A chemical belt automatic deviation correction control method and system

Abstract

The present invention relates to the technical field of chemical belt deviation correction, and discloses a chemical belt automatic deviation correction control method and system. The method comprises: collecting operating status data of the chemical belt and evaluating the deviation degree of the chemical belt; optimizing model parameters of a constructed belt deviation correction control model; if the operating deviation degree of the chemical belt exceeds a preset deviation threshold, generating deviation correction control parameters for the chemical belt using the belt deviation correction control model after model parameter optimization, and automatically controlling the chemical belt. The present invention calculates the operating deviation degree of the chemical belt based on the difference between the operating status data and the normal operating value, selects more accurate deviation correction control parameters for the chemical belt using a reinforcement learning method, and introduces iterative weight coefficients and oscillation function values ​​in the model parameter optimization process to improve global search and local search capabilities, thereby improving the accuracy of the belt deviation correction control model.

Description

Technical Field

[0001] The present invention relates to the field of chemical belt deviation correction, and in particular to an automatic deviation correction control method and system for a chemical belt. Background Art

[0002] Belt conveyors are widely used in industries such as the chemical, mining, steel, and energy industries, and are essential equipment for material transportation. In the chemical industry in particular, belt conveyors not only transport large quantities of materials but also directly impact the efficiency and safety of production lines. However, conveyor belts often deviate during operation, leading to belt wear, reduced transport efficiency, and even equipment failure, impacting the entire production process. Therefore, effectively addressing belt deviation has long been a key concern for belt conveyors. With the continuous advancement of modern industrial automation, traditional manual control and simple mechanical adjustment methods are no longer able to meet the increasingly complex demands of production. Traditional belt deviation correction methods often rely on mechanical correction devices (such as guide rollers and idlers), which often only provide corrections within a certain range and cannot precisely control the direction and magnitude of belt deviation. Summary of the Invention

[0003] The present invention provides a chemical belt automatic deviation correction control method and system, which monitors the running status of the belt in real time, automatically adjusts equipment parameters, and accurately corrects the belt deviation problem, thereby ensuring the efficient and safe operation of the belt conveyor.

[0004] To achieve the above object, the present invention provides a chemical belt automatic deviation correction control method, comprising the following steps:

[0005] S1: Collect the running status data of the chemical belt, evaluate the deviation degree of the chemical belt, and obtain the running deviation degree of the chemical belt;

[0006] S2: Constructing a belt deviation correction control model, wherein the belt deviation correction control model takes the calculated running deviation degree and running status data as input and takes the deviation correction control parameters as output;

[0007] S3: Optimizing the model parameters of the constructed belt deviation correction control model to obtain the belt deviation correction control model after model parameter tuning, wherein the model parameter tuning adopts the strategy-enhanced particle swarm method;

[0008] S4: If the running deviation of the chemical belt exceeds the preset deviation threshold, the belt deviation control model after model parameter optimization is used to generate the deviation control parameters of the chemical belt, and the chemical belt is automatically corrected and controlled.

[0009] As a further improvement method of the present invention:

[0010] Optionally, a plurality of sensors are used to obtain the operating status data of the chemical belt;

[0011] The monitoring position of the chemical belt is obtained by using a photoelectric sensor, wherein the photoelectric sensor includes a light source and a receiver. The light source is used to send an infrared beam to the edge positions on both sides of the chemical belt. The infrared beam reaches the edge position of the chemical belt to form a reflected light. The receiver is located in the reflection path of the reflected light and is used to receive the reflected light and convert the reflected light into an electrical signal. The amplitude of the electrical signal is used as the monitoring position of the chemical belt. ;

[0012] Place the speed meter on the surface of the chemical belt to measure the speed of the chemical belt ;

[0013] Place the temperature sensor near the chemical belt to measure the temperature of the chemical belt ; Specifically, the temperature sensor is a thermistor;

[0014] Place the microphone near the chemical belt to measure the chemical belt noise .

[0015] Optionally, a deviation degree of the chemical belt is evaluated based on the operating status data to obtain an operating deviation degree of the chemical belt. The operating deviation degree is obtained by weighting four deviation indicators, where the first to fourth deviation indicators are a position deviation indicator, a speed deviation indicator, a temperature deviation indicator, and a noise anomaly indicator, respectively. The deviation indicator of the chemical belt is calculated using the deviation degree evaluation method, wherein the deviation degree evaluation process is as follows:

[0016] Based on chemical belt monitoring position Calculate the position deviation index of chemical belts :

[0017] ;

[0018] in:

[0019] Indicates the monitoring position of the chemical belt under normal operation of the chemical belt. Indicates the preset maximum position deviation value, Indicates the deviation value of the chemical belt monitoring position;

[0020] Based on chemical belt speed Calculate the speed deviation index of chemical belts :

[0021] ;

[0022] in:

[0023] Indicates the set speed;

[0024] Based on chemical belt temperature Calculate the temperature deviation index of chemical belts :

[0025] ;

[0026] in:

[0027] Indicates the temperature of the chemical belt under normal operation. Indicates the preset maximum temperature;

[0028] The chemical belt noise For noise signals, chemical belt noise Perform Fourier transform to obtain the spectrum of Len frequency components, calculate the power of each spectrum, and obtain the power of the average spectrum of the chemical belt noise under normal operation of the chemical belt , which constitutes the abnormal noise index of chemical belt , where Len represents the chemical belt noise The signal length, Indicates the The power of the spectrum, ;

[0029] The weight coefficients of different deviation indicators are determined by using the hierarchical analysis method, and the deviation indicators are weighted. The weighted results are used as the running deviation degree of the chemical belt. .

[0030] Optionally, the belt correction control model is a reinforcement learning model structure, including a state space, an action space and a strategy distribution. The state space includes all states of the chemical belt, and the state is composed of a set of operating state data and the operating deviation degree corresponding to the operating state data. The action space includes all correction control parameters of the chemical belt. The strategy distribution is the probability distribution of selecting different correction control parameters under different states. The strategy distribution is the model parameter to be tuned in the belt correction control model.

[0031] Optionally, a reward function is constructed for evaluating the corrective control effect of the corrective control parameters selected in the strategy distribution, wherein the variables of the reward function include the state of the chemical belt and the corrective control parameters, wherein the expression of the reward function is:

[0032] ;

[0033] ;

[0034] in:

[0035] S represents the status of the chemical belt, A represents the deviation correction control parameter of the chemical belt, Indicates the reward function value after selecting the correction control parameter A to correct the chemical belt in state S;

[0036] Indicates the jth deviation index of the chemical belt after the deviation correction control parameter A is selected to correct the deviation of the chemical belt in state S, where ,The 1st to 4th deviation indicators are position deviation indicator, speed deviation indicator, temperature deviation indicator and noise anomaly indicator, represents the indicator weight coefficient of the j-th deviation indicator;

[0037] Indicates that in state S, after selecting the deviation correction control parameter A to correct the deviation of the chemical belt, the running deviation of the chemical belt is: represents the exponential decay control parameter, represents the interaction effect term of selecting the correction control parameter A under state S, Represents the effect control parameter.

[0038] Optionally, a training loss function for the policy distribution is constructed based on the reward function:

[0039] ;

[0040] ;

[0041] ;

[0042] ;

[0043] in:

[0044] is the strategy distribution, represents the nth state in the state space, , N represents the number of states in the state space, represents the nth correction control parameter in the action space, , M represents the number of correction control parameters in the action space; Indicates status Select the correction control parameters probability;

[0045] express The log probability gradient of Show The logarithmic probability of Indicates that the status Next, select the correction control parameters The reward function value after correcting the chemical belt;

[0046] Indicates that the status Under the deviation control parameters advantage value.

[0047] Optionally, the training loss function is solved, and the solution result is used as the model parameter tuning result of the belt deviation correction control model, wherein the solution process of the loss function is:

[0048] Initialize H groups of particles. Each group of particles includes particle position and particle velocity. The particle position is the corresponding result of a set of strategy distributions. Set the maximum number of iterations of the particle iteration to Max.

[0049] Extracting the particle positions of the iteratively obtained particles and normalizing the particle positions so that the sum of the element values ​​in each row of the particle positions is 1, using the normalized particle positions as the strategy distribution in the training loss function, using the function value of the training loss function as the fitness function value of the particles, and selecting the particle position with the highest fitness function value as the optimal particle position obtained by iteration, and iterating the particles;

[0050] The positions of the H groups of particles that reach the maximum number of iterations are obtained, and the positions of the H groups of particles are standardized. The particle position with the highest fitness function value is selected as the model parameter tuning result of the belt deviation correction control model, and the model parameter tuning result is used as the strategy distribution of the belt deviation correction control model to construct the belt deviation correction control model after model parameter tuning.

[0051] Optionally, the deviation correction control parameter is the offset of the center of the chemical belt to the right. If the offset is negative, it is offset to the left, and the left offset is the absolute value of the offset. If the running deviation degree of the chemical belt calculated in step S1 is If the deviation exceeds the preset threshold, the speed of the chemical belt is reduced until the preset deviation speed is reached, and the belt deviation correction control model after the model parameters are tuned is used to generate the deviation correction control parameters of the chemical belt.

[0052] In order to solve the above problems, the present invention also provides a chemical belt automatic deviation correction control system, which includes a server and a data storage device. The server includes an operation deviation evaluation module and an automatic deviation correction module:

[0053] The operation deviation evaluation module is used to collect the operation status data of the chemical belt and evaluate the deviation degree of the chemical belt;

[0054] The automatic deviation correction module is used to dispatch the belt deviation correction control model after the model parameters are tuned from the data storage device. If the running deviation degree of the chemical belt exceeds the preset deviation threshold, the belt deviation correction control model after the model parameters are tuned is used to generate the chemical belt deviation control parameters, and automatically correct the chemical belt;

[0055] The data storage device is used to optimize the model parameters of the constructed belt deviation correction control model and store the belt deviation correction control model after the model parameters are optimized;

[0056] To realize any of the above-mentioned chemical belt automatic deviation correction control methods.

[0057] In order to solve the above problem, the present invention further provides an electronic device, comprising:

[0058] a memory storing at least one instruction;

[0059] Communication interfaces to enable electronic equipment to communicate; and

[0060] The processor executes the instructions stored in the memory to implement the above-mentioned chemical belt automatic deviation correction control method.

[0061] In order to solve the above problems, the present invention also provides a computer-readable storage medium, which stores at least one instruction. The at least one instruction is executed by a processor in an electronic device to implement the above-mentioned chemical belt automatic deviation correction control method.

[0062] Compared with the existing technology, the present invention proposes a chemical belt automatic deviation correction control method and system, which has the following advantages:

[0063] First, this scheme proposes a method for quantifying operation deviation and constructing a reward function. It uses multiple sensors to obtain the operation status data of the chemical belt in multiple dimensions during operation. Based on the difference between the operation status data and the normal value, the operation deviation degree that characterizes the degree to which the chemical belt deviates from the normal value is calculated. The operation status data and the operation deviation degree of the chemical belt are used as the state, and the correction control parameters that accurately control the belt correction range are used as the action. The reward function is constructed using reinforcement learning to perform automatic correction control of the chemical belt. In the process of constructing the reward function, the square term is used to characterize various deviation indicators to improve the sensitivity to large deviation indicators, thereby reducing the deviation value of large deviation indicators. The exponential decay term is used to reward the state with overall low operation deviation, and the correction control parameters are encouraged to reduce the overall operation deviation. The interaction effect term is used to capture the mutual influence between different deviation indicators, and it is encouraged to reduce the negative interaction between deviation indicators.

[0064] At the same time, this scheme proposes a strategy distribution tuning method, which uses the strategy-enhanced particle swarm method to tune the strategy distribution, generates the iterative weight coefficient of the particle speed based on the fitness function value of the particle, and adaptively adjusts the iterative weight of each particle speed, which is beneficial to protecting particles with better particle positions and also beneficial to particles with poor particle positions moving toward better particle positions. At the same time, an oscillation function value of the particle position is proposed, and the difference between the particle position and the optimal particle position is oscillated based on the number of particle iterations. When the number of particle iterations is small, the oscillation coefficient is large, so that the particle speed jumps out of the local optimum and improves the global search ability of the particle. When the number of particle iterations is large, the oscillation coefficient is small, which improves the local search ability of the particle. If the degree of deviation of the chemical belt exceeds the preset deviation threshold, the belt correction control model after model parameter tuning is used to generate the chemical belt correction control parameters, and perform more precise correction control on the chemical belt. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 A schematic flow chart of a method for automatically correcting deviation of a chemical belt provided by one embodiment of the present invention;

[0066] Figure 2 This is a functional module diagram of a chemical belt automatic deviation correction control system provided by one embodiment of the present invention;

[0067] Figure 2 In: 100 chemical belt automatic deviation correction control system, 101 operation deviation evaluation module, 102 automatic deviation correction module, 103 data storage device;

[0068] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0069] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0070] The present embodiment provides a method for automatically correcting the deviation of a chemical belt. The method can be executed by at least one of the following electronic devices, including a server and a terminal, that can be configured to execute the method provided by the present embodiment. In other words, the method can be executed by software or hardware installed on a terminal or server device, where the software can be a blockchain platform. The server can include, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0071] Reference Figure 1 Embodiment 1 of the present invention provides a method for automatically correcting the deviation of a chemical belt, comprising the following steps:

[0072] S1: Collect the running status data of the chemical belt, evaluate the deviation degree of the chemical belt, and obtain the running deviation degree of the chemical belt.

[0073] Using multiple sensors to obtain the operating status data of the chemical belt;

[0074] The monitoring position of the chemical belt is obtained by using a photoelectric sensor, wherein the photoelectric sensor includes a light source and a receiver. The light source is used to send an infrared beam to the edge positions on both sides of the chemical belt. The infrared beam reaches the edge position of the chemical belt to form a reflected light. The receiver is located in the reflection path of the reflected light and is used to receive the reflected light and convert the reflected light into an electrical signal. The amplitude of the electrical signal is used as the monitoring position of the chemical belt. Specifically, the light source is a laser diode, and the receiver includes a photodiode and a signal processing circuit. The photodiode is used to convert reflected light into current, and the signal processing circuit is used to amplify, filter, and shape the current to obtain an electrical signal. If the electrical signal has large fluctuations, it indicates that there is a deviation between the edges of the two sides of the chemical belt.

[0075] Place the speed meter on the surface of the chemical belt to measure the speed of the chemical belt ;

[0076] Place the temperature sensor near the chemical belt to measure the temperature of the chemical belt ; Specifically, the temperature sensor is a thermistor;

[0077] Place the microphone near the chemical belt to measure the chemical belt noise .

[0078] The deviation degree of the chemical belt is evaluated based on the operating status data to obtain the operating deviation degree of the chemical belt. The operating deviation degree is obtained by weighting four deviation indicators. The first to fourth deviation indicators are position deviation indicator, speed deviation indicator, temperature deviation indicator and noise abnormality indicator, respectively. The deviation indicator of the chemical belt is calculated using the deviation degree evaluation method. The deviation degree evaluation process is as follows:

[0079] Based on chemical belt monitoring position Calculate the position deviation index of chemical belts :

[0080] ;

[0081] in:

[0082] Indicates the monitoring position of the chemical belt under normal operation of the chemical belt. Indicates the preset maximum position deviation value, Indicates the deviation value of the chemical belt monitoring position;

[0083] Based on chemical belt speed Calculate the speed deviation index of chemical belts :

[0084] ;

[0085] in:

[0086] Indicates the set speed;

[0087] Based on chemical belt temperature Calculate the temperature deviation index of chemical belts :

[0088] ;

[0089] in:

[0090] Indicates the temperature of the chemical belt under normal operation. Indicates the preset maximum temperature;

[0091] The chemical belt noise For noise signals, chemical belt noise Perform Fourier transform to obtain the spectrum of Len frequency components, calculate the power of each spectrum, and obtain the power of the average spectrum of the chemical belt noise under normal operation of the chemical belt , which constitutes the abnormal noise index of chemical belt , where Len represents the chemical belt noise The signal length, Indicates the The power of the spectrum, ;

[0092] The weight coefficients of different deviation indicators are determined by using the hierarchical analysis method, and the deviation indicators are weighted. The weighted results are used as the running deviation degree of the chemical belt. .

[0093] S2: Constructing a belt deviation correction control model, wherein the belt deviation correction control model takes the calculated running deviation degree and running status data as input and takes the deviation correction control parameters as output.

[0094] The belt deviation correction control model is a reinforcement learning model structure, including a state space, an action space and a strategy distribution. The state space includes all states of the chemical belt, and the state is composed of a set of operating state data and the operating deviation degree corresponding to the operating state data. The action space includes all deviation correction control parameters of the chemical belt. The strategy distribution is the probability distribution of selecting different deviation correction control parameters under different states. The strategy distribution is the model parameter to be tuned in the belt deviation correction control model.

[0095] A reward function is constructed for evaluating the corrective control effect of the corrective control parameters selected in the strategy distribution. The variables of the reward function include the state of the chemical belt and the corrective control parameters. The expression of the reward function is:

[0096] ;

[0097] ;

[0098] in:

[0099] S represents the status of the chemical belt, A represents the deviation correction control parameter of the chemical belt, Indicates the reward function value after selecting the correction control parameter A to correct the chemical belt in state S;

[0100] Indicates the jth deviation index of the chemical belt after the deviation correction control parameter A is selected to correct the deviation of the chemical belt in state S, where ,The 1st to 4th deviation indicators are position deviation indicator, speed deviation indicator, temperature deviation indicator and noise anomaly indicator, represents the indicator weight coefficient of the j-th deviation indicator;

[0101] Indicates that in state S, after selecting the deviation correction control parameter A to correct the deviation of the chemical belt, the running deviation of the chemical belt is: represents the exponential decay control parameter, represents the interaction effect term of selecting the correction control parameter A under state S, Represents the effect control parameter. Specifically, this scheme uses square terms to characterize various deviation indicators, increasing sensitivity to large deviation indicators and thereby reducing the deviation values ​​of large deviation indicators. It uses exponential decay terms to reward states with low overall operating deviations, encouraging the correction control parameters to reduce the overall operating deviation. It uses interaction effect terms to capture the mutual influence between different deviation indicators and encourage the reduction of negative interactions between deviation indicators.

[0102] S3: Optimizing model parameters of the constructed belt deviation correction control model to obtain the belt deviation correction control model after the model parameters are optimized.

[0103] Construct the training loss function of the policy distribution based on the reward function:

[0104] ;

[0105] ;

[0106] ;

[0107] ;

[0108] in:

[0109] is the strategy distribution, represents the nth state in the state space, , N represents the number of states in the state space, represents the nth correction control parameter in the action space, , M represents the number of correction control parameters in the action space; Indicates status Select the correction control parameters probability;

[0110] express The log probability gradient of Show The logarithmic probability of Indicates that the status Next, select the correction control parameters The reward function value after correcting the chemical belt;

[0111] Indicates that the status Under the deviation control parameters advantage value.

[0112] The training loss function is solved, and the solution result is used as the model parameter tuning result of the belt deviation correction control model. The solution process of the loss function is:

[0113] Initialize H groups of particles. Each group of particles includes particle position and particle velocity. The particle position is the corresponding result of a set of strategy distributions. Set the maximum number of iterations of the particle iteration to Max.

[0114] Extracting the particle positions of the iteratively obtained particles and normalizing the particle positions so that the sum of the element values ​​in each row of the particle positions is 1, using the normalized particle positions as the strategy distribution in the training loss function, using the function value of the training loss function as the fitness function value of the particles, and selecting the particle position with the highest fitness function value as the optimal particle position obtained by iteration, and iterating the particles;

[0115] As a preferred embodiment of the present invention, the present invention generates an iterative weight coefficient of particle velocity based on the fitness function value of the particle, and adaptively adjusts the iterative weight of each particle velocity, which is beneficial to protecting particles with better particle positions and also beneficial to particles with worse particle positions moving toward better particle positions. At the same time, an oscillation function value of the particle position is proposed, and the difference between the particle position and the optimal particle position is oscillated based on the number of particle iterations. When the number of particle iterations is small, the oscillation coefficient is large, thereby causing the particle speed to jump out of the local optimum and improve the global search ability of the particle. When the number of particle iterations is large, the oscillation coefficient is small, thereby improving the local search ability of the particle. The iterative formula of the hth group of particles is:

[0116] ;

[0117] ;

[0118] ;

[0119] ;

[0120] in:

[0121] represents the particle position of the hth group of particles obtained in the tth iteration, represents the particle velocity of the hth group of particles obtained in the tth iteration, , ;

[0122] represents the fitness function value of the hth group of particles obtained in the tth iteration, Indicates particle speed The corresponding iterative weight coefficient, Represents the preset minimum iteration weight coefficient, Indicates the preset maximum iteration weight coefficient, represents the mean value of the fitness function of the H group of particles obtained in the t-th iteration, It represents the minimum fitness function value among the H group of particles obtained in the t-th iteration;

[0123] represents the control parameter, Represents a random number between 0 and 1. represents the optimal particle position after the tth iteration, Indicates the oscillation coefficient of the t-th iteration. If t is less than ,but , if t is greater than or equal to ,but ;

[0124] Represents the particle position The oscillation function value of ;

[0125] The positions of the H groups of particles that reach the maximum number of iterations are obtained, and the positions of the H groups of particles are standardized. The particle position with the highest fitness function value is selected as the model parameter tuning result of the belt deviation correction control model, and the model parameter tuning result is used as the strategy distribution of the belt deviation correction control model to construct the belt deviation correction control model after model parameter tuning.

[0126] S4: If the running deviation of the chemical belt exceeds the preset deviation threshold, the belt deviation control model after model parameter optimization is used to generate the deviation control parameters of the chemical belt, and the chemical belt is automatically corrected and controlled.

[0127] The deviation correction control parameter is the offset of the center of the chemical belt to the right. If the offset is negative, it will offset to the left, and the left offset is the absolute value of the offset. If the deviation of the chemical belt calculated in step S1 is If the deviation exceeds the preset threshold, the speed of the chemical belt is reduced until the preset deviation speed is reached, and the belt deviation correction control model after the model parameters are tuned is used to generate the deviation correction control parameters of the chemical belt.

[0128] Example 2:

[0129] like Figure 2 1 is a functional module diagram of a chemical belt automatic deviation correction control system 100 provided in one embodiment of the present invention, which can implement the chemical belt automatic deviation correction control method in Example 1.

[0130] Depending on the functions to be implemented, the chemical belt automatic deviation correction control system 100 may include an operating deviation assessment module 101, an automatic deviation correction module 102, and a data storage device 103. The modules described herein, also known as units, are a series of computer program segments that can be executed by an electronic device processor and perform a fixed function.

[0131] The operation deviation evaluation module 101 is used to collect the operation status data of the chemical belt and evaluate the deviation degree of the chemical belt;

[0132] The automatic deviation correction module 102 is used to dispatch the belt deviation correction control model after the model parameters are tuned from the data storage device. If the running deviation degree of the chemical belt exceeds the preset deviation threshold, the belt deviation correction control model after the model parameters are tuned is used to generate the chemical belt deviation control parameters, and automatically correct the chemical belt.

[0133] The data storage device 103 is used to optimize the model parameters of the constructed belt deviation correction control model and store the belt deviation correction control model after the model parameters are optimized.

[0134] In detail, the modules in the chemical belt automatic deviation correction control system 100 in the embodiment of the present invention are used in the same manner as above. Figure 1 The same technical means are used as the automatic deviation correction control method for chemical belts described in , and can produce the same technical effects, so I will not go into details here.

[0135] It should be understood that the embodiment is for illustration only and the scope of the patent application is not limited to this structure.

[0136] It should be noted that the serial numbers of the above-mentioned embodiments of the present invention are for descriptive purposes only and do not represent the advantages or disadvantages of the embodiments. In addition, the terms "including", "comprising" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "including a ..." does not exclude the presence of other identical elements in the process, device, article or method comprising the element.

[0137] Through the description of the above embodiments, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus the necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better embodiment. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in each embodiment of the present invention.

[0138] The above are only preferred embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A chemical belt automatic deviation correction control method, characterized in that: The method comprises: S1: Collecting running status data of the chemical belt, evaluating the deviation degree of the chemical belt, and obtaining the running deviation degree of the chemical belt, wherein the running status data includes the monitoring position of the chemical belt, the speed of the chemical belt, the temperature of the chemical belt, and the noise of the chemical belt; S2: Constructing a belt deviation correction control model, wherein the belt deviation correction control model takes the calculated running deviation degree and running status data as input and takes the deviation correction control parameters as output; S3: Optimizing model parameters of the constructed belt deviation correction control model to obtain a belt deviation correction control model after the model parameters are optimized; S4: If the running deviation of the chemical belt exceeds the preset deviation threshold, the belt deviation control model after model parameter optimization is used to generate the deviation control parameters of the chemical belt, and the chemical belt is automatically corrected; The belt deviation correction control model is a reinforcement learning model structure, including a state space, an action space, and a strategy distribution. The state space includes all states of the chemical belt, and the state consists of a set of operating state data and the operating deviation degree corresponding to the operating state data. The action space includes all the deviation correction control parameters of the chemical belt. The strategy distribution is the probability distribution of selecting different deviation correction control parameters under different states. The strategy distribution is the model parameter to be tuned in the belt deviation correction control model. A reward function is constructed for evaluating the corrective control effect of the corrective control parameters selected in the strategy distribution. The variables of the reward function include the state of the chemical belt and the corrective control parameters. The expression of the reward function is: ； ； in: S represents the state of the chemical belt, A represents the deviation correction control parameter of the chemical belt, Indicates the reward function value after selecting the correction control parameter A to correct the chemical belt in state S; Indicates the jth deviation index of the chemical belt after the deviation correction control parameter A is selected to correct the deviation of the chemical belt in state S, where ,The 1st to 4th deviation indicators are position deviation indicator, speed deviation indicator, temperature deviation indicator and noise anomaly indicator, represents the indicator weight coefficient of the j-th deviation indicator; Indicates that in state S, after selecting the deviation correction control parameter A to correct the deviation of the chemical belt, the running deviation of the chemical belt is: represents the exponential decay control parameter, represents the interaction effect term of selecting the correction control parameter A under state S, Represents the effect control parameter.

2. A chemical belt automatic deviation correction control method according to claim 1, characterized in that: Using multiple sensors to obtain the operating status data of the chemical belt; The monitoring position of the chemical belt is obtained by using a photoelectric sensor, wherein the photoelectric sensor includes a light source and a receiver. The light source is used to send an infrared beam to the edge positions on both sides of the chemical belt. The infrared beam reaches the edge position of the chemical belt to form a reflected light. The receiver is located in the reflection path of the reflected light and is used to receive the reflected light and convert the reflected light into an electrical signal. The amplitude of the electrical signal is used as the monitoring position of the chemical belt. ; Place the speed meter on the surface of the chemical belt to measure the speed of the chemical belt ; Place the temperature sensor near the chemical belt to measure the temperature of the chemical belt ; Specifically, the temperature sensor is a thermistor; Place the microphone near the chemical belt to measure the chemical belt noise .

3. A chemical belt automatic deviation correction control method according to claim 2, characterized in that: Based on the running status data, the deviation degree of the chemical belt is evaluated to obtain the running deviation degree of the chemical belt. The running deviation degree is obtained by weighting four deviation indicators. The first to fourth deviation indicators are position deviation indicator, speed deviation indicator, temperature deviation indicator and noise abnormality indicator, respectively. The deviation index of the chemical belt is calculated using the deviation degree evaluation method. The deviation degree evaluation process is as follows: Based on chemical belt monitoring position Calculate the position deviation index of chemical belts : ； in: Indicates the monitoring position of the chemical belt under normal operation of the chemical belt. Indicates the preset maximum position deviation value, Indicates the deviation value of the chemical belt monitoring position; Based on chemical belt speed Calculate the speed deviation index of chemical belts : ； in: Indicates the set speed; Based on chemical belt temperature Calculate the temperature deviation index of chemical belts : ； in: Indicates the temperature of the chemical belt under normal operation. Indicates the preset maximum temperature; The chemical belt noise For noise signals, chemical belt noise Perform Fourier transform to obtain the spectrum of Len frequency components, calculate the power of each spectrum, and obtain the power of the average spectrum of the chemical belt noise under normal operation of the chemical belt , which constitutes the abnormal noise index of chemical belt , where Len represents the chemical belt noise The signal length, Indicates the The power of the spectrum, ; The weight coefficients of different deviation indicators are determined by using the hierarchical analysis method, and the deviation indicators are weighted. The weighted results are used as the running deviation degree of the chemical belt. .

4. A chemical belt automatic deviation correction control method according to claim 1, characterized in that: Construct the training loss function of the policy distribution based on the reward function: ； ； ； ； in: is the strategy distribution, represents the nth state in the state space, , N represents the number of states in the state space, represents the nth correction control parameter in the action space, , M represents the number of correction control parameters in the action space; Indicates status Select the correction control parameters probability; express The log probability gradient of Show The logarithmic probability of Indicates that the status Next, select the correction control parameters The reward function value after correcting the chemical belt; Indicates that the status Under the deviation control parameters advantage value.

5. A chemical belt automatic deviation correction control method according to claim 4, characterized in that: The training loss function is solved, and the solution result is used as the model parameter tuning result of the belt deviation correction control model. The solution process of the loss function is: Initialize H groups of particles. Each group of particles includes particle position and particle velocity. The particle position is the corresponding result of a set of strategy distributions. Set the maximum number of iterations of the particle iteration to Max. Extracting the particle positions of the iteratively obtained particles and normalizing the particle positions so that the sum of the element values ​​in each row of the particle positions is 1, using the normalized particle positions as the strategy distribution in the training loss function, using the function value of the training loss function as the fitness function value of the particles, and selecting the particle position with the highest fitness function value as the optimal particle position obtained by iteration, and iterating the particles; The positions of the H groups of particles that reach the maximum number of iterations are obtained, and the positions of the H groups of particles are standardized. The particle position with the highest fitness function value is selected as the model parameter tuning result of the belt deviation correction control model, and the model parameter tuning result is used as the strategy distribution of the belt deviation correction control model to construct the belt deviation correction control model after model parameter tuning.

6. A chemical belt automatic deviation correction control method according to claim 1, characterized in that: The deviation correction control parameter is the offset of the center of the chemical belt to the right. If the offset is negative, it will offset to the left, and the left offset is the absolute value of the offset. If the deviation of the chemical belt calculated in step S1 is If the deviation exceeds the preset threshold, the speed of the chemical belt is reduced until the preset deviation speed is reached, and the belt deviation correction control model after the model parameters are tuned is used to generate the deviation correction control parameters of the chemical belt.

7. A chemical belt automatic deviation correction control system, characterized in that: The chemical belt automatic deviation correction control system includes a server and a data storage device. The server includes an operation deviation evaluation module and an automatic deviation correction module: The operation deviation evaluation module is used to collect the operation status data of the chemical belt and evaluate the deviation degree of the chemical belt; The automatic deviation correction module is used to dispatch the belt deviation correction control model after the model parameters are tuned from the data storage device. If the running deviation degree of the chemical belt exceeds the preset deviation threshold, the belt deviation correction control model after the model parameters are tuned is used to generate the chemical belt deviation control parameters, and automatically correct the chemical belt; The data storage device is used to optimize the model parameters of the constructed belt deviation correction control model and store the belt deviation correction control model after the model parameters are optimized; To realize an automatic deviation correction control method for a chemical belt as described in any one of claims 1-6.

Images