A method, apparatus, and electronic device for predicting the state of charge of retired batteries.

Abstract

This invention relates to a method, apparatus, and electronic device for predicting the state of charge (SOC) of retired batteries. The method includes: acquiring data from several retired batteries at different SOC states; performing curve fitting using the data; obtaining feature parameters based on the curve fitting results; constructing a neural network model; training the neural network model using the feature parameters to obtain a fully trained neural network model; acquiring real-time feature parameters of the retired battery to be predicted; and obtaining the SOC of the retired battery to be predicted based on the real-time feature parameters and the fully trained neural network model. The SOC prediction method for retired batteries disclosed in this invention improves the accuracy of SOC prediction for retired batteries.

Description

Technical Field

[0001] This invention relates to the field of retired battery technology, and in particular to a method, apparatus, electronic device, and computer-readable storage medium for predicting the state of charge of retired batteries. Background Technology

[0002] Batteries, as energy storage power sources, have been widely used in fields such as communications, power systems, and transportation. However, different fields have different requirements for batteries, resulting in a large number of retired batteries. If retired batteries cannot be recycled for a second or even multiple times, it will cause a lot of energy waste. The state of charge of retired batteries is the basis for recycling and is used to determine whether they are suitable for recycling.

[0003] It is impractical to perform individual testing on the state of charge of retired batteries due to the enormous workload. Therefore, the commonly used methods for predicting the state of charge of retired batteries include the ampere-hour integration method, but the ampere-hour integration method is easily affected by the accuracy of current measurement and has low accuracy. Summary of the Invention

[0004] In view of this, it is necessary to provide a method, apparatus, electronic device and computer-readable storage medium for predicting the state of charge of retired batteries, so as to solve the problem of inaccurate prediction of the state of charge of retired batteries in the prior art.

[0005] To address the above problems, this invention provides a method for predicting the state of charge of retired batteries, comprising:

[0006] Data on several retired batteries under different states of charge are obtained, and curve fitting is performed using the data. Characteristic parameters are obtained based on the curve fitting results.

[0007] Construct a neural network model, and train the neural network model according to the feature parameters to obtain a fully trained neural network model;

[0008] The real-time feature parameters of the battery to be decommissioned are obtained, and the state of charge of the battery to be decommissioned is obtained based on the real-time feature parameters and the fully trained neural network model.

[0009] Furthermore, curve fitting is performed using the data, including:

[0010] The data was used to perform Nyquist curve fitting.

[0011] Furthermore, characteristic parameters are obtained based on the curve fitting results, including:

[0012] Obtain the initial feature parameters of the curve fitting result, and then filter the initial feature parameters to obtain the filtered feature parameters.

[0013] Further, the initial feature parameters are filtered, including:

[0014] Obtain the corresponding state of charge (SOC) value of the retired battery, and use the Pearson correlation coefficient to calculate the correlation between the initial feature parameters and the corresponding SOC value of the retired battery.

[0015] When the absolute value of the correlation is less than a set threshold, the initial feature parameters are removed.

[0016] Furthermore, a neural network model is constructed, including:

[0017] Construct a DBN-DNN neural network model.

[0018] Furthermore, the DBN-DNN neural network model includes one input layer, three hidden layers, and one output layer. The DBN-DNN neural network model is composed of three restricted Boltzmann machine units. Each restricted Boltzmann machine unit has two layers: an upper hidden layer and a lower explicit layer. The hidden layer (output layer) of the previous restricted Boltzmann machine unit serves as the explicit layer (input layer) of the next restricted Boltzmann machine unit.

[0019] Further, training the neural network model based on the feature parameters includes:

[0020] The neural network model is trained by using the aforementioned feature parameters as input to the input layer and the corresponding state of charge value of the decommissioned battery as output to the output layer.

[0021] The present invention also provides a device for predicting the state of charge of retired batteries, including a data acquisition module, a model training module and a state prediction module;

[0022] The data acquisition module is used to acquire data of several retired batteries under different states of charge, perform curve fitting using the data, and obtain characteristic parameters based on the curve fitting results.

[0023] The model training module is used to construct a neural network model and train the neural network model according to the feature parameters to obtain a fully trained neural network model.

[0024] The state prediction module is used to obtain the real-time feature parameters of the battery to be predicted for retirement, and to obtain the state of charge of the battery to be predicted for retirement based on the real-time feature parameters and the fully trained neural network model.

[0025] The present invention also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, it implements the method for predicting the state of charge of decommissioned batteries as described in any of the above technical solutions.

[0026] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the method for predicting the state of charge of decommissioned batteries as described in any of the above technical solutions.

[0027] The beneficial effects of the above embodiments are as follows: The method for predicting the state of charge of retired batteries provided by the present invention performs curve fitting on the data of retired batteries under different states of charge, obtains feature parameters based on the curve fitting results, and uses the feature parameters to train the constructed neural network model to obtain a fully trained neural network model. The accuracy of predicting the state of charge of retired batteries is improved by using the fully trained neural network model to predict the state of charge of retired batteries. Attached Figure Description

[0028] Figure 1 A flowchart illustrating an embodiment of the method for predicting the state of charge of decommissioned batteries provided by the present invention;

[0029] Figure 2 This is a schematic diagram of the DBN-DNN neural network model provided in an embodiment of the present invention;

[0030] Figure 3 This is a structural block diagram of an embodiment of the retired battery state of charge prediction device provided by the present invention;

[0031] Figure 4 A structural block diagram of an embodiment of the electronic device provided by the present invention. Detailed Implementation

[0032] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0033] This invention provides a method, apparatus, electronic device, and computer-readable storage medium for predicting the state of charge of retired batteries, which are described in detail below.

[0034] This invention provides a method for predicting the state of charge of decommissioned batteries, the flowchart of which is shown below. Figure 1 As shown, the method for predicting the state of charge of decommissioned batteries includes:

[0035] Step S101: Obtain data of several retired batteries under different states of charge, use the data to perform curve fitting, and obtain characteristic parameters based on the curve fitting results;

[0036] Step S102: Construct a neural network model and train the neural network model according to the feature parameters to obtain a fully trained neural network model;

[0037] Step S103: Obtain the real-time feature parameters of the battery to be predicted for retirement, and obtain the state of charge of the battery to be predicted for retirement based on the real-time feature parameters and the fully trained neural network model.

[0038] In one specific embodiment, step S101 includes selecting a number of retired batteries, adjusting the retired batteries to different SOC states at 0.05C to 1C, letting them stand for a sufficient time, measuring the open-circuit voltage U of the retired batteries, measuring the AC impedance of the retired batteries at different SOCs using an impedance meter at a temperature of (25±2)℃ and a frequency of 0.05 to 100KHz, importing the measured impedance spectrum data into impedance spectrum analysis software, selecting a suitable equivalent circuit model for Nyquist curve fitting, and using the values ​​of each component in the fitted equivalent circuit and the open-circuit voltage U as initial characteristic parameters.

[0039] As a preferred embodiment, curve fitting using the data includes:

[0040] The data was used to perform Nyquist curve fitting.

[0041] As a preferred embodiment, the characteristic parameters are obtained based on the curve fitting results, including:

[0042] Obtain the initial feature parameters of the curve fitting result, and then filter the initial feature parameters to obtain the filtered feature parameters.

[0043] It should be noted that filtering the initial feature parameters can greatly reduce the workload, and the filtered feature parameters are more accurate, making the prediction of the state of charge of retired batteries more accurate.

[0044] As a preferred embodiment, the initial feature parameters are screened, including:

[0045] Obtain the corresponding state of charge (SOC) value of the retired battery, and use the Pearson correlation coefficient to calculate the correlation between the initial feature parameters and the corresponding SOC value of the retired battery.

[0046] When the absolute value of the correlation is less than a set threshold, the initial feature parameters are removed.

[0047] In one specific embodiment, the formula for calculating the Pearson correlation coefficient is as follows: Where Cov(X,Y) is the covariance between the initial feature parameter X and the corresponding state of charge value Y of the decommissioned battery, Var[X] is the variance of X, Var[Y] is the variance of Y, r is the correlation value, and the threshold setting is defined according to the specific situation, usually set to 98%.

[0048] As a preferred embodiment, constructing a neural network model includes:

[0049] Construct a DBN-DNN neural network model.

[0050] A schematic diagram of the DBN-DNN neural network model in a specific embodiment is shown below. Figure 2 As shown, Figure 2 As can be seen, the DBN-DNN neural network model includes one input layer, three hidden layers, and one output layer, which includes three restricted Boltzmann machine units RBM1, RBM2, and RBM3, forming a DBN structure. The output of the DBN is used as the input of the BP neural network to form the DBN-DNN neural network model.

[0051] In a preferred embodiment, the DBN-DNN neural network model includes one input layer, three hidden layers, and one output layer. The DBN-DNN neural network model is composed of three restricted Boltzmann machine units. Each restricted Boltzmann machine unit has two layers: an upper hidden layer and a lower explicit layer. The hidden layer (output layer) of the previous restricted Boltzmann machine unit serves as the explicit layer (input layer) of the next restricted Boltzmann machine unit.

[0052] It should be noted that the DBN-DNN neural network model constructed in this invention is a DNN structure built with a deep belief network. By greedily pre-training restricted Boltzmann machine units layer by layer, the training efficiency is greatly improved, and the problem of local optima is improved. The solution process is also relatively simple and easy to apply.

[0053] In a preferred embodiment, training the neural network model based on the feature parameters includes:

[0054] The neural network model is trained by using the aforementioned feature parameters as input to the input layer and the corresponding state of charge value of the decommissioned battery as output to the output layer.

[0055] This invention also provides a device for predicting the state of charge of retired batteries, the structural block diagram of which is shown below. Figure 3 As shown, the retired battery state of charge prediction device includes a data acquisition module 301, a model training module 302, and a state prediction module 303.

[0056] The data acquisition module 301 is used to acquire data of several retired batteries under different states of charge, perform curve fitting using the data, and obtain characteristic parameters based on the curve fitting results.

[0057] The model training module 302 is used to construct a neural network model and train the neural network model according to the feature parameters to obtain a fully trained neural network model.

[0058] The state prediction module 303 is used to obtain the real-time feature parameters of the battery to be predicted for retirement, and to obtain the state of charge of the battery to be predicted for retirement based on the real-time feature parameters and the fully trained neural network model.

[0059] like Figure 4 As shown in the above-described method for predicting the state of charge of decommissioned batteries, this invention also provides an electronic device, which can be a mobile terminal, desktop computer, laptop, handheld computer, server, or other computing device. The electronic device includes a processor 403, a display 402, and a memory 401.

[0060] In some embodiments, memory 401 may be an internal storage unit of a computer device, such as a hard disk or memory. In other embodiments, memory 401 may be an external storage device of a computer device, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc. Furthermore, memory 401 may include both internal and external storage units of the computer device. Memory 401 is used to store application software and various types of data installed on the computer device, such as program code for installing the computer device. Memory 401 can also be used to temporarily store data that has been output or will be output. In one embodiment, memory 401 stores a retired battery state of charge prediction program 404, which can be executed by processor 403 to implement the retired battery state of charge prediction method of various embodiments of the present invention.

[0061] In some embodiments, processor 403 may be a central processing unit (CPU), microprocessor or other data processing chip, used to run program code stored in memory 401 or process data, such as executing a retired battery state of charge prediction program.

[0062] In some embodiments, display 402 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. Display 402 is used to display information from the computer device and to display a visual user interface. Components 401-403 of the computer device communicate with each other via a system bus.

[0063] In one embodiment, when processor 403 executes retired battery state of charge prediction program 404 in memory 401, the following steps are performed:

[0064] Data on several retired batteries under different states of charge are obtained, and curve fitting is performed using the data. Characteristic parameters are obtained based on the curve fitting results.

[0065] Construct a neural network model, and train the neural network model according to the feature parameters to obtain a fully trained neural network model;

[0066] The real-time feature parameters of the battery to be decommissioned are obtained, and the state of charge of the battery to be decommissioned is obtained based on the real-time feature parameters and the fully trained neural network model.

[0067] This embodiment also provides a computer-readable storage medium storing a retired battery state of charge prediction program, which, when executed by a processor, performs the following steps:

[0068] Data on several retired batteries under different states of charge are obtained, and curve fitting is performed using the data. Characteristic parameters are obtained based on the curve fitting results.

[0069] Construct a neural network model, and train the neural network model according to the feature parameters to obtain a fully trained neural network model;

[0070] The real-time feature parameters of the battery to be decommissioned are obtained, and the state of charge of the battery to be decommissioned is obtained based on the real-time feature parameters and the fully trained neural network model.

[0071] This invention provides a method, apparatus, electronic device, and computer-readable storage medium for predicting the state of charge (SOC) of retired batteries. By performing curve fitting on data of retired batteries under different SOC states, characteristic parameters are obtained based on the curve fitting results. The characteristic parameters are then used to train a constructed neural network model to obtain a fully trained neural network model. The accuracy of predicting the SOC of retired batteries is improved by using the fully trained neural network model.

[0072] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0073] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for predicting the state of charge of retired batteries, characterized in that, include: Data on several retired batteries under different states of charge are obtained, and curve fitting is performed using the data. Characteristic parameters are obtained based on the curve fitting results. A DBN-DNN neural network model is constructed, and the neural network model is trained according to the feature parameters to obtain a fully trained neural network model. The DBN-DNN neural network model includes one input layer, three hidden layers, and one output layer. The DBN-DNN neural network model is composed of three restricted Boltzmann machine units. Each restricted Boltzmann machine unit includes two layers: the upper layer is the hidden layer and the lower layer is the explicit layer. The hidden layer (output layer) of the previous restricted Boltzmann machine unit serves as the explicit layer (input layer) of the next restricted Boltzmann machine unit. The real-time feature parameters of the battery to be decommissioned are obtained, and the state of charge of the battery to be decommissioned is obtained based on the real-time feature parameters and the fully trained neural network model.

2. The method for predicting the state of charge of decommissioned batteries according to claim 1, characterized in that, Curve fitting using the data includes: The data was used to perform Nyquist curve fitting.

3. The method for predicting the state of charge of decommissioned batteries according to claim 1, characterized in that, The characteristic parameters are obtained based on the curve fitting results, including: Obtain the initial feature parameters of the curve fitting result, and then filter the initial feature parameters to obtain the filtered feature parameters.

4. The method for predicting the state of charge of decommissioned batteries according to claim 3, characterized in that, The initial feature parameters are filtered, including: Obtain the corresponding state of charge (SOC) value of the retired battery, and use the Pearson correlation coefficient to calculate the correlation between the initial feature parameters and the corresponding SOC value of the retired battery. When the absolute value of the correlation is less than a set threshold, the initial feature parameters are removed.

5. The method for predicting the state of charge of decommissioned batteries according to claim 4, characterized in that, Training the neural network model based on the feature parameters includes: The neural network model is trained by using the aforementioned feature parameters as input to the input layer and the corresponding state of charge value of the decommissioned battery as output to the output layer.

6. A device for predicting the state of charge of a retired battery, characterized in that, It includes a data acquisition module, a model training module, and a state prediction module; The data acquisition module is used to acquire data of several retired batteries under different states of charge, perform curve fitting using the data, and obtain characteristic parameters based on the curve fitting results. The model training module is used to construct a DBN-DNN neural network model and train the neural network model according to the feature parameters to obtain a fully trained neural network model. The DBN-DNN neural network model includes one input layer, three hidden layers, and one output layer. The DBN-DNN neural network model is composed of three restricted Boltzmann machine units. Each restricted Boltzmann machine unit includes two layers: an upper hidden layer and a lower explicit layer. The hidden layer (output layer) of the previous restricted Boltzmann machine unit serves as the explicit layer (input layer) of the next restricted Boltzmann machine unit. The state prediction module is used to obtain the real-time feature parameters of the battery to be predicted for retirement, and to obtain the state of charge of the battery to be predicted for retirement based on the real-time feature parameters and the fully trained neural network model.

7. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, it implements the method for predicting the state of charge of decommissioned batteries as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, It stores a computer program, which, when executed by a processor, implements the method for predicting the state of charge of decommissioned batteries as described in any one of claims 1-5.

Images