Battery replacement cabinet temperature control method and system based on genetic algorithm optimized PID parameters

Abstract

The application discloses a battery replacement cabinet temperature control method and system based on genetic algorithm optimization of PID parameters, and relates to the technical field of intelligent temperature control, and comprises the following steps: constructing a data-driven model through historical charging and discharging data and multi-source temperature data of batteries in a battery replacement cabinet, and determining whether the current cabinet environment temperature can meet the preset charging and discharging requirements of the batteries; if not, generating a PID parameter adjustment signal; establishing a battery replacement cabinet PID temperature control model according to the relationship between temperature change and control input, adjusting the PID parameters through an improved genetic algorithm, obtaining an optimal PID parameter combination, and generating a temperature control signal; and using the temperature control signal to control the temperature of the battery replacement cabinet, capturing temperature abnormal events during the temperature control process, and giving early warning. The application realizes real-time and intelligent processing of battery replacement cabinet temperature sampling and data, synchronous adjustment of PTC heating with temperature change, accurate control of the cabinet temperature, and provision of a stable and reliable charging and discharging environment for the batteries.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent temperature control, more particularly, to a battery swap cabinet temperature control method and system based on genetic algorithm optimized PID parameters. BACKGROUND

[0002] With the popularity of electric bicycles, the battery swap cabinet as an important facility for electric bicycle energy supply directly affects user experience and battery life. In cold northern regions, low ambient temperature can cause battery performance to decline, reducing charging and discharging efficiency, and even affecting battery life. Currently, the common heating method of the battery swap cabinet mainly uses PTC heating or resistance wire heating, and uses simple on-off control or temperature interval control.

[0003] However, the common heating method of the battery swap cabinet has the following disadvantages: low temperature control precision: on-off control and temperature interval control are difficult to achieve precise temperature control, which can easily cause large fluctuations in the temperature inside the cabinet, affecting battery charging and discharging efficiency and life. High energy consumption: traditional heating methods lack intelligent control, which can easily cause energy waste. Slow response speed: when the ambient temperature or battery temperature changes, traditional heating methods are difficult to respond quickly, causing temperature regulation lag. Therefore, there is an urgent need for a battery swap cabinet temperature control system based on a PID temperature closed-loop algorithm to mainly solve the current battery swap cabinet temperature control capability deficiency in low temperature environments, which seriously affects the normal operation of the battery, and has the problems of scattered temperature control function, imperfect protection mechanism, temperature protection misoperation, and charging and discharging, to ensure the 1C high-power continuous charging and discharging demand of lithium ion batteries. SUMMARY

[0004] To solve the above technical problems, the present application provides a battery swap cabinet temperature control method and system based on genetic algorithm optimized PID parameters, which centrally samples the temperature of external environment, cabinet environment, battery and charger, and processes data in real time and intelligently, adjusts the environment heating synchronously with temperature changes, realizes precise control of the temperature inside the cabinet, and provides a stable and reliable charging and discharging environment for the battery.

[0005] The present application provides a battery swap cabinet temperature control method based on genetic algorithm optimized PID parameters in the first aspect, comprising the following steps:

[0006] A data-driven model is constructed by using historical charging and discharging data of the battery in the battery swap cabinet and multi-source temperature data to determine whether the current cabinet environment temperature can meet the preset charging and discharging requirements of the battery;

[0007] If the current temperature environment cannot meet the preset charging and discharging requirements of the battery, a PID parameter adjustment signal is generated;

[0008] A PID temperature control model of the battery swap cabinet is established according to the relationship between temperature change and temperature control, the improved genetic algorithm is used to set the PID parameters, the optimal PID parameter combination is obtained, and the temperature control signal output by the PID temperature control model of the battery swap cabinet is obtained by using the optimal PID parameter combination;

[0009] The temperature control signal is used for temperature control of the battery swap cabinet, and the temperature of the battery swap cabinet is continuously monitored during the temperature control process, and the temperature abnormality event is prewarned and visually displayed.

[0010] In the scheme, a data-driven model is constructed by using historical charge and discharge data of the battery in the battery swap cabinet and multi-source temperature data, specifically:

[0011] The historical charge and discharge data of the battery swap cabinet and the monitored external environment temperature, the cabinet environment temperature, the battery and the charger temperature in the historical charge and discharge process are obtained, and the normal charge and discharge data meeting the preset battery charge and discharge requirements are selected from the historical charge and discharge data after data preprocessing;

[0012] The selected normal charge and discharge data are matched with the multi-source temperature data monitored in the historical charge and discharge process to generate data samples, the data samples are clustered, the temperature intervals are divided according to the climate information of the region where the battery swap cabinet is located and the temperature rise change of the battery charge and discharge in the cabinet, and the initial cluster centers are constructed using the temperature intervals;

[0013] The data samples are divided into the nearest initial cluster centers to generate cluster clusters, and the cluster clusters and cluster centers are updated through iterative clustering, and the clustering results of the last clustering are read and output after the preset iteration times are met, and the temperature working condition labels are set for the data samples;

[0014] The data samples with temperature working condition labels are used to train a stacked denoising autoencoder to construct a data-driven model, the input data samples are randomly set to zero, the features are extracted through a plurality of denoising autoencoders connected in an end-to-end manner, the output of the last autoencoder is imported into a extreme learning machine to judge whether the preset charge and discharge requirements of the battery are met, and when the classification performance of the data-driven model meets the performance standard, the trained data-driven model is output.

[0015] In the scheme, whether the current cabinet environment temperature can meet the preset charge and discharge requirements of the battery is judged, specifically:

[0016] The current multi-source temperature data of the battery swap cabinet and the battery charge and discharge data are collected and imported into the data-driven model, the features are extracted using the data-driven model, the temperature working condition label corresponding to the temperature environment of the battery swap cabinet is judged, and the data is reconstructed using the extracted features;

[0017] The multi-source temperature estimation value and the charge-discharge data estimation value of the battery in the normal charge-discharge state under the temperature working condition label are obtained by the decoder according to the feature distribution of the input data, and the residual error of the multi-source temperature estimation value and the charge-discharge data estimation value and the input data is obtained by using the extreme learning machine;

[0018] The residual error corresponding to the multi-source temperature data and the battery charge-discharge data is analyzed, and when the residual error is greater than the preset residual error threshold, it is determined that the current cabinet environment temperature cannot meet the preset charge-discharge requirement of the battery, and a PID parameter adjustment signal is generated according to the residual error corresponding to the multi-source temperature data.

[0019] In the scheme, a battery swap cabinet PID temperature control model is established according to the relationship between temperature change and temperature control, specifically:

[0020] The given temperature target value and the temperature measured value in the historical temperature adjustment of the battery swap cabinet under low temperature conditions are obtained, and a data set is constructed by combining the temperature difference between the given temperature target value and the temperature measured value with the corresponding control amount; and the training set and the validation set are divided according to the preset proportion;

[0021] The battery swap cabinet PID temperature control model is constructed by a deep belief network, the restricted Boltzmann machine in the deep belief network is trained by the training set, the PID parameters are calculated in each layer of the restricted Boltzmann machine, and the control amount is generated according to the PID parameters and is imported into the next layer of the restricted Boltzmann machine for layer-by-layer training;

[0022] After layer-by-layer iterative training, the weights of different layers of the restricted Boltzmann machine are updated, a deep nonlinear mapping between temperature change and temperature control is constructed, the constructed battery swap cabinet PID temperature control model is verified by the validation set, and the weights of the model are adjusted by using the verification result.

[0023] In the scheme, the PID parameters are set by an improved genetic algorithm in the battery swap cabinet PID temperature control model, and the optimal PID parameter combination is obtained, specifically:

[0024] In the training process of the battery swap cabinet PID temperature control model, the PID controller parameters of the deep belief network are optimized by the improved genetic algorithm, the PID parameters are genetically coded by chaotic mapping, and the chromosome population is initialized by elite reverse learning;

[0025] According to the temperature control accuracy, the temperature control speed and the temperature control stability, an fitness function is constructed, the population position of the chromosome is updated by selection and crossover operation according to the fitness value, the current optimal solution is selected, and the current optimal solution is disturbed and updated by introducing Cauchy mutation in the iteration process.

[0026] The fitness values before and after the disturbance are compared, the optimal solution is selected through the comparison result, the fitness of the individual is continuously calculated through iteration, and the next generation population is generated through selection, crossover and Cauchy mutation operation of the chromosome, when the fitness value is unchanged for a continuous number of times reaches a preset number of threshold values, the optimal PID parameter combination is obtained, and the existing parameters of the battery replacement cabinet PID temperature control model are replaced.

[0027] In the scheme, the temperature control signal output by the battery replacement cabinet PID temperature control model using the optimal PID parameter combination is used for temperature control, specifically:

[0028] According to the temperature change value of the given temperature target value and the current temperature measured value parsed by the PID parameter adjustment signal, the battery replacement cabinet PID temperature control model is imported for differential, proportional and integral operation, and the control quantity is obtained to generate a temperature control signal and send it to the temperature rise equipment for battery replacement cabinet environment temperature control.

[0029] In the scheme, the temperature of the battery replacement cabinet is continuously monitored during the temperature control process, and temperature abnormal events are warned and visualized, specifically:

[0030] The historical temperature rise data of the battery replacement cabinet temperature control equipment is obtained, the temperature change value in a unit time is obtained based on the historical temperature rise data, the threshold information of the temperature control monitoring task is established according to the temperature change value in the unit time, and it is judged whether the temperature change value of the cabinet environment temperature in the unit time is less than the threshold information, and if it is less than the threshold information, the temperature control equipment abnormal information is generated;

[0031] The temperature data critical value before and after the battery thermal runaway is extracted through the battery thermal runaway example, the cumulative probability distribution function of the temperature data critical value is constructed, and the thermal runaway occurrence probability corresponding to the current cabinet environment temperature is analyzed according to the cumulative probability distribution function;

[0032] If the thermal runaway occurrence probability is greater than a preset probability threshold, battery thermal runaway abnormal information is generated, the temperature control equipment abnormal information and the battery thermal runaway abnormal information are warned, and the warning is visualized.

[0033] Compared with the prior art, the beneficial effects of the present application are:

[0034] The present application adjusts the heating power in real time according to the temperature deviation, realizes the accurate control of the temperature in the battery replacement cabinet, avoids the temperature fluctuation, improves the battery charging and discharging efficiency and service life, dynamically adjusts the heating power according to the actual demand, avoids the energy waste, improves the heating efficiency, quickly responds to the changes of the environment temperature and the battery temperature, adjusts the heating power in time, ensures the stability of the temperature in the battery replacement cabinet, realizes the automatic control of the battery replacement cabinet heating system by using the genetic algorithm and the deep belief network for PID parameter self-tuning, guarantees the speed, control accuracy and stability of the temperature control, reduces the manual intervention, and improves the system reliability. BRIEF DESCRIPTION OF DRAWINGS

[0035] In order to more clearly illustrate the technical solutions in the embodiments or examples of the present application, the drawings needed to be used in the embodiments or examples will be briefly introduced. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative effort based on these drawings.

[0036] Figure 1 A flowchart of the battery replacement cabinet temperature control method based on genetic algorithm optimization of PID parameters is shown.

[0037] Figure 2 A flowchart of judging whether the current cabinet internal environment temperature can meet the preset charging and discharging requirements of the battery is shown.

[0038] Figure 3 A flowchart of adjusting PID parameters in the battery replacement cabinet PID temperature control model by the improved genetic algorithm is shown.

[0039] Figure 4 A block diagram of the battery replacement cabinet temperature control system based on genetic algorithm optimization of PID parameters is shown. DETAILED DESCRIPTION

[0040] In order to more clearly illustrate the technical solutions in the embodiments or examples of the present application, the drawings needed to be used in the embodiments or examples will be briefly introduced. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative effort based on these drawings.

[0041] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, however, the present application can also be implemented in other ways different from those described herein, therefore, the scope of protection of the present application is not limited by the specific embodiments disclosed below.

[0042] Figure 1 A flowchart of the battery replacement cabinet temperature control method based on genetic algorithm optimization of PID parameters is shown.

[0043] As Figure 1 shown, the battery replacement cabinet temperature control method based on genetic algorithm optimization of PID parameters is provided in the first embodiment of the present application, including:

[0044] S102, a data-driven model is constructed by historical charging and discharging data of the battery in the battery replacement cabinet and multi-source temperature data, and it is judged whether the current cabinet internal environment temperature can meet the preset charging and discharging requirements of the battery.

[0045] S104, if the current temperature environment cannot meet the preset charging and discharging requirements of the battery, a PID parameter adjustment signal is generated;

[0046] S106, a battery swap cabinet PID temperature control model is established according to the relationship between temperature change and temperature control, the PID parameters are tuned by an improved genetic algorithm, the optimal PID parameter combination is obtained, and the temperature control signal output by the battery swap cabinet PID temperature control model is obtained using the optimal PID parameter combination;

[0047] S108, the temperature control signal is used to control the temperature of the battery swap cabinet, and the temperature of the battery swap cabinet is continuously monitored during the temperature control process, and the temperature abnormality event is prewarned and visualized.

[0048] It should be noted that through the multi-point high-precision temperature sensor in the cabinet, the temperature control system can realize centralized sampling of the temperature of the external environment, the cabinet environment, the battery and the charger, and real-time and intelligent processing of data, and can adjust the PTC heater or resistance wire heating synchronously with the temperature change to realize accurate control of the temperature in the cabinet. According to the historical temperature control tasks, the historical charging and discharging data of the battery swap cabinet and the monitored external environment temperature, cabinet environment temperature, battery and charger temperature during the historical charging and discharging process are obtained, and after normalization, noise reduction and other data preprocessing, normal charging and discharging data that meet the preset battery charging and discharging requirements are selected from the historical charging and discharging data, such as charging data that meet the 1C high-power charging requirements of lithium ion batteries; the selected normal charging and discharging data are matched with the multi-source temperature data monitored during the historical charging and discharging process to generate data samples, the data samples are clustered, the temperature corresponding to the low temperature environment is divided into intervals according to the climate information of the region where the battery swap cabinet is located, the temperature rise change of the battery charging and discharging in the cabinet in different low temperature environment intervals is obtained, the divided intervals are corrected, and different low temperature working condition labels are obtained. Construct an initial clustering center using the final temperature interval; use the Euclidean distance as the measurement function to divide the data samples into the nearest initial clustering center to generate clustering clusters, and continuously update the clustering clusters and clustering centers through iterative clustering. After satisfying the preset iteration times, the clustering results of the last clustering are read and output, and the low temperature working condition label is set for the data samples.

[0049] The data-driven model is constructed by training the stacked denoising autoencoder using the data sample with the low-temperature temperature condition label, a plurality of denoising autoencoders are connected end to end to extract features one by one in the pre-training process, the input data and the output data of each denoising autoencoder are of the same dimension, and the hidden layer data of each denoising autoencoder is used as the input data of the next denoising autoencoder. The input data sample is randomly zeroed, the features are extracted by the several denoising autoencoders connected end to end, the output of the last autoencoder is introduced into the extreme learning machine to determine whether the preset charging and discharging requirements of the battery are met, and the hidden layer node number, input data zeroing ratio, learning rate and batch size of the stacked denoising autoencoder are set after iterative training using the training data. The original stacked denoising autoencoder is improved by means of the high-dimensional feature classification and fast training capability of the extreme learning machine, the training time of the data-driven model is reduced, the detection accuracy of the model is ensured, and the robustness of the data-driven model is enhanced. When the classification performance of the data-driven model meets the performance standard, the trained data-driven model is output.

[0050] Figure 2 A flowchart for determining whether the current cabinet environment temperature can meet the preset charging and discharging requirements of the battery is shown.

[0051] According to the embodiment of the present application, whether the current cabinet environment temperature can meet the preset charging and discharging requirements of the battery is determined, specifically:

[0052] S202, the current multi-source temperature data and the battery charging and discharging data of the battery swap cabinet are collected and introduced into the data-driven model, the data-driven model is used for feature extraction, the temperature condition label corresponding to the temperature environment of the battery swap cabinet is determined, and the extracted features are used for data reconstruction;

[0053] S204, the multi-source temperature estimation value and the charging and discharging data estimation value of the battery in the normal charging and discharging state under the temperature condition label are obtained by the decoder according to the feature distribution of the input data, and the residual error of the multi-source temperature estimation value and the charging and discharging data estimation value and the input data is obtained by the extreme learning machine;

[0054] S206, the residual error corresponding to the multi-source temperature data and the battery charging and discharging data is analyzed, when the residual error is greater than the preset residual error threshold, it is determined that the current cabinet environment temperature cannot meet the preset charging and discharging requirements of the battery, and the PID parameter adjustment signal is generated according to the residual error corresponding to the multi-source temperature data.

[0055] It should be noted that the data-driven model constructed based on the stacked denoising autoencoder is trained using normal data samples that meet the preset charging and discharging requirements and temperature standards of the battery, and the model training aims to minimize the reconstruction data error. Therefore, the stacked denoising autoencoder generates estimated data samples close to the training data according to the latent space representation of the normal data samples, that is, the difference between the estimated data samples generated by the stacked denoising autoencoder and the normal data samples is small, and the difference with the abnormal samples is large. Therefore, the residual error between the obtained multi-source temperature estimation value and the charging and discharging data estimation value and the input data is evaluated using the extreme learning machine to realize temperature environment anomaly detection.

[0056] It should be noted that the existing PID control algorithm cannot be adjusted in real time, has poor self-adaptation ability, and cannot meet the control requirements of a high-precision temperature control system. The fuzzy PID controller has a slow response speed to sudden disturbance sources, which can easily lead the temperature control system to enter an unstable state. Therefore, an incremental PID algorithm is used to construct a battery swap cabinet PID temperature control model based on the self-learning and information processing capabilities of the deep belief network. The actual output result and the error value are used as the input signals of the battery swap cabinet PID temperature control model, and then the corresponding PID parameters are output after a multi-layer restricted Boltzmann machine calculation. The given temperature target value and the temperature measured value in the historical temperature adjustment of the battery swap cabinet under low temperature conditions are obtained, and a data set is constructed by combining the temperature difference between the given temperature target value and the temperature measured value with the corresponding control amount. The training set and the validation set are divided according to the preset proportion.

[0057] The deep belief network is stacked by multiple restricted Boltzmann machines (RBM). The battery swap cabinet PID temperature control model is constructed by the deep belief network. The restricted Boltzmann machines in the deep belief network are trained by the training set. Feature extraction and PID parameter calculation are performed in each layer of the restricted Boltzmann machine. The control amount generated according to the PID parameters is imported into the next layer of the restricted Boltzmann machine for layer-by-layer training. The weights of different layers of the restricted Boltzmann machine are updated after layer-by-layer iterative training by unsupervised learning. A deep nonlinear mapping between temperature change and temperature control is constructed. The constructed battery swap cabinet PID temperature control model is verified by the validation set. The weights of the model are adjusted by back propagation using the verification result. The goal is to minimize the control error. The control amount u(t) of the temperature control device is composed of three parts: the proportion K p , the integral K i and the differential K d , which is represented as:

[0058]

[0059] wherein e(t) is the error (the temperature difference between the given temperature target value and the temperature measured value).

[0060] Figure 3A flow chart of adjusting PID parameters in the battery swap cabinet PID temperature control model by the improved genetic algorithm is shown.

[0061] According to the embodiment of the present application, in the battery swap cabinet PID temperature control model, the PID parameters are adjusted by the improved genetic algorithm to obtain the optimal PID parameter combination, specifically:

[0062] In the training process of the battery swap cabinet PID temperature control model, the depth belief network PID controller parameters are optimized by the improved genetic algorithm, the PID parameters are genetically coded by chaotic mapping, and the chromosome population is initialized by elite reverse learning.

[0063] In S304, the fitness function is constructed according to the temperature control accuracy, temperature control speed and temperature control stability, the chromosome population position is updated by selection and crossover operation according to the fitness value, the current optimal solution is selected, and the current optimal solution is disturbed and updated by introducing Cauchy mutation in the iteration process.

[0064] In S306, the optimal solution is selected by comparing the fitness values before and after disturbance, the fitness of individuals is calculated by iteration, and the next generation population is generated by selection, crossover and Cauchy mutation operation of chromosomes. When the number of consecutive unchanged fitness values reaches the preset number of thresholds, the optimal PID parameter combination is obtained, and the existing parameters of the battery swap cabinet PID temperature control model are replaced.

[0065] It should be noted that in the training process of the battery swap cabinet PID temperature control model, the PID controller parameters in the restricted Boltzmann machine are trained by the genetic algorithm. The fuzzy PID controller has five fuzzy variables, including differential compensation value, proportional compensation value, integral compensation value, error and error change rate. In the process of defining the value range of the controller parameters and obtaining the PID parameters, the basic data of multiple indicators such as rise time, regulation time and overshoot are obtained. The fuzzy variables and corresponding fuzzy language values, index basic data and the like are genetically coded by Tent chaotic mapping, the chromosome population is initialized by elite reverse learning, the elite population is selected by calculating the fitness function value and arranging the fitness values from small to large, the reverse solution of the elite population is obtained by reverse learning and the fitness value is calculated, the fitness values before and after are compared, and the initial population is selected according to the comparison result. According to the temperature control accuracy, temperature control speed and temperature control stability, the fitness function is constructed, and the fitness function L is represented as:

[0066]

[0067] Where w1, w2 and w3 represent coefficients, e(t) represents error, dT i (t) represents the given temperature target value, and dT(t) represents the temperature change per unit time.

[0068] In the iteration process of genetic algorithm, Cauchy mutation is introduced to disturb and update the current optimal solution. Cauchy mutation has a longer mutation step length, which guides the population to jump out of the local optimal solution. After each iteration is completed, the fitness value of each population and the average fitness value of all populations are calculated. When the fitness value of each population is less than the average fitness value of all populations, Cauchy mutation is selected to disturb and update the current optimal solution. The mapping relationship between the PID control parameters and the temperature control performance is obtained by using the trained deep belief network. The optimal solution is obtained by optimizing the PID parameters through genetic algorithm, and the parameter setting task is executed to obtain the set control signal. The optimized PID parameter combination of the PID temperature control model of the battery replacement cabinet is better than the existing PID controller parameters, which can realize accurate control of the internal temperature of the battery replacement cabinet and avoid temperature fluctuations.

[0069] Preferably, in order to further enhance the global search ability, when the fitness value is continuously unchanged for a preset number of times, the current population range of the genetic algorithm is used as the initial population range of the swarm algorithm, the elite reverse learning is used to initialize the swarm, the swarm corresponding to the minimum fitness value is selected as the local optimal solution, and the swarm population position is updated through the two stages of chasing prey and attacking prey. The fitness value is recalculated and the current and global optimal solution positions are updated. The swarm position is updated by using the Levy flight strategy, the fitness of the updated swarm is obtained, and the smaller swarm position output PID parameter optimal solution is obtained by comparing the fitness. The Levy flight strategy is represented as:

[0070] g n =g b +a·Levy(λ)

[0071] Wherein g n represents the swarm updated by the Levy flight strategy, g b represents the swarm before the Levy flight strategy is updated, a represents a random step length, and Levy(λ) represents a random search path.

[0072] It should be noted that the temperature change value of the given temperature target value and the current temperature measured value is analyzed according to the PID parameter adjustment signal, and the differential, proportional and integral operations are performed on the battery replacement cabinet PID temperature control model to obtain the control amount to generate the temperature control signal and send it to the temperature rise equipment for battery replacement cabinet environment temperature control.

[0073] The historical temperature rise data of the temperature adjusting equipment of the battery swap cabinet is acquired, the temperature change value in a unit time is acquired based on the historical temperature rise data, the threshold information of the temperature control monitoring task is established according to the temperature change value in the unit time, whether the temperature change value of the cabinet environment temperature in the unit time is less than the threshold information is judged, and if less, temperature adjusting equipment abnormal information is generated; the temperature data critical value before and after the battery thermal runaway is extracted through the battery thermal runaway instance, the cumulative probability distribution function of the temperature data critical value is constructed, and the thermal runaway occurrence probability corresponding to the current cabinet environment temperature is analyzed according to the cumulative probability distribution function; if the thermal runaway occurrence probability is greater than a preset probability threshold, battery thermal runaway abnormal information is generated, the temperature adjusting equipment abnormal information and the battery thermal runaway abnormal information are prewarned, and the prewarning is visually displayed. The temperature control system and the aerosol safety fire extinguishing system realize linkage, when the battery thermal runaway abnormal information appears, the aerosol safety fire extinguishing system is activated to be ready for action, the wide voltage and large current safe charging of the battery swap battery can be ensured, and the 1C high-power continuous charging demand of the lithium ion battery can be met.

[0074] Figure 4 A block diagram of a battery swap cabinet temperature control system based on genetic algorithm optimization of PID parameters is shown.

[0075] The second embodiment of the application provides a battery swap cabinet temperature control system 4 based on genetic algorithm optimization of PID parameters, which comprises a temperature data acquisition unit 401, a temperature monitoring unit 402, a temperature control unit 403 and a temperature abnormality warning unit 404.

[0076] The temperature data acquisition unit is responsible for acquiring the temperature data and charging and discharging data of the external environment, the warehouse environment, the cabinet environment, the battery and the charger, and pre-processing the multi-source temperature data and charging and discharging data.

[0077] The temperature monitoring unit is responsible for constructing a data-driven model, reading normal data from historical multi-source temperature data and historical charging and discharging data for model training, judging whether the current cabinet environment temperature meets the preset charging and discharging demand of the battery, and generating a PID parameter adjustment signal if not.

[0078] The temperature control unit establishes a battery swap cabinet temperature control model according to the relationship between temperature change and control input, adjusts the PID parameters through an improved genetic algorithm, acquires an optimal PID parameter combination, and generates a temperature control signal for temperature control according to the optimal PID parameter combination.

[0079] The temperature abnormality warning unit is responsible for continuously monitoring the temperature in the temperature control process of the battery swap cabinet, capturing temperature abnormality events for prewarning and realizing visual display.

[0080] The third embodiment of the present application provides a computer readable storage medium, wherein a PID parameter optimization method based on a genetic algorithm for temperature control of a battery swap cabinet is stored in the computer readable storage medium.

[0081] In several embodiments provided in the present application, it should be understood that the disclosed method can be implemented in other ways. The above-described embodiments are only illustrative, for example, the division of the units is only a logical function division, and actual implementation can have another division mode, for example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed components can be indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0082] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by program instruction related hardware, and the foregoing program can be stored in a computer readable storage medium, and the program is executed to execute the steps of the above-mentioned method embodiments; and the foregoing storage medium includes a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk and various storage program codes.

[0083] Alternatively, the integrated units of the present application can be stored in a computer readable storage medium if they are realized in the form of software function units and sold or used as independent products. Based on this understanding, the technical solutions of the embodiments of the present application can be embodied in the form of software products, which are stored in a storage medium and include a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the embodiments of the present application. The foregoing storage medium includes a mobile storage device, a ROM, a RAM, a magnetic disk or an optical disk and various storage program codes.

[0084] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered within the protection scope of the present application.

Claims

1. A battery replacement cabinet temperature control method based on genetic algorithm optimization of PID parameters, characterized by, The method comprises the following steps: A data-driven model is constructed based on historical charging and discharging data and multi-source temperature data of the battery in the battery swap cabinet to determine whether the current temperature environment in the cabinet can meet the preset charging and discharging requirements of the battery; If the current temperature environment cannot meet the preset charging and discharging requirements of the battery, a PID parameter adjustment signal is generated; A PID temperature control model of the battery swap cabinet is established based on the relationship between temperature change and temperature control, and an improved genetic algorithm is used to set the PID parameters to obtain an optimal PID parameter combination, and a temperature control signal output by the PID temperature control model of the battery swap cabinet is obtained using the optimal PID parameter combination; The temperature control signal is used to control the temperature of the battery swap cabinet, and the temperature of the battery swap cabinet is continuously monitored during the temperature control process, and temperature abnormal events are warned and visually displayed; The PID temperature control model of the battery swap cabinet is established based on the relationship between temperature change and temperature control, specifically as follows: The given temperature target value and the measured temperature value in the historical temperature adjustment of the battery swap cabinet under low-temperature conditions are obtained, a data set is constructed by combining the temperature difference between the given temperature target value and the measured temperature value and the corresponding control amount, and the training set and the validation set are divided according to a preset proportion; A deep belief network is used to construct the PID temperature control model of the battery swap cabinet, the restricted Boltzmann machine in the deep belief network is trained through the training set, the PID parameters are calculated in each layer of the restricted Boltzmann machine, the control amount is generated according to the PID parameters and is imported into the next layer of the restricted Boltzmann machine for layer-by-layer training; After the layer-by-layer iterative training, the weights of different layers of the restricted Boltzmann machine are updated, a deep nonlinear mapping between temperature change and temperature control is constructed, the constructed PID temperature control model of the battery swap cabinet is verified through the validation set, and the weights of the model are adjusted through the verification result; In the PID temperature control model of the battery swap cabinet, the improved genetic algorithm is used to set the PID parameters to obtain an optimal PID parameter combination, specifically as follows: In the training process of the PID temperature control model of the battery swap cabinet, the improved genetic algorithm is used to optimize the PID controller parameters of the deep belief network, the PID parameters are genetically coded through chaotic mapping, and the chromosome population is initialized by elite reverse learning; An adaptability function is constructed based on temperature control accuracy, temperature control speed and temperature control stability, the chromosome population position is updated by selection and crossover operation according to the adaptability value, the current optimal solution is selected, and the current optimal solution is disturbed and updated by introducing Cauchy mutation in the iteration process; The adaptability values before and after disturbance are compared, the optimal solution is selected through the comparison result, the adaptability of individuals is continuously calculated, and the next generation population is generated by selection, crossover and Cauchy mutation operation of the chromosome; 2. The battery swap cabinet temperature control method based on genetic algorithm optimization of PID parameters according to claim 1, characterized in that, When the adaptability value is unchanged for a continuous number of times reaching a preset number of times threshold, the optimal PID parameter combination is obtained, and the existing parameters of the PID temperature control model of the battery swap cabinet are replaced. A data-driven model is constructed based on historical charging and discharging data and multi-source temperature data of the battery in the battery swap cabinet, specifically as follows: The historical charging and discharging data of the battery swap cabinet and the monitored external environment temperature, the cabinet internal environment temperature, the battery and the charging and discharging device temperature in the historical charging and discharging process are obtained, and after data preprocessing, the normal charging and discharging data meeting the preset battery charging and discharging requirement are screened from the historical charging and discharging data; The screened normal charging and discharging data are matched with the multi-source temperature data monitored in the historical charging and discharging process to generate data samples, the data samples are clustered, the temperature interval is divided according to the climate information of the region where the battery swap cabinet is located and the temperature rise change of the battery charging and discharging in the cabinet, and the initial clustering center is constructed using the temperature interval; The data samples are divided into the nearest initial clustering center to generate a clustering cluster, and the clustering cluster and the clustering center are continuously updated through iterative clustering, and the clustering result of the last clustering is read and output after the preset iteration number is met, and the temperature working condition label is set for the data samples; The data samples with the temperature working condition label are used to train the stacked denoising autoencoder to construct a data-driven model, the input data samples are randomly zeroed, the features are extracted through a plurality of denoising autoencoders connected end to end, the output of the last autoencoder is introduced into the extreme learning machine to judge whether the preset charging and discharging requirement of the battery is met, and when the classification performance of the data-driven model meets the performance standard, the trained data-driven model is output.

3. The battery swap cabinet temperature control method based on genetic algorithm optimization of PID parameters according to claim 1, characterized in that, It is judged whether the current cabinet internal environment temperature can meet the preset charging and discharging requirement of the battery, specifically: The current multi-source temperature data and the battery charging and discharging data of the battery swap cabinet are collected and introduced into the data-driven model, the features are extracted using the data-driven model, the temperature working condition label corresponding to the temperature environment of the battery swap cabinet is judged, and the extracted features are used for data reconstruction; The multi-source temperature estimation value and the charging and discharging data estimation value of the battery in the normal charging and discharging state under the temperature working condition label are obtained according to the feature distribution of the input data through the decoder, and the residual error of the multi-source temperature estimation value and the charging and discharging data estimation value and the input data is obtained using the extreme learning machine respectively; The residual error corresponding to the multi-source temperature data and the battery charging and discharging data is analyzed, when the residual error is greater than the preset residual error threshold, it is determined that the current cabinet internal environment temperature cannot meet the preset charging and discharging requirement of the battery, and the PID parameter adjustment signal is generated according to the residual error corresponding to the multi-source temperature data.

4. The battery swap cabinet temperature control method based on genetic algorithm optimization of PID parameters according to claim 1, characterized in that, The temperature control signal output by the PID temperature control model of the battery swap cabinet is obtained using the optimal PID parameter combination for temperature control, specifically: The temperature change value of the given temperature target value and the current temperature measured value is analyzed according to the PID parameter adjustment signal, the differential, proportional and integral operations are performed on the battery swap cabinet PID temperature control model, the control quantity is obtained to generate the temperature control signal and send it to the temperature rise device for temperature control of the battery swap cabinet environment.

5. The battery swap cabinet temperature control method based on genetic algorithm optimization of PID parameters according to claim 1, characterized in that, During the temperature control process, the temperature of the battery swap cabinet is continuously monitored, the temperature abnormal event is warned and visualized, specifically: The historical temperature rise data of the temperature adjusting device of the battery swap cabinet is obtained, the temperature change value per unit time is obtained based on the historical temperature rise data, the threshold information of the temperature control monitoring task is established according to the temperature change value per unit time, and it is judged whether the temperature change value per unit time of the cabinet internal environment temperature is less than the threshold information, if yes, the temperature adjusting device abnormal information is generated. A temperature data critical value before and after the battery thermal runaway is extracted through a battery thermal runaway example, a cumulative probability distribution function of the temperature data critical value is constructed, and a thermal runaway occurrence probability corresponding to a current cabinet environment temperature is analyzed according to the cumulative probability distribution function; If the thermal runaway occurrence probability is greater than a preset probability threshold, battery thermal runaway abnormal information is generated, the temperature adjusting equipment abnormal information and the battery thermal runaway abnormal information are prewarned, and the prewarning is visually displayed.

6. A battery replacement cabinet temperature control system based on genetic algorithm optimization of PID parameters, characterized in that, The battery replacement cabinet temperature control method based on the genetic algorithm optimized PID parameter is realized, the system includes a temperature data acquisition unit, a temperature monitoring unit, a temperature control unit, and a temperature abnormality early warning unit; The temperature data acquisition unit is responsible for collecting temperature data and charging and discharging data of external environment, warehouse environment, cabinet environment, battery and charger, and pre-processing the multi-source temperature data and charging and discharging data; The temperature monitoring unit is responsible for constructing a data-driven model, reading normal data from historical multi-source temperature data and historical charging and discharging data for model training, judging whether the current cabinet environment temperature meets the preset charging and discharging requirements of the battery, and generating a PID parameter adjustment signal if not; The temperature control unit establishes a battery replacement cabinet temperature control model according to the relationship between temperature change and control input, adjusts the PID parameter through the improved genetic algorithm, obtains the optimal PID parameter combination, generates a temperature control signal according to the optimal PID parameter combination, and controls the temperature; The temperature abnormality early warning unit is responsible for continuously monitoring the temperature in the battery replacement cabinet temperature control process, capturing temperature abnormality events for prewarning and realizing visual display.

Images