EP4064792B1 - Headlight provided with improved dynamic lighting - Google Patents

Description

Technical Domain

The present invention relates to the field of headlamps equipped with so-called Reactive Lighting technology, and in particular a headlamp comprising an accelerometric sensor.

State of the art

The applicant for the present patent application has marketed a portable lamp, of the headlamp type, equipped with so-called reactive or dynamic lighting, the operating principle of which is illustrated in the figure 1 . This headlamp features an electronic circuit with a sensor that analyzes the external brightness to instantly deliver a set lighting power and beam shape optimal for the situation.

This type of lamp has proven to be particularly suitable for committed and intensive sports because it relieves the user of the manual mode adjustments that would be necessary to switch between different beam power thresholds.

Thanks to this reactive lighting technique, the user has their hands free and their mind totally focused on their activity, whatever the lighting situation considered.

Thus, in proximity lighting , the user can observe or examine an object at short distance (reading a map, making a rope knot or pitching a tent for example) and the lamp can protect a Very wide and low-power light beam, automatically set to a minimum threshold value thanks to this dynamic lighting technique. The lighting automatically adapts to the distance of the object.

On the contrary, in a situation of movement , for example when the user is walking and/or running, the beam becomes mixed: wide at the level of the feet and focused to see a few meters and anticipate the relief.

In addition, when in a distant vision situation, the user raises his head to see into the distance - for example to look for a beacon during a run or a relay attached to a wall, the lighting power increases considerably and the beam becomes focused to best assist the user of the lamp.

Finally, we note that reactive or dynamic lighting technology ( Reactive Lighting) has proven to be particularly economical in use and allows for advantageously increasing battery life since its implementation, under the control of a computer, aims to optimize battery consumption, offering greater battery life for your lamp.

As we can see, this reactive or dynamic lighting technology undeniably constitutes a significant advance in the field of headlamps, and more generally of portable lighting, in particular in that it allows the lighting to be constantly adapted to the lighting conditions.
However, practitioners have noted drawbacks in certain very specific situations.
Indeed, in so-called trail or running activities, the presence of numerous reflective surfaces on shoes, technical clothing and signage causes pumping phenomena in the projected light, thus degrading the quality of the lighting and creating an area of discomfort for the user.
When cycling with a headlamp, or even other very dynamic activities (ski touring or others), the minimum level of light may prove to be insufficient in practice to guarantee safe lighting when crossing a luminous or natural obstacle (car headlights, tree branches, etc.). The discomfort zone noted above may then turn out to be a danger zone.
These are the defects and drawbacks which the present invention aims to remedy.
US 2020/187331 A1 (BERTKEN DENNIS [US] ET AL ) describes a lantern or flashlight type lighting device having an accelerometer, but does not describe a headlamp, nor does it describe or suggest the use of a headlamp control module configured to select an activity profile chosen from a set of predetermined physical activity profiles. This prior art does not describe or suggest the use of a LUT lookup table providing at least one value or parameter for adjusting the power regulation of the lamp.
WO 2012/146256 A2 (LIGHTEN APS [DK], PEDERSEN STEEN HVIDTFELDT HESSELLUND [DK]) describes an ambient lighting system for initiating a circadian or well-being change. This prior art describes the use of a human or mammal activity sensor or detector and the use of an accelerometer with processing means to determine whether the person is running, walking, sitting, typing, lying down, etc. However, this document does not describe or suggest the use of the accelerometer and the light source within a single component that constitutes a headlamp attached to a user's head.
US 2016/258599 A1 (GENTHON FABIEN [FR]) describes a headlamp with a circuit for adjusting the geometry of the light beam.

Summary of the invention

The present invention aims to provide a significant improvement to dynamic lighting technology by making it possible to take into consideration specific lighting situations requiring additional lighting.

Another object of the present invention is to provide a headlamp having reactive or dynamic type light regulation having improved adjustment of light power.

It is another object of the present invention to provide a headlamp improved by the addition of an accelerometer to refine the reactive or dynamic regulation mechanism used by the lamp.

• une source lumineuse ;
• un module de puissance pour l'alimentation de la source lumineuse à partir d'une information ou un signal de commande;
• un module de commande pour le réglage de la puissance générée par la source lumineuse, comportant :
  un capteur de lumière permettant de capter la lumière de l'environnement du porteur de la lampe, le module de commande étant configuré pour générer l'information ou le signal de commande en fonction de l'information générée par le capteur de lumière.

The invention achieves these aims by means of a lamp, such as a headlamp defined in claim 1, comprising

• a light source;
• a power module for supplying the light source from information or a control signal;
• a control module for adjusting the power generated by the light source, comprising:
  a light sensor for sensing light from the environment of the wearer of the lamp, the control module being configured to generate the information or the control signal based on the information generated by the light sensor.

The control module further comprises an accelerometer configured to provide at regular intervals data representative of an acceleration of the headlamp along at least one horizontal axis and one vertical axis; and wherein the control module comprises a circuit configured to store and process accelerometry data in order to select a physical activity profile chosen from a set of predetermined physical activity profiles stored in a memory.

The selected physical activity profile is then used as an input pointer to read a LUT lookup table stored in an internal memory of the lamp, and which provides at least one value or parameter used to generate information or the control signal setting the light output. Such that the value or parameter read from the LUT lookup table is used together with the information generated by the light sensor to determine the information or the control signal for the light output.

Preferably, the set of predetermined accelerometer profiles include profiles representative of walking, running and cycling.

Preferably, the power of the light beam set by the control unit varies between a low threshold and a high threshold, and the low threshold is set by a value that is extracted directly from the LUT lookup table from the automatically selected profile.

Preferably, the processing of the accelerometer data allowing the selection of the predetermined profile uses a statistical processing method based on a calculation of the variance of the accelerometer data along the two horizontal axes and along the vertical axis.

The invention also allows the production of a light regulation method for a headlamp as defined in claim 10.

In a particular embodiment, the data extracted from the LUT table makes it possible to define a minimum threshold of luminous power and a specific geometry of the light beam chosen between a wide beam, a pointed beam and/or both.

Preferably, the lamp is a headlamp configured to process accelerometer data to detect a user's fall in addition to their physical activity, and configured to communicate with a mobile phone for the purpose of transmitting an alert message.

In a particular embodiment, in the event of a fall, the control module is configured to command a light alert sequence aimed at calling for help.

Description of the drawings

• La figure 1 illustre le schéma de principe de l'éclairage dynamique ou réactif.
• La figure 2 représente un mode de réalisation d'une lampe frontale conforme à la présente invention, qui incorpore un capteur de luminosité ainsi qu'un capteur accélérométrique pour fixer les seuils d'éclairage réactif ou dynamique.
• Les figures 3a à 3c illustrent les chronogrammes typiques, sur les trois axes x, y et z, des signaux accélérométriques pour trois activités physiques considérées: marche à pied, bicyclette et course à pied (jogging).
• Les figures 4a, 4b et 4c illustrent plus particulièrement les accélérations suivant chacun des axes xx', yy' et zz' et ce pour chacune des trois activités physiques considérées dans un mode de réalisation.
• La figure 5 illustre un mode de réalisation d'un procédé de régulation de puissance d'éclairage conforme à la présente invention.
• La figure 6 illustre le réglage du seuil minimum d'éclairage dyanmique en fonction de l'activité physique détectée.
• La figure 7 illustre un procédé de traitement du signal et de détermination du profil de mouvement du capteur d'accélération 3D
• Les figures 8A, 8B, 8C et 8D illustrent respectivement les signaux utiles du triplet (S1 _u (t, µ), S2 _u (t, µ), S3 _u (t, µ)) pour différents profils de mouvements µ0,µ1,µ2 et µ3 du capteur d'accélération 3D 110.

Other characteristics, aims and advantages of the invention will appear on reading the description and the drawings below, given solely as non-limiting examples. In the attached drawings:

• There figure 1 illustrates the basic diagram of dynamic or reactive lighting.
• There figure 2 depicts an embodiment of a headlamp according to the present invention, which incorporates a brightness sensor as well as an accelerometric sensor to set the thresholds for reactive or dynamic lighting.
• THE Figures 3a to 3c illustrate typical chronograms, on the three axes x, y and z, of the accelerometric signals for three physical activities considered: walking, cycling and running (jogging).
• THE Figures 4a, 4b and 4c more particularly illustrate the accelerations along each of the axes xx', yy' and zz' and this for each of the three physical activities considered in one embodiment.
• There figure 5 illustrates an embodiment of a lighting power regulation method according to the present invention.
• There figure 6 illustrates the adjustment of the minimum dynamic lighting threshold based on the physical activity detected.
• There figure 7 illustrates a method of signal processing and determining the motion profile of the 3D acceleration sensor
• THE Figures 8A, 8B, 8C and 8D respectively illustrate the useful signals of the triplet ( S 1 _u ( t , µ ), S 2 _u ( t, µ ), S 3 _u ( t, µ )) for different motion profiles µ 0, µ 1, µ 2 and µ 3 of the 3D acceleration sensor 110.

Description of preferred embodiments

We now describe how we can significantly improve a headlamp equipped with a reactive or dynamic lighting system, such as that marketed in the “RL” lamps of the PETZL company, for example the headlamps marketed under the name NAO ^™ , or SWIFT RL ^™ , and which include an automatic mechanism for regulating the power generated based on information produced by a brightness sensor.

By virtue of the present invention, the light power regulation mechanism is arranged so as to integrate, in addition to the information emanating from the brightness sensor, other complementary information generated by an accelerometric sensor providing acceleration signals on one or more axes X1, Y1 or Z1.

A specific algorithm, which will be described in detail below, makes it possible to set lighting thresholds generated by the light power regulation system, and in particular a minimum lighting threshold.

I. General architecture

There figure 2 illustrates the general architecture of an embodiment of a lamp 100 - assumed to be frontal - comprising a reactive or dynamic regulation system of the light intensity based on a sensor 120 making it possible to measure the ambient brightness and/or a part of the flux reflected by the lighting of the frontal lamp.

The lamp 100 also comprises an accelerometric sensor, and preferably a three-dimensional (3D) acceleration sensor 110 making it possible to generate accelerometric information along at least one axis and preferably three axes X1, Y1, Z1 in particular illustrated in the figure. figure 8 , the X1 and Z1 axes being horizontal and the Y1 axis being vertical.

More specifically, the lamp 100 comprises a power module 210 associated with a control module 220 and a lighting unit 230 comprising at least one light-emitting diode LED and, optionally, a transceiver module 240 coupled to the control module and a battery module 250 also coupled to the control module 220.

In the example of the figure 2 , the lighting unit 230 comprises a single LED diode 231 with its power supply circuit 232 connected to the power module 210. Clearly, several diodes may be considered to obtain a high-luminosity beam. Generally speaking, the LED diode(s) may be associated with a specific focal optic 233 making it possible to ensure collimation of the generated light beam.

In a specific embodiment, the power supply of the LED diode 231 via the circuit 232 is carried out by the power module under the control of information or a control signal generated by the control module 220 via a link which may take the form of a conductor or a set of conductors constituting a bus. The figure shows more particularly the particular example of a conductor 225.

The power module 210 specifically comprises all the components that are conventionally found in an LED lighting lamp for producing a high-intensity light beam, and generally based on Pulse Width Modulation (PWM), well known to a person skilled in the art and similar to that found in class D audio circuits. This PWM modulation is controlled by means of the control signal 225 generated by the control module 220. Generally speaking, it will be noted that the term " signal " mentioned above refers to an electrical quantity - current or voltage - making it possible to cause the control of the power module, and in particular the PWM modulation used to supply current to the LED diode 231. This is only a particular embodiment, it being understood that it will be possible to substitute for the " control signal 225 " any " control information ", for example logic information stored in a register and transmitted as has been said by any means suitable for the power module 210 for the purpose of controlling the emission power of the light beam. The control signal can therefore be emitted on different media depending on whether it is a signal or information. These media can be a bus-type communication line coupling the control module and the power module or a simple electronic circuit for transferring a control voltage or intensity. In a particular embodiment, it could even be envisaged that the two control and power modules are integrated into the same module or integrated circuit.

A person skilled in the art will therefore easily understand that when referring to a " control signal 225" , this includes without distinction the embodiments using an electrical control quantity - current or voltage - as well as the embodiments in which the control is carried out by means of logic information transmitted within the power circuit. For this reason, we will hereinafter speak without distinction of control signal or information .

Generally speaking, the components that make up the power module 210 - switches and circuits - are well known to a person skilled in the art and the description will be deliberately shortened in this regard for the sake of conciseness. Likewise, the reader will be referred to general works dealing with the various aspects of MLI (or PWM) modulation.

Coming back to the figure 2 , it can be seen that the control module 220 comprises a processor 221 as well as volatile memories 222 of the RAM type and non-volatile (flash, EEPROM) 223 as well as one or more input/output circuits 224. The RAM and non-volatile memories are intended for storing data and micro-program or firmware instructions. Furthermore, the non-volatile memory 223 is also used for storing data representative of physical activity profiles which will be used in conjunction with the accelerometer data provided by the accelerometer sensor 110 as will be described later.

The headlamp also includes a battery module 250 having a controller 252 and a battery 251, for example of the Ion-Lithium type.

Generally speaking, the control module 220 can access each of the other modules present in the lamp, and in particular the power module 210, the battery module 250, the two brightness sensors 120 and accelerometer 110 as well as, where appropriate, the communication module 240 allowing two-way (upstream - downstream) wireless communication with a smartphone 300 or any other wireless communication device.

Access of the control module 220 to the various components of the headlamp may take various forms, either by means of specific circuits and/or conductors or a set of conductors forming a bus. As an illustration, the link 225 is shown in the figure 2 in the form of a conductor while a real data/address/command bus 226 is used for the exchange of information between the control module 220, the battery module 250 and the transmitter/receiver module 240. This is however only a particular embodiment, it being understood that a person skilled in the art will be able to make various modifications and/or adaptations where appropriate to take into account the requirements specific to the intended application.

By accessing the various modules making up the headlamp, the control module 220 can both read and collect information contained in each of these modules and/or conversely, transfer information, data and/or commands there, as will become more clear in the remainder of the presentation.

This is how the control module 220 can send to the power module a control signal as represented by the signal transmitted on the link 225 and, more generally, can read the current value of the supply current of the diode 231 passing via the conductors 232 (via circuits and/or buses not shown in the figure).

Similarly, the control module 220 can access the battery module 250 via the bus 226 to read either the different voltage values (depending on the charge or discharge cycle in progress) at the terminals thereof and/or the value of the intensity delivered in order to be able to calculate a state of charge (SOC or State Of Charge in English literature).

Il. Communication module 240

The control module 220 comprises a communication module 240 enabling a wireless bidirectional connection with a mobile information processing system or mobile telephone 300. In a preferred embodiment, the transmitter and the receiver will be compatible with the Bluetooth standard, preferably with the Bluetooth 4.0 Low energy standard. In another embodiment, the WIFI or IEEE802.11 standard will be adopted instead. The module 240 comprises a baseband unit (not illustrated) coupled to a receiver and a wireless transmitter, enabling an uplink communication channel to be organized towards the mobile telephone 300 and, in the opposite direction, a downlink communication channel towards this same telephone. For this purpose, the communication module 240 may be required to perform various processing operations, in series or in parallel, on the digital representation of the signal received and to be transmitted, and in particular, filtering, statistical calculation, demodulation, channel coding/decoding operations making it possible to make the communication robust to noise, etc. Such operations are well known in the field of signal processing, particularly when it comes to isolating a particular component of a signal, likely to carry digital information, and it will not be necessary here to weigh down the description.

Once detected, these packets are transferred to the processor 221 within the control module 220.

The processor 221 is therefore responsible for interpreting the packets received as well as formatting packets for transmission according to a format specific to the standard used. Thus, in the case of the Bluetooth Low Energy standard, these packets will have a structure around the standardized Generic Attribute Profile (GATT) that will not be detailed here. Depending on the interpretation of the data bits included in the packets received, the processor will reconstruct any information or commands received on the downlink from the mobile information processing system 300. Having interpreted this information or commands, the processor 221 will then relay or convert this information or command to the module concerned. Thus in the basic embodiment, the processor 221 identifies commands for the power module 210 in order to modify the light intensity and in response to this identification is capable of generating control information on the conductor 225 intended for the power module 210 so that the latter proceeds to modify the light intensity generated by the lighting unit 230.

Additionally, the processor 221 may also identify read requests issued by the associated mobile information processing system 300 so that the headlamp sends certain parameters to the telephone 300 on the uplink.

These requests can thus be a request for the state of charge of the battery or the value of the current light power. In this case, the processor 221 will retrieve the necessary information directly from the module concerned and after having carried out any additional calculations on this information to obtain the final required information (in the case of the state of charge for example as seen above), will format a corresponding data packet for transmission by the transceiver module 240.

It is clear that the figure 2 describes a basic embodiment, and that many other embodiments are possible and within the reach of a person skilled in the art. For example, in a more sophisticated mode, other modules may be added within the headlamp and these modules will also be coupled to the processor 221 via the bus 226 for example. These modules will then also be able to exchange data or commands in uplink or downlink with the associated mobile information processing system 300 which will then be able to communicate with the headlamp and transmit various configuration commands to it by means of a dedicated application running on the smartphone. This dedicated application then makes it possible to coordinate the various functionalities of the headlamp by offering in particular a user-friendly interface with the user by means of which the latter can either enter operating parameters or directly control the headlamp or select different options for the functionalities offered.

III. Dynamic or reactive lighting regulation

The control module 220 of the headlamp 100 implements a dynamic or reactive lighting technique. This technique consists of substituting the well-known manual adjustment modes - based on various preset light power values such as low , medium or high , a more automatic technique allowing the adjustment of the light power to be left to the control module 220 and more specifically to a regulation algorithm executed by the processor 221 under the control of a regulation micro-software stored in non-volatile memory 223.

Following the principle of dynamic or reactive lighting, the processor 221 adjusts the light power according to the value of the ambient brightness measured by the sensor 120, for example by selecting a value chosen from a set of N predefined threshold values. Such a mechanism of regulation is therefore similar to a mechanism for adjusting by discrete steps within a finite set of power values, allowing the control module 220 to control the headlamp by successively moving from one adjustment value to another value chosen from the set of predetermined values.

With a set of three predetermined adjustment values, corresponding to three powers, for example “ low ”, “ medium ” or “ high ”, the reactive or dynamic brightness mechanism therefore allows the automatic adjustment of the headlamp to the correct value within the N predetermined values.

In the same way, the geometry of the light beam can be adjusted automatically by selecting, via the control module 220, a diffusion mode chosen from a set of several predetermined modes: for example wide , narrow , or both at the same time.

Such dynamic or reactive regulation, by discrete steps, proves to be simple and inexpensive to implement and allows automatic switching between predefined threshold values.

However, a person skilled in the art may envisage a more sophisticated regulation mechanism based on a real servocontrol integrating the brightness value within a feedback loop which may be linear or not, in order to set the power of the light beam generated by the module 230. In this regard, error correction mechanisms may be appropriately integrated within the feedback loop, in particular a proportional (P), proportional-integral (PI), or even Proportional Integral Differential (PID) correction, etc., used with appropriate parameters.

Whatever the type of light regulation envisaged, by discrete steps or by means of linear or non-linear control, the regulation of dynamic or reactive lighting can be advantageously improved by introducing an exploitation of the accelerometer data µx, µy and µz generated by the three-dimensional accelerometer sensor 110, as will now be described.

IV. Collaboration of the accelerometer 110 with the dynamic light regulation mechanism

The three-dimensional accelerometer module 110 provides accelerometer signals µx, µy and µz along three trigonometric axes X1, Y1 and Z1. As shown in the figure 8 , the X1 and Z1 axes are horizontal while the Y1 axis is a vertical axis and, moreover, the X1 and Y1 axes are arranged in a sagittal plane relative to the user.

There Figure 3a illustrates typical timing diagrams of the µx, µy and µz signals for a physical walking activity.

There figure 3b illustrates typical timing diagrams of the same µx, µy and µz signals for a physical cycling activity.

Finally, the figure 3c illustrates typical timing diagrams of µx, µy and µz signals for running physical activity.

There Figure 4a illustrates more particularly the profile of the µx acceleration, while the Figures 4b and 4c illustrate the profiles of the accelerations µy and µz, respectively.

As seen in these figures, the profiles of these accelerations µx, µy and µz are very characteristic and are clearly distinguished according to the three physical activities considered: Walking; cycling or biking; running or jogging.

In order to significantly improve the reactive or dynamic regulation mechanism, the headlamp control module 100 is configured to come to execute a method of detecting a physical activity profile, detected within a set of N predetermined profiles.

In this regard, the control module 220 is configured so that the non-volatile memory 223 comprises a memory area in which data representative of several physical activity profiles are stored, and preferably the data representative of the activities " walking ", " running " and " cycling " . Furthermore, the non-volatile memory 223 also comprises an area intended for storing a micro-program allowing the processing of the accelerometer data µx, µy and µz generated on the fly by the 3D accelerometer module 110. This algorithm will, as will be detailed later in relation to the figure 5 , compare the µx, µy and µz data generated in real time with data stored in memory 223 which are characteristic of the predetermined profiles (walking, cycling, running) stored in the memory. The algorithm aims to bring together, at regular intervals, the accelerometer data of a determined profile so as to bring the signals generated by the accelerometer into the predefined physical activity category, i.e. that corresponding to the different profiles stored in the memory of the headlamp.

There figure 5 illustrates a method of light regulation in accordance with the present invention, based jointly on the detection of ambient brightness and the exploitation of accelerometer data.

In a step 510, the method generates at regular intervals, for example every 20 milliseconds, a set of accelerometer data µx, µy and µz provided by the 3D accelerometer module 110. Optionally, the method may be limited to only part of the accelerometer data, for example only the µy data along the vertical direction Y1.

In a step 520, the method stores the data µx, µy and µz in the RAM 222.

Then, in a step 530, the accelerometer data µx, µy and µz are subject to digital processing to select a physical activity profile from a set of N predetermined profiles stored in non-volatile memory 223. Several methods can be used to select or detect the physical activity profile and will be explained in more detail in section V of this description.

In a step 540, the method uses the profile selected in step 530 as an input pointer to access a look-up table (LUT) in which values and parameters specific to the dynamic or reactive regulation mechanism applied by the control module 220 of the headlamp 100 are stored, and allowing the generation of the information or the control signal 225 transmitted to the power module 210.

In a particular embodiment, the parameters read from the LUT lookup table correspond to threshold values loaded into registers used by the reactive or dynamic regulation algorithm.

More specifically, the parameters are reduced to a threshold value corresponding to a minimum of lighting considered by the dynamic regulation algorithm.

Alternatively, in the case where the dynamic regulation algorithm uses a set of separate registers in which threshold values corresponding to various brightnesses are stored, reading the look-up table provides these threshold values. Thus, the minimum brightness value, but possibly also the maximum light power value, are defined based on the accelerometer data µx, µy and µz.

As will be understood, a person skilled in the art will be able to design various variants in the use of the values extracted from the correspondence table. It should be noted that these values may be used to set more general parameters than thresholds, and in particular variables used in linear or non-linear automatic regulation mechanisms, for example parameters or integral correction variables, or proportional-integral etc., in order to more finely adapt the reactive or dynamic regulation mechanism to the detected physical activity profile.

Then in a step 550, the method reads the LUT table and extracts the parameter(s) stored therein and, in the case of the preferred embodiment which is particularly economical to implement, the method extracts the minimum threshold value which should be applied to the reactive or dynamic light regulation mechanism.

In a step 560, the reactive or dynamic light regulation mechanism is executed using the value(s) extracted from the LUT table so as to precisely adapt this regulation, and where appropriate the feedback loop fixing the light power generated by the headlamp to adapt it to the physical activity identified in step 530. Thus the control information or the control signal transmitted via the conductor 225 is generated from the value(s) extracted from the LUT, together with the information provided by the light sensor 120.

In the preferred embodiment based on reading a single minimum threshold value within the LUT table, the dynamic or reactive regulation is therefore applied so as to ensure, in all cases, a minimum light power corresponding to the threshold value extracted from the LUT table.

It should be noted that various variants may be envisaged by a person skilled in the art, and in particular variants relating to the adjustment of the beam geometry. Indeed, the LUT table may conveniently include, in addition to the minimum threshold value mentioned above, one or more additional parameters making it possible to set the beam geometry, and in particular the use of wide or narrow collimation, or even a combination of the two. It could even be advantageous to extract from the LUT table the proportions of distribution of the light power on the three beams: wide, mixed and pointed depending on the physical activity detected.

Then, in a step 570, the method loops to step 510 to perform the reading and processing of new accelerometer data µx, µy and µz.

As can be seen, the reactive or dynamic light regulation mechanism is advantageously enriched by the contribution of the accelerometer data obtained on the fly from the accelerometer 110, and which the control module 220 processes to bring the processed data closer to a predetermined physical activity profile stored in the non-volatile memory 223 which, once identified, makes it possible to consult the LUT table so as to extract the most appropriate parameters and adjustment values for the light regulation.

In this way, it is possible to advantageously make the use of the ambient brightness captured by the sensor 120 cooperate with the raw accelerometer data µx, µy and µz generated directly by the 3D accelerometer sensor 110.

There figure 6 illustrates the effect of the process just described where we see that the low level threshold set without the input of accelerometer data remains at the same level regardless of the activity considered, for example walking (left part of the figure), running (middle part of the figure) and cycling or mountain biking (right part of the figure). If this low level does not pose any difficulty for a walking type activity, we observe on the other hand that this same low level presents a discomfort zone for a running activity and even becomes a danger zone for a mountain biking type activity.

As just described, the process described in the figure 5 allows the low level threshold to be raised, to adapt it to a first higher level for a running activity and to raise it to a second even higher level for a mountain bike type activity, so that the user is never in the discomfort zone represented in the middle part of the figure 6 and even less in the danger zone of the right part of this same figure.

We can therefore see in the end that the process allows a finer adaptation of the light power determined according to a reactive or dynamic regulation process, which takes into account the physical activity profile considered.

It should be noted that a set of three activity profiles has been described but that the invention could conveniently be used for a greater number of profiles (climbing, alpine skiing, Nordic skiing, etc.)

V. Physical activity detection method

• .- un premier accéléromètre élémentaire, configuré pour mesurer l'évolution d'une première composante d'accélération µx, longitudinale, de la lampe suivant un premier axe (X1) sensiblement parallèle à la direction du mouvement de la lampe,
• .- un deuxième accéléromètre élémentaire, configuré pour mesurer l'évolution d'une deuxième composante d'accélération µy, verticale, de la lampe suivant un deuxième axe (Y1) sensiblement parallèle à la direction verticale terrestre locale,
• .- un troisième accéléromètre élémentaire, configuré pour mesurer l'évolution d'une troisième composante d'accélération µz, latérale, suivant un troisième axe (Z1) perpendiculaire aux premier et deuxième axes.

Physical activity detection is based on a 3D 110 three-dimensional acceleration sensor which comprises three elementary accelerometers:

• .- a first elementary accelerometer, configured to measure the evolution of a first longitudinal acceleration component µx of the lamp along a first axis (X1) substantially parallel to the direction of movement of the lamp,
• .- a second elementary accelerometer, configured to measure the evolution of a second acceleration component µy, vertical, of the lamp along a second axis (Y1) substantially parallel to the local terrestrial vertical direction,
• .- a third elementary accelerometer, configured to measure the evolution of a third lateral acceleration component µz, along a third axis (Z1) perpendicular to the first and second axes.

The X1 and Y1 axes are placed in a sagittal plane relative to the user.

Each elementary accelerometer is configured to provide a time series of elementary acceleration values along their corresponding axis. The first time series, provided by the first elementary accelerometer, forms a first elementary raw signal, designated by S 1 _b ( t, µ ), which varies as a function of time t and of the motion profile µ of the 3D acceleration sensor relative to the local terrestrial reference frame. The second time series, provided by the second elementary accelerometer, forms a second elementary raw signal, designated by S 2 _b ( t, µ ), which varies as a function of time t and of the motion profile µ of the 3D acceleration sensor relative to the local terrestrial reference frame. The third time series, provided by the third elementary accelerometer, forms a third elementary raw signal, designated by S 3 _b ( t , µ ), which varies as a function of time t and of the motion profile µ of the 3D acceleration sensor relative to the local terrestrial reference frame. The motion profile µ of the 3D acceleration sensor is for example that of a walker, designated by µ 1 , that of a cyclist, designated by µ 2, or that of a runner, designated by µ 3. As illustrated in particular in the Figures 8a to 8d .

The control module 220 comprises a digital electronic circuit - which may advantageously be produced using the processor 221 associated with its memory or using any other specialized digital signal processor (DSP ), and which is configured to process only one or at least two of the raw signals S 1 _b ( t , µ ), S 2 _b ( t , µ ), S 3 _b ( t, µ ) provided by the 3D acceleration sensor according to a method 700 or algorithm for processing the signal and determining the movement profile of the 3D acceleration sensor illustrated in the figure. figure 7 , and finally allowing the detection of physical activity useful to the process of the figure 5 .

The 700 process of the figure 7 comprises an optional initial filtering step 710, followed by a feature extraction step 720, then a decision step 730 by thresholding.

In the initial step 710 of the processing method 700, referred to as the " filtering step ", one or more of the raw signals S 1 _b ( t , µ ), S 2 _b ( t , µ ), S 3 _b ( t , µ ) are filtered respectively into new signals, called useful signals and referred to as S 1 _u ( t , µ ), S 2 _u ( t , µ ), S 3 _u ( t , µ ), in which the useful information is still present but where the harmful information, called "noise" (here electronic noise of the 3D acceleration sensor), is either removed or weakened. The overall information contained in the signal therefore has a certain degree of specialization at this level. In the case where the initial filtering step 710 is omitted, the raw signals S 1 _b ( t , µ ) , S 2 _b ( t , µ ) , S 3 _b ( t , µ ) are respectively identical to the useful signals S 1 _u ( t , µ ), S 2 _u ( t , µ ), S 3 _u ( t , µ )

Following the Figures 8A, 8B, 8C and 8D the useful signals of the triplet ( S 1 _u ( t, µ ), S 2 _u ( t , µ ), S 3 _u ( t , µ )) are illustrated respectively for different motion profiles µ 0, µ 1, µ 2 and µ 3 of the 3D acceleration sensor 110.

Following the Figure 8A , the useful signals S 1 _u ( t , µ 0 ), S 2 _u ( t , µ 0 ) and S 3 _u ( t , µ 0 ), illustrated respectively on a first curve 802, a second curve 804, a third curve 806, are typically those of a 3D acceleration sensor having the shape 808 of a reference movement profile µ 0 , corresponding to a low amplitude or quasi-zero movement of the 3D acceleration sensor.

Following the Figure 8B , the useful signals S 1 _u ( t, µ 1), S 2 _u ( t, µ 1) and S 3 _u ( t , µ 1) , illustrated respectively on a fourth curve 822, a fifth curve 824 and a sixth curve 826 are typically those of a 3D acceleration sensor having the shape 828 of a movement profile µ 1 of a walker (in English “walking”).

Following the Figure 8C , the useful signals S 1 _u ( t , µ 2 ), S 2 _u ( t , µ 2 ) and S 3 _u ( t , µ 2 ), illustrated respectively on a seventh curve 842, an eighth curve 844 and a ninth curve 846 are typically those of a 3D acceleration sensor having the shape 848 of a movement profile µ 2 of a cyclist (in English “biking”).

Following the Figure 8D , the useful signals S 1 _u ( t, µ 3), S 2 _u ( t, µ 3) and S 3 _u ( t, µ 3), illustrated respectively on a tenth curve 862, an eleventh curve 864 and a twelfth curve 866 are typically those of a 3D acceleration sensor having the shape 868 of a movement profile µ 3 of a runner (in English “jogging”).

The purpose of the feature extraction step 720 is to extract from at least one of the useful signals S 1 _u ( t , µ ), S 2 _u ( t , µ ), S 3 _u ( t , µ ) a finite set of one or more parameters, if possible independent, representative of the observed phenomenon, and making it possible to describe it.

The extraction of features implemented in step 720 allows in other words the passage of a useful vector or scalar signal to data.

The difference between these two types is important: a signal can be seen as a set of points for which each point has a high rate of dependence (deterministic or statistical) with its neighbors. Data represents a set of points where this notion of neighborhood is less important. In reality, the passage from the signal to the data is most often done in several stages. The intermediate entities are then indifferently called signal, estimator, or data. The main goal of feature extraction is to arrive, from the useful signal, at data that are independent of each other and exhaustively represent the phenomenon to be interpreted.

In general, the useful signals studied here can be characterized by elementary estimators which are the moments of these signals: the mean (moment of order 1), and the pseudo-standard deviation (moment of order 2) are the best known and most used. For example, an estimator can be a function of one or more moments of the same useful signal.

dans laquelle :

• .-Nech désigne le nombre total d'instants d'échantillonnage équiréparties dans la fenêtre d'échantillonnage courante,
• .- mS2 désigne la moyenne statistique du signal utile S2 calculée dans la fenêtre d'échantillonnage courante calculée à partir des mesures de signal utile S2 aux mêmes instants d'échantillonnage tk.

According to a first embodiment, the useful signal S 2 _u ( t, µ ) which measures the evolution of the second vertical acceleration component of the lamp can characterize by itself the motion profile of the lamp from its second-order moment, that is to say its variance. According to the first embodiment, the estimator allowing the motion profile of the lamp to be characterized is written on a current and sliding sampling window of predetermined duration T _{and is} by the following equation: $$\mathit{<figure-callout\; id="2"\; label="East\; S"\; filenames="imgf0004.png"\; state="\{\{state\}\}">East\; </figure-callout>}\left(S\right)\left(2\right)\left(\mu \right)={\displaystyle \sum _{k=1\mathrm{has}\mathit{<figure-callout\; id="2"\; label="Nech\; S"\; filenames="imgf0004.png"\; state="\{\{state\}\}">Nech</figure-callout>}}{\left(S\right)}^{}{\left(2\left(\mathit{tk},\mu \right)-\mathit{<figure-callout\; id="2"\; label="mS"\; filenames="imgf0004.png"\; state="\{\{state\}\}">mS</figure-callout>}2\right)}^2/\mathit{Nech}}$$

in which:

• .-Nech denotes the total number of equally distributed sampling instants in the current sampling window,
• .- mS2 denotes the statistical mean of the useful signal S2 calculated in the current sampling window calculated from the useful signal S2 measurements at the same sampling times tk.

Here the elementary estimator considered Est(S2) is the statistical variance of the useful signal S 2 _u ( t , µ ).

Then, in the thresholding decision step 730, the type of lamp displacement profile is determined by thresholding on the estimator Est ( S 2 )( µ ).

These elementary estimators taken in isolation may not always be sufficient to provide a good description of a complex problem. In order to systematically choose estimators that are consistent and useful for the interpretation of a signal, more sophisticated analysis methods may prove useful.

For complex problems, efficient feature extraction is very often reduced by statisticians to determining the dimension of the problem. This dimension is given by the minimum number of parameters allowing the problem to be represented exhaustively. These parameters are then called variables of the problem. By definition, these variables are variables that are independent of each other, this reduces the dimension of the problem by 1. In practice, for complex problems, it is very difficult to construct the vector of variables. Indeed, it is rare that the estimators that we know how to extract from a signal are totally independent of each other. In addition, the construction of these estimators requires a "perfect" mathematical model of the problem (in the sense of physics), which is not always possible. A certain number of analysis methods make it possible to extract, to construct a vector of parameters from any vector. These methods are grouped under the generic term of factor analysis.

• ;- l'analyse en composantes principales,
• .- l'analyse factorielle des correspondances,
• .- l'analyse factorielle des correspondantes multiples
• .- l'analyse factorielle discriminante,
• .- la régression linéaire,
• .- la classification par k-moyennes (en anglais k-means),
• .- la caractérisation par géométrie fractale.

Factor analysis is based on geometric reasoning on the data. The signal is considered as a "point cloud" in an N-dimensional space, and we seek to determine the geometric characteristics of this cloud: principal axes (eigenvectors), spread, form factors, etc. To do this, the approach is to calculate the eigenvectors of the point cloud, then to change space, so as to express the coordinates of the points of the cloud, as well as all the known relationships on these points, in the space of eigenvectors. Among the statistical methods of factor analysis are:

• ;- principal component analysis,
• .- factorial analysis of correspondences,
• .- factor analysis of multiple correspondents
• .- discriminant factor analysis,
• .- linear regression,
• .- k-means classification,
• .- characterization by fractal geometry.

dans laquelle les paramètres a, b, c sont déterminés par apprentissage sur le signaux utiles d'apprentissages {S1 _u (t, µ0), S2 _u (t, µ0), S3 _u (t, µ0)}, {S1 _u (t, µ1), S2 _u (t, µ1), S3 _u (t, µ1)} , { S1 _u (t, µ2),S2 _u (t, µ2),S3 _u (t, µ2)} , et { S1 _u (t, µ3), S2 _u (t, µ3) et S3 _u (t, µ3).For example, according to a second embodiment, the dimension of the problem of estimating the lamp displacement profile is considered equal to 3. The three elementary variables are formed by the respective statistical variances Est(S1)(µ), Est(S2)(µ), Est(S3)(µ), of the useful signals S 1 _u ( t, µ ), S 2 _u ( t , µ ), S 3 _u ( t, µ ). A scalar estimator denoted Est(S1,S2,S3)(µ) of the vector useful signal ( S 1 _u ( t , µ ), S 2 _u ( t, µ ), S 3 _u ( t, µ )) is determined as a linear combination of the statistical variances Est(S1)(µ), Est(S2)(µ), Est(S3)(µ) according to the equation: $$\mathit{<figure-callout\; id="1"\; label="East\; S"\; filenames="imgf0004.png,imgf0007.png"\; state="\{\{state\}\}">East\; </figure-callout>}\left(S\right)\left(1,\mathrm{<figure-callout\; id="2"\; label="S"\; filenames="imgf0004.png"\; state="\{\{state\}\}">S</figure-callout>}2,\mathrm{<figure-callout\; id="3"\; label="S"\; filenames="imgf0004.png"\; state="\{\{state\}\}">S</figure-callout>}3\right)\left(\mu \right)=\mathrm{has}\ast \mathit{<figure-callout\; id="1"\; label="East\; S"\; filenames="imgf0004.png,imgf0007.png"\; state="\{\{state\}\}">East\; </figure-callout>}\left(S\right)\left(1\right)\left(\mu \right)+b\ast \mathit{<figure-callout\; id="2"\; label="East\; S"\; filenames="imgf0004.png"\; state="\{\{state\}\}">East\; </figure-callout>}\left(S\right)\left(2\right)\left(\mu \right)+c\ast \mathit{<figure-callout\; id="3"\; label="East\; S"\; filenames="imgf0004.png"\; state="\{\{state\}\}">East\; </figure-callout>}\left(S\right)\left(3\right)\left(\mu \right)$$

in which the parameters a, b, c are determined by learning on the useful learning signals { S 1 _u ( t, µ 0), S 2 _u ( t, µ 0), S 3 _u ( t, µ 0)}, { S 1 _u ( t, µ 1), S 2 _u ( t , µ 1), S 3 _u ( t , µ 1)}, { S 1 _u ( t, µ 2 ), S 2 _u ( t , µ 2), S 3 _u ( t, µ 2)}, and { S 1 _u ( t , µ 3), S 2 _u ( t , µ 3) and S 3 _u ( t , µ 3).

Then, in the thresholding decision step 308, the type of lamp displacement profile is determined by thresholding on the scalar estimator Est ( S 1, S 2, S 3)( µ ).

It should be noted that these more complex achievements, using the combination of several variables, make the detection process more robust, particularly with regard to a possible rotation of the user's head relative to one of the axes.

VI. Further improvements and advantages of the invention

In a preferred embodiment, the physical activity profile identified by the control module 220 is transmitted by the wireless link to the mobile telephone 300 so that the latter can inform, at any time, of the physical activity detected automatically according to the technique previously in order, if necessary, to allow the user to correct the detection and allow adaptive learning of the physical activity detection method.

Furthermore, in a particular embodiment, the headlamp is configured to read accelerometer data µx, µy and µz on the fly to determine the user's fall and, in this case, to trigger an emergency procedure. In particular, the procedure may be based on sending an alert signal to the mobile phone so as to initiate the generation of an emergency message, such as an SMS or email.

Alternatively, or cumulatively, the alert procedure will include activating the lamp to generate an alert light sequence, such as MORSE coding of the well-known S.O.S. sequence.

Any other alert procedure may be considered once the headlamp control module 220 has detected the user's fall.

Finally, it is useful to note that the invention is not limited to headlamps only and can be used applied to a hand lamp.

Claims (12)

1. A headlamp (100) comprising

   - a light source (231) comprising one or more LED-type diodes;

   - a power module (230) for supplying current to said light source (231), said power module being controlled by a control information or a control signal;

   - a control module (220) for adjusting the light intensity generated by said light source;

   - a light sensor (120) for sensing light from the environment of the lamp holder, said control module (220) being configured to generate said control information or said control signal according to the information generated by said light sensor (120),

   - an accelerometer (110) configured to provide at regular intervals data representative of an acceleration of the headlamp along at least one horizontal axis and one vertical axis;

   wherein said control module (220) includes circuitry (221, 222, 223) configured to digitally store and process data representative of said acceleration;

   characterized in that

   - said control module (220) is configured for selecting a physical activity profile selected from a set of predetermined physical activities stored within a memory (223) from data representatives of said acceleration being digitally processed; and

   wherein said control module (220) comprises a LUT look-up table stored within said memory (223) providing at least one value or adjustment parameter serving for generating said control information or said control signal;

   wherein the physical activity profile serves as an entry pointer into said LUT;

   wherein the value or adjustment parameter read from said LUT is used in conjunction with information generated by the light sensor to determine said control information or said control signal.
2. The headlamp according to claim 1 wherein said accelerometer (110) is configured for generating accelerometer data along two horizontal axes X1, Z1 and along a vertical axis Y1; and
   wherein the set of predetermined accelerometry profiles include profiles representative of walking, running and cycling.
3. The headlamp according to claim 1 wherein the power of the light beam adjusted by the control unit (220) varies between two thresholds, respectively low and high, and wherein said low threshold is set by a value extracted directly from said LUT table .
4. The Headlamp according to claim 1 wherein said circuit (221, 222, 223) configured to store and process data representative of said acceleration uses a digital and statistical processing method based on the measurement of the variance of the component of vertical acceleration µy1 of said headlamp.
5. The headlamp according to claim 4 wherein said circuit (221, 222, 223) configured to store and process data representative of said acceleration uses a digital and statistical processing method based on the measurement of the variance of two components of acceleration of said headlamp.
6. The headlamp according to claim 1 wherein the data extracted from the LUT table make it possible to define a minimum light power threshold and a specific geometry of the light beam chosen between a wide beam, a narrow focusing beam and/or both at the same time.
7. The headlamp according to claim 1 wherein said control module (220) is configured to process data from the accelerometer to detect the fall of a user.
8. The headlamp according to claim 7 wherein said control module transmits the user fall information in order to generate an electronic alert transmitted to a mobile phone.
9. The headlamp according to claim 7 wherein the control module is configured to control a light alert sequence aimed at calling for help.
10. A light regulation process for a headlamp as defined in one of the preceding claims, comprising the steps:

    - generating (510) at regular intervals a set of accelerometer data µx1, µy1 and µz1 according respectively a first predetermined horizontal axis X1, a predetermined vertical axis Y1 and a second predetermined horizontal axis Z1 provided by said accelerometer;

    - storing (520) said data µx1, µy1 and µz1 within a random access memory (222);

    - performing digital processing (530) on said accelerometer data µx1, µy1 and µz1 in order to determine a physical activity profile selected from a set of N predetermined profiles stored in a non-volatile memory (223);

    - using (540) the selected profile as an input pointer to access a LUT correspondence table in which are stored values and parameters specific to the mechanism allowing the generation of said control information or said control signal used to adjust the light power;

    - reading the LUT table (550) and extracting the parameter(s) or values stored therein;

    - determining said control information or said control signal from the value or values extracted from the LUT, together with the information provided by said light sensor (120);

    - return to the first step to read and process new accelerometer data.
11. The process according to claim 10 wherein the reading of the LUT correspondence table provides a value defining the minimum power generated by the lamp.
12. A headlamp according to claim 1 wherein the data extracted from the LUT table allows the definition of a minimum threshold of the light power and the setting of a specific geometry of the light beam by the selection via said control module (220) of an operating mode selected among a first mode of wide beam, a second mode of narrow beam and a third mode of simultaneous wide and narrow beams.

Images