CN118213207B - Intelligent control method and equipment for capacitor element nailing machine - Google Patents
Intelligent control method and equipment for capacitor element nailing machine Download PDFInfo
- Publication number
- CN118213207B CN118213207B CN202410596771.5A CN202410596771A CN118213207B CN 118213207 B CN118213207 B CN 118213207B CN 202410596771 A CN202410596771 A CN 202410596771A CN 118213207 B CN118213207 B CN 118213207B
- Authority
- CN
- China
- Prior art keywords
- capacitor
- state
- value
- stage
- equipment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 239000003990 capacitor Substances 0.000 title claims abstract description 99
- 238000000034 method Methods 0.000 title claims abstract description 31
- 238000004519 manufacturing process Methods 0.000 claims abstract description 49
- 238000001514 detection method Methods 0.000 claims abstract description 48
- 238000004804 winding Methods 0.000 claims abstract description 18
- 230000002159 abnormal effect Effects 0.000 claims abstract description 7
- 239000010410 layer Substances 0.000 claims description 67
- 241000894007 species Species 0.000 claims description 54
- 239000013598 vector Substances 0.000 claims description 26
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000013528 artificial neural network Methods 0.000 claims description 5
- 238000009826 distribution Methods 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 4
- 238000003860 storage Methods 0.000 claims description 4
- 230000002457 bidirectional effect Effects 0.000 claims description 3
- 230000002431 foraging effect Effects 0.000 claims description 3
- 244000062645 predators Species 0.000 claims description 3
- 239000002356 single layer Substances 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- 239000011888 foil Substances 0.000 abstract description 5
- 238000012797 qualification Methods 0.000 abstract description 3
- 238000005457 optimization Methods 0.000 description 3
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 2
- 229910052782 aluminium Inorganic materials 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005096 rolling process Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H01—ELECTRIC ELEMENTS
- H01G—CAPACITORS; CAPACITORS, RECTIFIERS, DETECTORS, SWITCHING DEVICES, LIGHT-SENSITIVE OR TEMPERATURE-SENSITIVE DEVICES OF THE ELECTROLYTIC TYPE
- H01G13/00—Apparatus specially adapted for manufacturing capacitors; Processes specially adapted for manufacturing capacitors not provided for in groups H01G4/00 - H01G11/00
- H01G13/02—Machines for winding capacitors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/082—Learning methods modifying the architecture, e.g. adding, deleting or silencing nodes or connections
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Power Engineering (AREA)
- Computing Systems (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Molecular Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Health & Medical Sciences (AREA)
- Manufacturing & Machinery (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Feedback Control In General (AREA)
- Fixed Capacitors And Capacitor Manufacturing Machines (AREA)
Abstract
The invention discloses an intelligent control method and equipment for a capacitor element nailing machine, wherein the method comprises the steps of detecting the internal state of production equipment, detecting the quality of capacitor products, calculating the working state prediction of the equipment according to the quality of the capacitor products and the detection result of the internal state of the production equipment, obtaining a working state prediction value, and when the working state prediction value is larger than a threshold value, indicating that the working state is abnormal, and entering the step of adjusting equipment parameters; and adjusting the movement speed and the tension of the feeding mechanism, the positive electrode nailing machine, the negative electrode nailing machine and the winding machine based on an improved swarm intelligence algorithm. The invention realizes the automatic control of the capacitor element nailing machine, effectively improves the production quality of the capacitor, reduces the probability of generating foil leakage and deflection nails, and improves the qualification rate and the service life of the capacitor.
Description
Technical Field
The invention relates to the field of automatic production control of capacitor elements, in particular to an intelligent control method and equipment of a capacitor element nailing and rolling machine.
Background
The nailing machine is an important device for producing capacitor elements and is used for completing connection of leads and winding of electrolytic paper; the core of the electrolytic capacitor is formed by winding positive aluminum foil, electrolytic paper and negative aluminum foil. The capacitor coil needs to be performed under a certain tension, and if the coil tension is uneven or the coil speed is too high, the foil leakage phenomenon is easy to occur, and the qualification rate and the service life of the electrolytic capacitor are greatly influenced.
Disclosure of Invention
In order to solve the technical problems, the invention provides an intelligent control method and equipment for a capacitor element nailing machine.
In order to solve the technical problems and achieve the aim of the invention, the invention is realized by the following technical scheme:
an intelligent control method of a capacitor element nailing machine comprises the following steps:
S1: detecting the internal state of production equipment, including detecting the movement speed and the tension of a feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine and a winding machine;
S2: detecting the quality of a capacitor product; the quality of the capacitor product comprises the appearance and the electrical performance of the capacitor, wherein the appearance comprises the detection of the position error of a capacitor pin, the diameter error and the height error of the capacitor; the electrical property detection comprises detection of capacitance value errors and leakage current errors;
S3: calculating equipment working state prediction according to the quality of the capacitor product and the internal state detection result of the production equipment to obtain a working state prediction value, and when the working state prediction value is smaller than a threshold value, indicating that the working state is normal; when the predicted value of the working state is larger than the threshold value, the working state is abnormal, and the parameter equipment is adjusted;
S4: and adjusting equipment parameters, and adjusting the movement speeds and the tension of the feeding mechanism, the positive electrode nailing machine, the negative electrode nailing machine and the winding machine based on an improved group intelligent algorithm.
Further, the step S3 includes:
s31: determining a capacitor product quality predictor and a production facility internal state predictor based on the improved LSTM network;
s32: calculating the weight proportion of the capacitor product quality and the internal state detection result of production equipment;
s33: calculating an operating state predicted value Y based on:
，
Q 1 is the weight occupied by the quality of the capacitor product, q 2 is the weight occupied by the detection result of the internal state of the production equipment, p 1 is the predicted value of the quality of the capacitor product, and p 2 is the predicted value of the internal state of the production equipment;
S34: judging whether to enter a step of adjusting equipment parameters according to the working state predicted value; specifically, when the predicted value of the working state is smaller than the threshold value, the working state is normal; and when the predicted value of the working state is larger than the threshold value, indicating that the working state is abnormal, and entering the step of adjusting the parameters of the equipment.
Further, the step S31 includes: the improved LSTM network structure comprises an input layer, a hidden layer and an output layer, wherein the hidden layer comprises two hidden layers in two directions, namely a front hidden layer and a rear hidden layer, and input layer data can be transmitted to the front hidden layer or can be input to the rear hidden layer.
Further, the step S31 further includes:
S311: establishing an improved LSTM network structure, initializing a network parameter w f、wi、wc and a bias value b f、bi、bc;
Wherein w f、wi、wc is the weight of the forgetting gate, the output gate and the cell state respectively, and b f、bi、bc is the bias of the forgetting gate, the output gate and the cell state respectively;
S312: respectively importing the existing data comprising respective detection parameters, corresponding capacitor product quality predicted values and production equipment internal state predicted values, and dividing the data into a training set and a testing set;
S313: inputting the input vector into an improved LSTM network, calculating the current hidden layer state value of the single-layer hidden layer network, and updating the forgetting gate, the input gate, the output gate and the hidden layer output result of the network;
S314: combining bidirectional loops of forward and backward stacked unidirectional LSTM neural networks to obtain improved LSTM network model output ；
The state of the previous hidden layer is indicated,Representing the state of the post hidden layer;
S315: calculating a weight corresponding to each output vector;
s316: summarizing the input sequence data by using a weighted average formula to obtain a new output value;
S317: finally, the output value is transmitted to an output layer to obtain a prediction result, namely a final output result;
And (3) repeatedly calculating the steps S311-S317 to obtain a capacitor product quality predicted value p 1 and a production equipment internal state predicted value p 2.
Further, the step S315 further includes calculating a weight corresponding to each output vector based on the following formula:
，
Wherein v T is an output vector, h t is the output of the cell at time t, W is an output layer weight, and U is an input layer weight; n is the total number of output vectors, h j is the j-th output vector, softmax () is the normalization function;
the step S316 calculates V based on the following formula:
；
And N is the total number of output vectors, r m is the weight corresponding to the mth output vector, and h m is the mth output vector.
Further, the step S32 includes:
And determining the quality of the capacitor product and the weight distribution of the internal state detection result of the production equipment based on an entropy weight method.
Further, the step S4 includes: a first stage, a foraging stage and a second stage, a defensive strategy stage for predators.
Further, the first stage includes:
s431: generating a new location of the ith species according to the following formula:
；
Wherein, For the state value of the ith species in the jth dimension in the first stage, x i,j is the state value of the previous state of the ith species in the jth dimension, r is a random number of [0,1], PZ j is the position of the optimal species in the jth dimension, I is a variation range parameter, I=round (1+rand), rand is a random number of [0,1], and the larger I is, the larger the variation range is; p1 represents the first stage;
S432: updating the position of the ith species according to the calculation result of the fitness function, specifically calculating the fitness of the updated species group, comparing with the fitness of the current position, if the fitness after updating is better than the fitness at the current position, defining the position of the individual as the updated position, specifically calculating as follows:
，
Wherein F i is the fitness function value corresponding to the original position, For the fitness function value corresponding to the new position of the first stage, X i is the position of the previous state of the ith species,New positions for the ith species of the first stage; p1 represents the first stage;
The second stage comprises:
S441: generating random numbers ；
S442: if the random number P s is less than or equal to 0.5, the new state of the ith species is calculated using the formula:
，
Wherein, For the state value of the ith species in the jth dimension in the second stage, R is a constant, R is a random number of [0,1], T is the current iteration number, and T is the maximum iteration number; p2 represents the second stage;
S443: if the random number P s > 0.5, the new state of the ith species is calculated using the following formula:
，
Wherein, For the state value of the ith species in the jth dimension in the second stage, r is a random number of [0,1], AZ j is the position of the attacked species in the jth dimension, I=round (1+rand), rand is a random number of [0,1], and the larger the I is, the larger the variation range is;
S444: updating the position of the ith species according to the calculation result of the fitness function, specifically calculating the fitness of the updated species group, comparing with the fitness of the current position, if the fitness after updating is better than the fitness at the current position, defining the position of the individual as the updated position, specifically calculating as follows:
，
Wherein F i is the fitness function value corresponding to the original position, For the fitness function value corresponding to the new position of the second stage, X i is the position of the previous state of the ith species,New positions for the second stage i species; p2 represents the second stage.
An intelligent control device for a capacitor element nailing machine, comprising:
The production equipment internal state detection module is used for detecting the internal state of the production equipment and comprises a detection feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine and a winding machine;
A capacitor product quality detection module for detecting a capacitor product quality, the capacitor product quality including a capacitor appearance and an electrical property;
The working state prediction module is used for calculating equipment working state prediction according to the quality of capacitor products and the detection result of the internal state of production equipment and adjusting equipment parameters according to the equipment working state prediction value;
And the equipment parameter adjustment module is used for adjusting the speeds and the tensions of a feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine and a winding machine of the equipment based on the improved swarm intelligence algorithm.
In addition, to achieve the above object, the present invention also provides a computer-readable storage medium having stored thereon program instructions of a capacitor sub-reel machine intelligent control method, which can be executed by one or more processors to implement the steps of the capacitor sub-reel machine intelligent control method as described above.
Compared with the prior art, the invention has the beneficial effects that:
(1) The invention realizes the automatic control of the capacitor element nailing machine, effectively improves the production quality of the capacitor, reduces the probability of generating foil leakage and deflection nails, and improves the qualification rate and the service life of the capacitor.
(2) The invention is based on an improved LSTM network structure, through arranging the hidden layers in two directions, not only can realize the data input from the past to the future, but also increases the data input flow direction from the past, and the hidden layers of the past neural network and the hidden layers of the future are mutually independent, so that the advantage of better mining time sequence data characteristics can be realized.
(3) According to the method, the quality of the capacitor product and the weight distribution of the detection result of the internal state of the production equipment are determined based on the entropy weight method, and the subjective deviation caused by manual determination is avoided and the prediction accuracy is improved through objective weight distribution.
(4) The invention carries out speed and tension adjustment calculation based on an improved swarm intelligent algorithm, and solves the optimization problem by establishing proper balance between exploration and development.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention and do not constitute a limitation on the invention. In the drawings:
FIG. 1 is a flow chart of a method for intelligent control of a capacitor element nailing machine according to an embodiment of the present invention;
Fig. 2 is a schematic diagram of an improved LSTM network structure according to an embodiment of the present invention.
Detailed Description
Embodiments of the present disclosure are described in detail below with reference to the accompanying drawings.
Other advantages and effects of the present disclosure will become readily apparent to those skilled in the art from the following disclosure, which describes embodiments of the present disclosure by way of specific examples. It will be apparent that the described embodiments are merely some, but not all embodiments of the present disclosure. The disclosure may be embodied or practiced in other different specific embodiments, and details within the subject specification may be modified or changed from various points of view and applications without departing from the spirit of the disclosure. It should be noted that the following embodiments and features in the embodiments may be combined with each other without conflict. All other embodiments, which can be made by one of ordinary skill in the art without inventive effort, based on the embodiments in this disclosure are intended to be within the scope of this disclosure.
It should also be noted that the illustrations provided in the following embodiments merely illustrate the basic concepts of the disclosure by way of illustration, and only the components related to the disclosure are shown in the drawings and are not drawn according to the number, shape and size of the components in actual implementation, and the form, number and proportion of the components in actual implementation may be arbitrarily changed, and the layout of the components may be more complicated.
Referring to fig. 1, an intelligent control method of a capacitor element nailing machine comprises the following steps:
S1: detecting the internal state of the production equipment, including detecting the movement speed and the tension of a feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine and a winding machine.
S2: detecting the quality of a capacitor product;
Capacitor product quality includes capacitor appearance and electrical properties; specifically, the appearance detection comprises measuring capacitor pin position error c 1, capacitor diameter and height error c 2、c3; the electrical property detection includes detecting a capacitance value error d 1 and a leakage current error d 2.
Optionally, the appearance detection is based on machine vision;
s3: calculating equipment working state prediction according to the quality of the capacitor product and the internal state detection result of the production equipment to obtain a working state prediction value, and when the working state prediction value is smaller than a threshold value, indicating that the working state is normal; and when the predicted value of the working state is larger than the threshold value, indicating that the working state is abnormal, and entering the step of adjusting the parameters of the equipment.
The internal states of the production equipment comprise the movement speed and the tension of a feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine and a winding machine.
The method specifically comprises the following steps:
s31: determining capacitor product quality predictions and production facility internal state predictions
The method for determining the capacitor product quality predicted value and the internal state predicted value of the production equipment based on the improved LSTM algorithm specifically comprises the following steps:
S311: establishing an improved LSTM network structure, initializing a network parameter w f、wi、wc and a bias value b f、bi、bc;
Wherein w f、wi、wc is the weight of the forgetting gate, the output gate and the cell state respectively, and b f、bi、bc is the bias of the forgetting gate, the output gate and the cell state respectively.
The improved LSTM network structure is shown in figure 2, and the network comprises an input layer, a hidden layer and an output layer; the hidden layers comprise a hidden layer A and a hidden layer B in two directions, namely a front hidden layer and a rear hidden layer; the input layer data can be transmitted to a front hidden layer or can be input to a rear hidden layer, the structure can pay attention to the information of the upper layer sequence and the lower layer sequence at the same time, the front hidden layer is a forward flow LSTM network structure from head to tail, and the rear hidden layer is a reverse flow LSTM structure from tail to head; the output data of the forward hidden layer and the backward hidden layer are combined to form the input of the output layer node, and finally the sequence data is output.
The network model not only can realize data input from the past to the future, but also increases the flow direction of the data input from the past to the future, and the hidden layers of the past neural network and the hidden layers of the future are mutually independent, so that the advantage of better mining time sequence data characteristics can be realized.
S312: the existing data, including the respective detection parameters and the corresponding capacitor product quality predicted values and the production equipment internal state predicted values, are respectively imported, and the data are divided into a training set and a testing set.
S313: the input vector is input into the improved LSTM network, the value of the current hidden layer state of the single layer hidden layer network is calculated, and the forget gate, input gate, output gate and hidden layer output result of the network are updated.
S314: combining bidirectional loops of forward and backward stacked unidirectional LSTM neural networks to obtain improved LSTM network model output；
The state of the previous hidden layer is indicated,Representing the state of the post hidden layer.
S315: calculating the weight corresponding to each output vector:
，
Wherein v T is an output vector, h t is the output of the cell at time t, W is an output layer weight, and U is an input layer weight; n is the total number of output vectors, h j is the j-th output vector, softmax () is the normalization function;
s316: summarizing the input sequence data by using a weighted average formula to obtain a new output value V as follows:
；
And N is the total number of output vectors, r m is the weight corresponding to the mth output vector, and h m is the mth output vector.
S317: finally, the output value is transmitted to an output layer to obtain a prediction result, namely a final output result;
The capacitor product quality predicted value p 1 and the production equipment internal state predicted value p 2 are obtained by repeatedly calculating the steps.
S32: weight ratio calculation of capacitor product quality and internal state detection result of production equipment
Because the influence of each detection parameter on the prediction result is inconsistent, the method determines the quality of the capacitor product and the weight distribution of the detection result of the internal state of the production equipment based on the entropy weight method, and specifically comprises the following steps:
s321: data normalization
13 Detection parameters are respectively measured capacitor pin position error c 1, capacitor diameter and height error c 2、c3; the electrical property detection comprises detection capacitance value errors d 1, leakage current errors d 2, a feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine, a movement speed v 1、v2、v3、v4 of a winding machine and tension f 1、f2、f3、f4, and m groups of prior data are selected, wherein each group of prior data comprises 13 detection data.
All detection parameters are subjected to standardization treatment:
Let the original 13 detection parameters be X j={x1,x2,...,xm }, j=1 to 13, and the normalized value is y ij:
S322: solving the information entropy of each detection parameter:
Where p ij is the evaluation of data set i with respect to index j:
S323: calculating information deviation degree d j:
dj=1-Ej
S324: calculating the entropy weight w j of the j-th index:
s325: calculating the weight of capacitor product quality and the weight of internal state detection result of production equipment
W 1~w5 is the weight corresponding to the measured capacitor pin position error, capacitor diameter, height error, detected capacitance error and leakage current error.
W 6~w13 is the weight corresponding to the motion speed and the tension of the feeding mechanism, the positive electrode nailing machine, the negative electrode nailing machine and the winding machine respectively.
The weight occupied by the quality of the capacitor product and the detection result of the internal state of the production equipment is determined according to the importance degree of each index through an entropy weight method, so that human factors are avoided, and the weight is determined from the index, so that the method has objectivity.
S33: calculating a predicted value Y of the working state
Q 1 is the weight occupied by the quality of the capacitor product, q 2 is the weight occupied by the detection result of the internal state of the production equipment, p 1 is the predicted value of the quality of the capacitor product, and p 2 is the predicted value of the internal state of the production equipment.
S34: judging whether to enter a step of adjusting equipment parameters according to the working state predicted value; specifically, when the predicted value of the working state is smaller than the threshold value, the working state is normal; and when the predicted value of the working state is larger than the threshold value, indicating that the working state is abnormal, and entering the step of adjusting the parameters of the equipment.
S4: device parameter adjustment
Adjusting production equipment parameters, adjusting the speed and the tension of equipment, and optionally, performing speed and tension adjustment calculation based on an improved swarm intelligence algorithm, wherein the method specifically comprises the following steps of:
S41: and setting initial values, including the maximum iteration times T, the species population number N, and initializing the position of the species and the adaptability function of evaluation.
S42: renewing the pioneer species;
S43: entering a first stage and a foraging stage;
s431: generating a new location of the ith species according to the following formula:
；
Wherein, For the state value of the ith species in the jth dimension in the first stage, x i,j is the state value of the previous state of the ith species in the jth dimension, r is a random number of [0,1], PZ j is the position of the optimal species in the jth dimension, I is a variation range parameter, I=round (1+rand), rand is a random number of [0,1], and the larger I is, the larger the variation range is; p1 represents the first stage;
S432: updating the position of the ith species according to the calculation result of the fitness function, specifically calculating the fitness of the updated species group, comparing with the fitness of the current position, if the fitness after updating is better than the fitness at the current position, defining the position of the individual as the updated position, specifically calculating as follows:
，
Wherein F i is the fitness function value corresponding to the original position, For the fitness function value corresponding to the new position of the first stage, X i is the position of the previous state of the ith species,New positions for the ith species of the first stage; p1 represents the first stage;
S44: entering a second stage: a defensive strategy stage for predators;
441: generating random numbers ；
S442: if the random number P s is less than or equal to 0.5, the new state of the ith species is calculated using the formula:
，
Wherein, For the state value of the ith species in the jth dimension in the second stage, R is a constant, R is a random number of [0,1], T is the current iteration number, and T is the maximum iteration number; p2 represents the second stage;
S443: if the random number P s > 0.5, the new state of the ith species is calculated using the following formula:
，
Wherein, For the state value of the ith species in the jth dimension in the second stage, r is a random number of [0,1], AZ j is the position of the attacked species in the jth dimension, I=round (1+rand), rand is a random number of [0,1], and the larger the I is, the larger the variation range is;
S444: updating the position of the ith species according to the calculation result of the fitness function, specifically calculating the fitness of the updated species group, comparing with the fitness of the current position, if the fitness after updating is better than the fitness at the current position, defining the position of the individual as the updated position, specifically calculating as follows:
，
Wherein F i is the fitness function value corresponding to the original position, For the fitness function value corresponding to the new position of the second stage, X i is the position of the previous state of the ith species,New positions for the second stage i species; p2 represents the second stage;
s45: repeating the steps S43-S44 until N times of circulation are completed;
S46: saving the best candidate solution to date;
S47: repeating the steps S42-S46 until T iterations are completed;
s48: outputting the optimal solution, and adjusting the speeds and the tensions of a feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine and a winding machine of the equipment according to the optimal solution.
The improved swarm intelligence algorithm solves the optimization problem by establishing proper balance between exploration and development, and provides a solution for solving the multi-objective optimization problem.
In the embodiment, the working state is intelligently predicted based on an improved LSTM prediction algorithm by detecting the quality of the capacitor product and the internal state of production equipment, so that objective and accurate prediction of the working state of the nailing machine is realized, then the speed and the tension of the equipment are adjusted by adopting an improved swarm intelligent algorithm according to a prediction result, and proper equipment operation parameters are set, so that the production error of the capacitor nailing machine is reduced.
The embodiment of the invention also provides intelligent control equipment of the capacitor element nailing machine, which comprises the following components:
the production equipment internal state detection module is used for detecting the internal state of the production equipment and comprises a detection feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine and a winding machine.
And the capacitor product quality detection module is used for detecting the capacitor product quality, wherein the capacitor product quality comprises the appearance of the capacitor and the electrical performance.
A working state prediction module for calculating the equipment working state prediction according to the quality of the capacitor product and the detection result of the internal state of the production equipment, adjusting the equipment parameters according to the equipment working state prediction value,
And the equipment parameter adjustment module is used for adjusting the speeds and the tensions of a feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine and a winding machine of the equipment based on the improved swarm intelligence algorithm.
In addition, the embodiment of the invention also provides a computer readable storage medium, and the computer readable storage medium stores program instructions of the intelligent control method of the capacitor sub-nailing machine, and the program instructions of the intelligent control method of the capacitor sub-nailing machine can be executed by one or more processors to realize the steps of the intelligent control method of the capacitor sub-nailing machine.
The above examples are only illustrative of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention, and various modifications and improvements made by those skilled in the art to the technical solution of the present invention should fall within the scope of protection defined by the claims of the present invention without departing from the spirit of the present invention.
Claims (5)
1. The intelligent control method of the capacitor element nailing machine is characterized by comprising the following steps of:
S1: detecting the internal state of production equipment, including detecting the movement speed and the tension of a feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine and a winding machine;
S2: detecting the quality of a capacitor product; the quality of the capacitor product comprises the appearance and the electrical performance of the capacitor, wherein the appearance comprises detection of the position error of a capacitor pin, the diameter of the capacitor and the height error; the electrical property detection comprises detection of capacitance value errors and leakage current errors;
S3: calculating equipment working state prediction according to the quality of the capacitor product and the internal state detection result of the production equipment to obtain a working state prediction value, and when the working state prediction value is smaller than a threshold value, indicating that the working state is normal; when the predicted value of the working state is larger than the threshold value, the working state is abnormal, and the equipment parameter adjustment step is entered;
s4: adjusting equipment parameters, namely adjusting the movement speeds and the tension of a feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine and a winding machine based on an improved group intelligent algorithm;
the step S3 includes:
s31: determining a capacitor product quality predictor and a production facility internal state predictor based on the improved LSTM network;
s32: calculating the weight proportion of the capacitor product quality and the internal state detection result of production equipment;
s33: calculating an operating state predicted value Y based on:
，
Q 1 is the weight occupied by the quality of the capacitor product, q 2 is the weight occupied by the detection result of the internal state of the production equipment, p 1 is the predicted value of the quality of the capacitor product, and p 2 is the predicted value of the internal state of the production equipment;
S34: judging whether to enter a step of adjusting equipment parameters according to the working state predicted value; specifically, when the predicted value of the working state is smaller than the threshold value, the working state is normal; when the predicted value of the working state is larger than the threshold value, the working state is abnormal, and the step of adjusting the equipment parameters is entered;
the step S31 includes: the improved LSTM network structure comprises an input layer, a hidden layer and an output layer; the hidden layers comprise two hidden layers in two directions, namely a front hidden layer and a rear hidden layer; the data of the input layer is transmitted to the front hidden layer or is input into the rear hidden layer;
The step S31 further includes:
S311: establishing an improved LSTM network structure, initializing a network parameter w f、wi、wc and a bias value b f、bi、bc;
Wherein w f、wi、wc is the weight of the forgetting gate, the output gate and the cell state respectively, and b f、bi、bc is the bias of the forgetting gate, the output gate and the cell state respectively;
S312: respectively importing the existing data comprising respective detection parameters, corresponding capacitor product quality predicted values and production equipment internal state predicted values, and dividing the data into a training set and a testing set;
S313: inputting the input vector into an improved LSTM network, calculating the current hidden layer state value of the single-layer hidden layer network, and updating the forgetting gate, the input gate, the output gate and the hidden layer output result of the network;
S314: combining bidirectional loops of forward and backward stacked unidirectional LSTM neural networks to obtain improved LSTM network model output ；
The state of the previous hidden layer is indicated,Representing the state of the post hidden layer;
S315: calculating a weight corresponding to each output vector;
s316: summarizing the input sequence data by using a weighted average formula to obtain a new output value;
S317: finally, the output value is transmitted to an output layer to obtain a prediction result, namely a final output result;
Repeatedly calculating the steps S311-S317 to obtain a capacitor product quality predicted value p 1 and a production equipment internal state predicted value p 2;
the step S315 further includes calculating a weight corresponding to each output vector based on the following formula:
，
Wherein v T is an output vector, h t is the output of the cell at time t, W is an output layer weight, and U is an input layer weight; n is the total number of output vectors, h j is the j-th output vector, softmax () is the normalization function;
the step S316 calculates V based on the following formula:
；
The N is the total number of output vectors, r m is the weight corresponding to the m-th output vector, and h m is the m-th output vector;
The step S32 includes:
And determining the quality of the capacitor product and the weight distribution of the internal state detection result of the production equipment based on an entropy weight method.
2. The intelligent control method of a capacitor element nailing machine according to claim 1, wherein the speed and tension adjustment calculation based on the improved swarm intelligence algorithm in step S4 comprises: the first stage: a foraging stage and a second stage: a defensive strategy stage for predators.
3. The method of intelligent control of a capacitor element nailing machine of claim 2 wherein the first stage comprises:
s431: generating a new location of the ith species according to the formula;
，
Wherein, For the state value of the ith species in the jth dimension in the first stage, x i,j is the state value of the previous state of the ith species in the jth dimension, r is a random number of [0,1], PZ j is the position of the optimal species in the jth dimension, I is a variation range parameter, I=round (1+rand), rand is a random number of [0,1], and the larger I is, the larger the variation range is; p1 represents the first stage;
S432: updating the position of the ith species according to the calculation result of the fitness function, specifically calculating the fitness of the updated species group, comparing with the fitness of the current position, if the fitness after updating is better than the fitness at the current position, defining the position of the individual as the updated position, specifically calculating as follows:
，
Wherein F i is the fitness function value corresponding to the original position, For the fitness function value corresponding to the new position of the first stage, X i is the position of the previous state of the ith species,New positions for the ith species of the first stage; p1 represents the first stage;
The second stage comprises:
S441: generating random numbers ；
S442: if the random number P s is less than or equal to 0.5, the new state of the ith species is calculated using the formula:
，
Wherein, For the state value of the ith species in the jth dimension in the second stage, R is a constant, R is a random number of [0,1], T is the current iteration number, and T is the maximum iteration number; p2 represents the second stage;
s443: if random number The new status of the ith species was calculated using the following formula:
，
Wherein, For the state value of the ith species in the jth dimension in the second stage, r is a random number of [0,1], AZ j is the position of the attacked species in the jth dimension, I=round (1+rand), rand is a random number of [0,1], and the larger the I is, the larger the variation range is;
S444: updating the position of the ith species according to the calculation result of the fitness function, specifically calculating the fitness of the updated species group, comparing with the fitness of the current position, if the fitness after updating is better than the fitness at the current position, defining the position of the individual as the updated position, specifically calculating as follows:
Wherein F i is the fitness function value corresponding to the original position, For the fitness function value corresponding to the new position of the second stage, X i is the position of the previous state of the ith species,New positions for the second stage i species; p2 represents the second stage.
4. An apparatus based on the intelligent control method of the capacitor element nailing machine according to any one of claims 1-3, comprising:
The production equipment internal state detection module is used for detecting the internal state of the production equipment and comprises a detection feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine and a winding machine;
A capacitor product quality detection module for detecting a capacitor product quality, the capacitor product quality including a capacitor appearance and an electrical property;
The working state prediction module is used for calculating equipment working state prediction according to the quality of capacitor products and the detection result of the internal state of production equipment and adjusting equipment parameters according to the equipment working state prediction value;
And the equipment parameter adjustment module is used for adjusting the speeds and the tensions of a feeding mechanism, a positive electrode nailing machine, a negative electrode nailing machine and a winding machine of the equipment based on the improved swarm intelligence algorithm.
5. A computer readable storage medium having stored thereon program instructions of a capacitor sub-reel machine intelligent control method, the program instructions of the capacitor sub-reel machine intelligent control being executable by one or more processors to implement the steps of the capacitor sub-reel machine intelligent control method of any of claims 1-3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410596771.5A CN118213207B (en) | 2024-05-14 | 2024-05-14 | Intelligent control method and equipment for capacitor element nailing machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410596771.5A CN118213207B (en) | 2024-05-14 | 2024-05-14 | Intelligent control method and equipment for capacitor element nailing machine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118213207A CN118213207A (en) | 2024-06-18 |
| CN118213207B true CN118213207B (en) | 2024-08-06 |
Family
ID=91448842
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410596771.5A Active CN118213207B (en) | 2024-05-14 | 2024-05-14 | Intelligent control method and equipment for capacitor element nailing machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118213207B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117786516A (en) * | 2024-01-02 | 2024-03-29 | 国网湖北省电力有限公司超高压公司 | Oil immersed transformer fault diagnosis method based on multi-source data fusion |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017185347A1 (en) * | 2016-04-29 | 2017-11-02 | 北京中科寒武纪科技有限公司 | Apparatus and method for executing recurrent neural network and lstm computations |
| CN112613909A (en) * | 2020-12-21 | 2021-04-06 | 山东麦港数据系统有限公司 | Agricultural product short-term price prediction method and device based on improved LSTM model |
| CN116949546B (en) * | 2023-09-19 | 2023-12-26 | 滨州市华亿电器设备有限公司 | Cleaning control method for outgoing line of capacitor |
-
2024
- 2024-05-14 CN CN202410596771.5A patent/CN118213207B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117786516A (en) * | 2024-01-02 | 2024-03-29 | 国网湖北省电力有限公司超高压公司 | Oil immersed transformer fault diagnosis method based on multi-source data fusion |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118213207A (en) | 2024-06-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111259738B (en) | Face recognition model construction method, face recognition method and related device | |
| CN110473592B (en) | Multi-view human synthetic lethal gene prediction method | |
| CN112052948B (en) | Network model compression method and device, storage medium and electronic equipment | |
| CN108846340A (en) | Face identification method, device and disaggregated model training method, device, storage medium and computer equipment | |
| CN113128671B (en) | Service demand dynamic prediction method and system based on multi-mode machine learning | |
| CN109886343B (en) | Image classification method and device, equipment and storage medium | |
| CN114398049B (en) | Self-adaptive dynamic updating method for digital twin model of discrete manufacturing workshop | |
| CN112529683A (en) | Method and system for evaluating credit risk of customer based on CS-PNN | |
| CN113591988B (en) | Knowledge cognitive structure analysis method, system, computer equipment, medium and terminal | |
| CN112633584A (en) | River sudden water pollution accident water quality prediction method based on improved LSTM-seq2seq model | |
| CN118213207B (en) | Intelligent control method and equipment for capacitor element nailing machine | |
| CN111191722B (en) | Method and device for training prediction model through computer | |
| CN114137967A (en) | Driving behavior decision method based on multi-network joint learning | |
| CN117579500B (en) | Network traffic prediction method, device, equipment and medium | |
| CN118195853A (en) | Intelligent campus information management method and system based on big data | |
| CN117000780A (en) | Self-adaptive cold-rolled strip multichannel shape forecasting method | |
| CN112529637B (en) | Service demand dynamic prediction method and system based on context awareness | |
| CN109993271A (en) | Grey neural network forecasting based on theory of games | |
| CN113095328B (en) | Semantic segmentation method guided by base index and based on self-training | |
| CN115816466A (en) | Method for improving control stability of visual observation robot | |
| CN108921213A (en) | A kind of entity classification model training method and device | |
| CN115293249A (en) | Power system typical scene probability prediction method based on dynamic time sequence prediction | |
| CN115423091A (en) | Conditional antagonistic neural network training method, scene generation method and system | |
| CN114186646A (en) | Block chain abnormal transaction identification method and device, storage medium and electronic equipment | |
| CN118690201B (en) | Feature backward compatible learning method between models, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |