplifier 95 in the feedback circuit 91, and a switch 97 is provided between the monitor cell 72 and the feedback circuit 91. The switch 96 is a normally open switch, and the open / close state is controlled by the switching signal 1. The switch 97 is a normally closed switch, and the open / close state is controlled by the switching signal 2.

  With the above configuration, when the normal A / F is detected, the switch 96 is turned off (opened) and the switch 97 is turned on (closed), whereby the A / F output voltage AFO corresponding to the A / F can be measured. . At the learning timing, the switch 96 is temporarily turned on (closed) and the switch 97 is turned off (opened), so that the offset error of the sensor control circuit 90 can be detected from the A / F output voltage AFO at that time.

  An A / F sensor having another configuration is shown in FIG. The sensor element 100 of FIG. 12 includes three layers of solid electrolytes 101, 102, and 103, and a pair of electrodes 104 and 105 are disposed opposite to the solid electrolyte 101, and a pair of electrodes 106 and 107 are disposed on the solid electrolyte 102. Opposed. In the present sensor element 100, the solid electrolyte 101 and the electrodes 104 and 105 constitute a pump cell 111 as a “first cell”, and the solid electrolyte 102 and the electrodes 106 and 107 constitute a monitor cell 112 as a “second cell”. Yes. Further, the solid electrolyte 103 constitutes a wall material for securing the oxygen reference chamber 108. The sensor element 100 has a laminated structure, which is the same as the sensor element 10 described above. In FIG. 12, reference numeral 109 denotes a porous diffusion layer, and reference numeral 110 denotes a gas detection chamber. The monitor cell 112 is generally called an electromotive force cell or an oxygen concentration detection cell, similarly to the monitor cell 72 of FIG. The present invention can be similarly applied to the sensor element 100 having the above configuration.

  Furthermore, the present invention can be applied not only to an A / F sensor having a laminated structure but also to an A / F sensor having a cup structure. The present invention can also be applied to a so-called O2 sensor in which an electromotive force is generated between the electrodes of the sensor element in accordance with the oxygen concentration in the exhaust gas.

  In addition to the A / F sensor whose oxygen concentration is a detection target, the present invention can be applied to a gas concentration sensor whose detection target is another component concentration. For example, a composite type gas concentration sensor has a plurality of cells formed of a solid electrolyte body, of which a pump cell discharges or draws oxygen in a detection gas introduced into the chamber and detects the oxygen concentration. In the sensor cell, the specific component concentration is detected from the gas after the oxygen is discharged. In addition, a gas concentration sensor having a monitor cell that outputs an electromotive force signal according to the residual oxygen concentration in the chamber may be used. This gas concentration sensor is embodied as, for example, a NOx sensor that detects the NOx concentration in the exhaust gas. By applying the present invention, it is possible to detect the NOx concentration suitable for the NOx sensor.

  The present invention is not limited to the description of the above embodiment, and may be implemented as follows, for example.

  The configuration shown in FIGS. 13 and 14 can be applied as a configuration in which the sensor control circuit is temporarily in a stoichiometric detection state. 13 and 14 show the configuration of the main part on the minus terminal side connected to the sensor element 10 in the configuration of FIG. FIG. 13 shows a detailed circuit configuration of the operational amplifier 34 connected to the sensor terminal side. Transistors 151 and 152 (output stage switching elements) provided in the output stage are operated by the transistors 153 and 154. The state is controlled to be switched. In this case, by turning on the switching signal, the transistors 153 and 154 are turned on, and the operations of the transistors 151 and 152 are turned off. As a result, the sensor control circuit is temporarily brought into the stoichiometric detection state.

  In FIG. 14, two MOSFETs 163 and 164 are provided between two transistors 161 and 162 with a connection point to the sensor terminal interposed therebetween, and the gate inputs of the MOSFETs 163 and 164 are switched and controlled by a switch 165. It is like that. In this case, the switch 165 is turned on by turning on the switching signal, and the MOSFETs 163 and 164 are turned off. As a result, the electrical path connected to the sensor terminal is opened, and the sensor control circuit is temporarily brought into the stoichiometric detection state.

  In FIG. 15, the sensor control circuit is in a stoichiometric detection state by setting the connection terminals on both the positive and negative sides connected to the sensor element 10 to the same potential. FIG. 15 shows a part of the configuration shown in FIG. That is, the switch 171 is provided between the applied voltage control circuit 36 and the operational amplifier 34, and the switch 171 is controlled to be switched by the switching signal. In this case, by switching the switch 171 from the x contact to the y contact, both connection terminals of the sensor element 10 have the same potential (both Vf), and the sensor applied voltage = 0V. Thereby, the sensor control circuit is brought into a stoichiometric detection state with an element current = 0 mA.

  When calculating the offset error, a reference output generation state other than the stoichiometric detection state can be set. For example, the reference output is generated in the vicinity of the stoichiometric state (however, in order to eliminate the influence of the gain error as much as possible, it is desirable to be in the vicinity of the stoichiometric as much as possible).

  Further, when calculating the gain error, it may be determined that the exhaust gas atmosphere has become an atmosphere other than the air, and the gain error may be calculated under the atmosphere. In short, the reference gas atmosphere is different from the gas atmosphere corresponding to the stoichiometric detection state (equivalent to the reference output generation state), and the original reference output at that time is known.

  When the temperature of the sensor control circuit changes, the output error amount changes because of a change in resistance value or the like. Therefore, the temperature of the sensor control circuit may be estimated (or detected) each time, and the offset error K1 may be stored and held for each temperature region when learning the offset error. Similarly, the gain error may be learned for each temperature region.

  For example, when the vehicle travels uphill, it is considered that the throttle ON state continues and the ECU temperature rises. Therefore, when the throttle ON state continues for a predetermined time or longer, it may be determined that it is the learning timing, and the learning of the output error amount (for example, the processing in steps S102 to S105 in FIG. 5) may be performed. In short, the temperature change of the sensor control circuit is monitored, and the output error amount is learned when the temperature change exceeds a predetermined value.

  Further, learning of the output error amount (for example, the processing in steps S102 to S105 in FIG. 5) may be performed at least one of immediately after the engine is started and when it is stopped. For example, the start-up learning of the output error amount is performed at a timing when about 0.5 seconds have elapsed after turning on the ignition switch. Further, when the engine is stopped, the learning of the output error amount is performed again in the main relay control that is executed by temporarily delaying the power-off to the microcomputer 20 or the like.

  In the second embodiment, the offset error K1 (first error amount) calculated with the sensor control circuit 30 in the stoichiometric detection state, and the atmospheric gain error K2 (first error) calculated under atmospheric conditions at the time of fuel cut. 2), the A / F output voltage AFO is corrected on the basis of the error amount. For example, instead of the configuration in which the sensor control circuit 30 is in the stoichiometric detection state and the offset error K1 (first error amount) is calculated, it is determined that the A / F sensor is in an inactive state immediately after the engine is started or stopped. The offset error K1 (first error amount) is calculated. In short, the first and second error amounts are calculated based on the gas concentration output generated in the sensor circuit and the original reference output at that time in two predetermined gas atmospheres or sensor circuit states corresponding to the gas atmosphere. Any configuration that corrects the gas concentration output based on the first and second error amounts may be used. If it is a requirement that the sensor is inactive, there is a timing constraint on the calculation of the first error amount. However, if the first error amount is calculated once at the time of engine start or the like, then the first error amount is calculated. Each time the error amount of 2 is calculated, the gas concentration output can be appropriately corrected.

  By the way, in a diesel engine, it is desired to reduce PM (particulate) and NOx contained in exhaust gas. In this case, the relationship between the PM concentration and the NOx concentration with respect to the O2 concentration is as shown in FIG. 17, and if the O2 concentration is controlled at the P1 point (for example, 10%), the concentrations of both PM and NOx can be maintained at a relatively low level. . However, for example, when the O2 concentration shifts to the P2 point due to an output error of the sensor control circuit, the NOx purification rate is particularly deteriorated. On the other hand, according to the embodiment of the present invention, the O2 concentration can always be controlled at a desired P1 point, so that the emission of PM and NOx can be suppressed and the exhaust emission can be improved.

  DESCRIPTION OF SYMBOLS 10 ... Sensor element, 11 ... Solid electrolyte, 20 ... Microcomputer, 30 ... Sensor control circuit, 32 ... Current detection resistor, 34 ... Operational amplifier, 35 ... Switch, 50 ... Sensor control circuit, 52 ... Current detection resistor, 55 ... Switch, DESCRIPTION OF SYMBOLS 60 ... Sensor element 61, 62 ... Solid electrolyte, 71 ... Pump cell, 72 ... Monitor cell, 80 ... Sensor control circuit, 87-89 ... Switch, 90 ... Sensor control circuit, 96, 97 ... Switch, 100 ... Sensor element, 101 DESCRIPTION OF SYMBOLS -103 ... Solid electrolyte, 111 ... Pump cell, 112 ... Monitor cell, 151,152 ... Transistor, 163,164 ... MOSFET, 171 ... Switch.

Claims (10)

A gas concentration sensor comprising a solid electrolyte body disposed in an exhaust system of an internal combustion engine, and a sensor circuit for measuring a current flowing when voltage is applied to the gas concentration sensor and generating a gas concentration output according to the measured current In the gas concentration detection device
State switching means for temporarily switching the sensor circuit from a normal state in which gas concentration detection is performed on exhaust gas from the internal combustion engine to a reference output generation state in which a predetermined reference output is generated regardless of the exhaust gas atmosphere; ,
A correction amount calculating means for calculating a gas concentration output correction amount based on a difference between a gas concentration output generated at that time and an original reference output after switching the sensor circuit to the reference output generation state;
With
Switching to a reference output generation state by the state switching means is carried out when the temperature of the sensor circuit changes more than a predetermined value.   2. The gas concentration detection according to claim 1, wherein, when switching to the reference output generation state, the state switching unit sets the sensor circuit to a stoichiometric detection state in which the same stoichiometric reference output is generated as when detecting gas after stoichiometric combustion. apparatus. In the sensor circuit, switch means is provided in a current path connected to the gas concentration sensor,
The gas concentration detection device according to claim 2, wherein the state switching unit sets the sensor circuit to the stoichiometric detection state by opening the switch unit.   3. The gas concentration according to claim 2, wherein the state switching unit forcibly turns off an output stage switch element of an amplification circuit connected to the gas concentration sensor, thereby bringing the sensor circuit into the stoichiometric detection state. Detection device. Of the connecting terminals on both the positive and negative sides connected to the gas concentration sensor, one is connected to a current detection resistor for measuring the current flowing through the gas concentration sensor, and the other is used to open or close the sensor current path. Connect the switch means,
The gas concentration detection device according to claim 2, wherein the state switching unit sets the sensor circuit to the stoichiometric detection state by opening the switch unit.   The gas concentration detection device according to claim 2, wherein the state switching unit sets the sensor circuit in the stoichiometric detection state by setting connection terminals on both positive and negative sides connected to the gas concentration sensor to the same potential.   The gas concentration detection device according to any one of claims 1 to 6, wherein the switching to the reference output generation state by the state switching means is performed in a state where the gas concentration output is not required in the control system of the internal combustion engine.   The gas concentration detection device according to any one of claims 1 to 7, wherein the switching to the reference output generation state by the state switching means is performed at least immediately after the start of the internal combustion engine and at a stop.   The gas concentration detection device according to claim 1, further comprising an update unit that sequentially updates and stores the gas concentration output correction amount calculated by the correction amount calculation unit.   The gas concentration detection device according to claim 9, wherein the updating unit stores the gas concentration output correction amount in association with a temperature of the sensor circuit.

Images