Battery state of charge estimation method and device

Abstract

The application provides a battery state of charge estimation method and device. It relates to the technical field of computers. The last acquisition data, the last estimation data and the current acquisition data are obtained, the acquisition data includes working condition data and environment data, the acquisition time interval of the last acquisition data and the current acquisition data is not more than a set threshold, and the last acquisition data, the last estimation data and the current acquisition data are input into a pre-trained neural network model to obtain the current estimation data. In this way, the battery dynamic characteristics can be simulated to estimate the battery state of charge, and the method is suitable for various batteries. In the case of a large amount of reference data training, a SOC estimation model suitable for different models and different batches can be established.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, in particular to a battery state of charge estimation method and device. BACKGROUND

[0002] The SOC accuracy of a lithium battery has a great influence on an energy storage power station. How much electricity can be discharged or how much electricity can be charged at the current state of the energy storage power station is determined based on the SOC. The SOC value of the battery is derived from a battery management system (BMS). The traditional battery management system basically calculates the SOC of the lithium battery based on the ampere-hour integration method. This method has a serious defect of cumulative error. The cumulative error is very large after a long time of charging and discharging, so correction is needed. The traditional correction method is based on the batch and model of the battery and adjusts through fixed parameters. With the development of energy storage, the lithium battery technology is also constantly updated and replaced. This correction method cannot adapt to the rapidly developing lithium battery technology, often leading to large deviations in the calculation results. SUMMARY

[0003] The purpose of the embodiments of the present application is to provide a battery state of charge estimation method and device to solve the technical problem of large estimation deviation in the prior art.

[0004] In a first aspect, a battery state of charge estimation method is provided. The method includes:

[0005] acquiring last time collection data, last time estimation data, and current time collection data, the collection data including working condition data and environment data, wherein the collection time interval of the last time collection data and the current time collection data is not more than a set threshold;

[0006] inputting the last time collection data, the last time estimation data, and the current time collection data into a pre-trained neural network model to obtain current time estimation data.

[0007] In some optional implementations, the working condition data includes voltage and current, and the environment data includes temperature.

[0008] In some optional implementations, the method further includes:

[0009] acquiring the cycle number at the current time calculation, the health state at the current time calculation, and the time interval between the last time collection and the current time collection;

[0010] The inputting the last time collection data, the last time estimation data, and the current time collection data into a pre-trained neural network model to obtain current time estimation data includes:

[0011] The last time collection data, last time estimated data, this time collection data, the number of cycles at this time calculation, the health status at this time calculation and the time interval between the last time collection and the present time collection are input into the pre-trained neural network model to obtain the present time estimated data.

[0012] In some optional implementations, the neural network model comprises an input layer, a hidden layer and an output layer; the activation function of the hidden layer is a monotonous differentiable Sigmoid function, and the activation function of the output layer is a linear purelin function.

[0013] In some optional implementations, the hidden layer is determined based on the following formula:

[0014]

[0015] wherein, P j is the output of the hidden layer, f is the activation function of the hidden layer, X i is the input vector, θ j is the parameter, W ij is the weight;

[0016] The output layer is determined based on the following formula:

[0017]

[0018] wherein, Y is the output of the output layer, W j , P j , and g is the activation function of the output layer.

[0019] In some optional implementations, the method further comprises:

[0020] determining a training sample; wherein the training sample is calculated from data collected in a preset collection period in a period with stable and constant power and small current change during the energy storage discharge or charge process;

[0021] training the initial neural network model based on the training sample to obtain the pre-trained neural network model.

[0022] In some optional implementations, the method further comprises:

[0023] The determination of the training sample comprises:

[0024] determining the true value state of charge ΔSOC in the training sample based on the following formula:

[0025] wherein, ΔSOC is determined as:

[0026] ΔSOC = S - IT

[0027] Wherein, S is the last calculated SOC value, IT is the discharged power in this time period.

[0028] In a second aspect, a battery state of charge estimation device is provided.

[0029] The acquisition module is configured to acquire last acquisition data, last estimation data, and current acquisition data, wherein the acquisition data includes working condition data and environmental data, and the acquisition time interval between the last acquisition data and the current acquisition data is not more than a set threshold.

[0030] The estimation module is configured to input the last acquisition data, the last estimation data, and the current acquisition data into a pre-trained neural network model to obtain current estimation data.

[0031] In a third aspect, an electronic device is provided, which includes a processor and a memory.

[0032] The memory stores a computer program, and the computer program, when executed by the processor, performs the method of any one of the preceding first aspect.

[0033] In a fourth aspect, a computer readable storage medium is provided, and the computer readable storage medium stores a computer program, and the computer program, when executed by a processor, performs the method of any one of the preceding first aspect.

[0034] The embodiments of the present application provide a battery state of charge estimation method and device. By acquiring last acquisition data, last estimation data, and current acquisition data, the acquisition data includes working condition data and environmental data, and the acquisition time interval between the last acquisition data and the current acquisition data is not more than a set threshold, and by inputting the last acquisition data, the last estimation data, and the current acquisition data into a pre-trained neural network model to obtain current estimation data, the battery dynamic characteristics can be simulated to estimate the battery state of charge, and the method is suitable for various batteries. In the case of a large amount of reference data training, an SOC estimation model suitable for different models and different batches can be established. BRIEF DESCRIPTION OF DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments of the present application. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope, and for those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0036] Figure 1is an example electronic device for implementing a battery state of charge estimation method according to an embodiment of the present application;

[0037] Figure 2 is a flowchart of a battery state of charge estimation method according to an embodiment of the present application;

[0038] Figure 3 is an effect schematic diagram of a battery state of charge estimation method according to the prior art;

[0039] Figure 4 is an effect schematic diagram of a battery state of charge estimation method according to an embodiment of the present application;

[0040] Figure 5 is a schematic diagram of a battery state of charge estimation device according to an embodiment of the present application. DETAILED DESCRIPTION

[0041] In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. The components of the embodiments of the present application described and shown in the drawings herein can be arranged and designed in various different configurations.

[0042] First, referring to Figure 1 , an example electronic device 100 for implementing a battery state of charge estimation method according to an embodiment of the present application will be described.

[0043] As shown in Figure 1 , the electronic device 100 includes one or more processors 102, one or more memories 104, an input device 106, an output device 108, and an image acquisition device 110, which are interconnected through a bus system 112 and / or other forms of connection mechanism (not shown). It should be noted that Figure 1 The components and structures of the electronic device 100 shown are only exemplary and are not limiting, and the electronic device can also have other components and structures as needed.

[0044] The processor 102 can be a central processing unit (CPU), a graphics processing unit (GPU), or other forms of processing units with data processing and / or instruction execution capabilities, and can control other components in the electronic device 100 to perform desired functions.

[0045] The memory 104 can include one or more computer program products that can include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory, for example, can include random access memory (RAM), cache memory, and / or the like. The non-volatile memory, for example, can include read only memory (ROM), hard disk, flash memory, and / or the like. One or more computer program instructions can be stored on the computer-readable storage media, and the processor 102 can execute the program instructions to implement the client functions (implemented by the processor) in the embodiments of the application described below and / or other desired functions. Various application programs and various data, such as various data used and / or generated by the application programs, and the like, can also be stored in the computer-readable storage media.

[0046] The input device 106 can be a device used by a user to input instructions, and can include one or more of a keyboard, a mouse, a microphone, a touch screen, and the like.

[0047] The output device 108 can output various information (e.g., images or sounds) to the outside (e.g., a user), and can include one or more of a display, a speaker, and the like.

[0048] According to an embodiment of the present application, an embodiment of a battery state of charge estimation method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described herein can be executed in an order different from that shown herein.

[0049] Figure 2 is a flowchart of a battery state of charge estimation method according to an embodiment of the present application, as shown in Figure 2 The method comprises the following steps:

[0050] S210, acquiring last acquisition data, last estimation data and this acquisition data, the acquisition data including working condition data and environment data, wherein the acquisition time interval of the last acquisition data and the this acquisition data is not more than a set threshold;

[0051] The working condition data includes voltage and current, and the environment data includes temperature.

[0052] S220, inputting the last acquisition data, the last estimation data and the this acquisition data into a pre-trained neural network model to obtain this estimation data.

[0053] As shown in Figure 3 The last result is fed back to the model calculation to optimize the model and improve the result accuracy through the backward feedback mechanism.

[0054] For example, such as Figure 4 As shown, the neural network model can include an input layer, a hidden layer, and an output layer; the activation function of the hidden layer can be a monotonically differentiable sigmoid function, and the activation function of the output layer can be a linear purelin function.

[0055] The input layer is the input vector X(m), specifically represented as:

[0056] X(m) = (V1,V2,I1,I2,T1,T2,S,L,H,T), where V1 represents the voltage in the previous calculation, V2 represents the voltage in the current calculation, I1 represents the current in the previous calculation, I2 represents the current in the current calculation, T1 represents the temperature in the previous calculation, T2 represents the temperature in the current calculation, S represents the SOC value fed back after the previous calculation, L represents the number of cycles in the current calculation, H represents the SOH in the current calculation, and T represents the time interval between the previous calculation and the current calculation. X(1) represents the first term in the X(m) vector, i.e., V1. X(2) represents the second term in the X(m) vector, i.e., V2, and so on.

[0057] The hidden layer can be determined based on the following formula (I):

[0058]

[0059] Among them, P j X is the output of the hidden layer, f is the activation function of the hidden layer, and X is the output of the hidden layer. i Let θ be the input vector, and i represent the i-th term in the vector X(m); j Let W be a parameter, representing the adjustment factor of the j-th hidden layer; ij denoted as weight, representing the weight factor of the i-th input vector in the j-th hidden layer;

[0060] The output layer can be determined based on the following formula (II):

[0061]

[0062] Where Y is the output of the output layer, W j P is the weight factor of the j-th hidden layer. j Let g represent the output of the j-th hidden layer in Equation (I), and g be the activation function of the output layer.

[0063] In some embodiments, the neural network can be pre-trained. For example, the neural network can be trained periodically. Based on this, training samples can be determined; wherein, the training samples are obtained by calculating data collected within a preset acquisition period during a time period of stable and constant power and small current change during energy storage discharge or charging; the initial neural network model is trained based on the training samples to obtain a pre-trained neural network model.

[0064] In some embodiments, the true state of charge (SOC) is 100% or 0% when the fully charged or fully discharged state is reached, and the neural network model is optimized based on the calibration value.

[0065] In some embodiments, determining the training samples during the charging and discharging process may further include: determining the true state of charge ΔSOC in the training samples based on the following formula;

[0066] Wherein, ΔSOC is determined:

[0067] ΔSOC=S-IT

[0068] Where S is the previously calculated SOC value, and IT is the amount of electricity released during the current time period.

[0069] In some embodiments, the method further includes: obtaining the number of iterations in the current calculation, the health status in the current calculation, and the time interval between the last data collection and the current data collection. Based on this, the above-described method of inputting the last collected data, the last estimated data, and the current collected data into a pre-trained neural network model to obtain the current estimated data includes: inputting the last collected data, the last estimated data, the current collected data, the number of iterations in the current calculation, the health status in the current calculation, and the time interval between the last data collection and the current data collection into a pre-trained neural network model to obtain the current estimated data.

[0070] In some embodiments, the method can be applied to a battery state-of-charge estimation system based on an energy storage big data center. The system may include a data acquisition device connected to a battery management system for collecting lithium battery operating condition data and environmental data, and wirelessly transmitting the lithium battery operating condition data and environmental data.

[0071] The energy storage big data center is used to acquire lithium battery operating condition data and environmental data uploaded by data acquisition devices, and to use the operating condition data and environmental data to calculate the state of charge (SOC) of lithium batteries.

[0072] The data acquisition device communicates with the battery management system via serial port and sends operating condition data and environmental data to the energy storage big data center via mobile communication network.

[0073] Energy storage big data centers include:

[0074] The data cleaning and storage module is used to clean and filter the received lithium battery operating condition data and environmental data, remove abnormal data, and store them in a big data storage medium.

[0075] The calculation module is used to calculate the SOC of the lithium battery. The calculation process is as follows:

[0076] Step 1: Obtain battery operating condition data and environmental data, as well as feedback data from the previous calculation, from the energy storage big data center.

[0077] Step 2: Calculate the difference in battery capacity between the current data and the previous data.

[0078] Step 3: Input the acquired lithium battery operating condition data, environmental data, feedback data from the previous calculation, and battery capacity difference into the neural network model.

[0079] Step 4: The neural network model uses the input data for iterative calculations and simultaneously corrects the computational model.

[0080] The wireless transmission module can be replaced by a wired transmission mode.

[0081] The data acquisition device can interface with the battery management system in a non-serial port mode. It can also interface with the energy management system (EMS) integrated into the energy storage site to obtain lithium battery operating condition data and environmental data from the energy management system.

[0082] Figure 5 This is a schematic diagram of a battery state of charge estimation device provided in an embodiment of the present invention.

[0083] like Figure 5 As shown, the device includes:

[0084] The acquisition module 501 is used to acquire the previous data, the previous estimated data, and the current data. The acquired data includes operating condition data and environmental data. The time interval between the previous data and the current data acquisition is not more than a set threshold.

[0085] The estimation module 502 is used to input the previously collected data, the previously estimated data, and the currently collected data into a pre-trained neural network model to obtain the currently estimated data.

[0086] In some embodiments, operating data includes voltage and current, and environmental data includes temperature.

[0087] In some embodiments, the acquisition module 501 is further configured to: acquire the number of cycles in the current calculation, the health status in the current calculation, and the time interval between the last acquisition and the current acquisition;

[0088] The estimation module 502 is also used to: input the last collected data, the last estimated data, the current collected data, the number of loops in the current calculation, the health status in the current calculation, and the time interval between the last collection and the current collection into the pre-trained neural network model to obtain the current estimated data.

[0089] In some embodiments, the neural network model includes an input layer, a hidden layer, and an output layer; the activation function of the hidden layer is a monotonically differentiable sigmoid function, and the activation function of the output layer is a linear purelin function.

[0090] In some embodiments, the hidden layer is determined based on the above formula (a):

[0091] The output layer is determined based on the above formula (II).

[0092] In some embodiments, the system further includes a training module, configured to: determine training samples; wherein the training samples are obtained by calculating data collected within a preset acquisition period during a time period of stable and constant power and small current change during energy storage discharge or charging; and train an initial neural network model based on the training samples to obtain a pre-trained neural network model.

[0093] In some embodiments, the system further includes: an optimization module for calculating calibration values ​​of the vented and fully charged states of charge based on the ampere-hour integral method; and optimizing the neural network model based on the calibration values.

[0094] In some embodiments, the training module is further configured to: determine training samples by: determining the true state of charge ΔSOC in the training samples based on the following formula.

[0095] The device provided in this embodiment of the invention has the same implementation principle and technical effect as the aforementioned method embodiment. For the sake of brevity, any parts not mentioned in the device embodiment can be referred to the corresponding content in the aforementioned method embodiment.

[0096] Furthermore, this embodiment also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the method provided in the foregoing method embodiments.

[0097] The computer program product of the face recognition method, device and system provided in the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the methods in the preceding method embodiments. For specific implementation, please refer to the method embodiments, which will not be repeated here.

[0098] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0099] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. Furthermore, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Additionally, the coupling or direct coupling or communication connection shown or discussed may be through some communication interface; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0100] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0101] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0102] If the functionality is implemented as a software functional unit and sold or used as an independent product, it can be stored in a processor-executable, non-volatile, computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0103] Finally, it should be noted that the above embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for estimating the state of charge of a battery, characterized in that, include: Acquire the last collected data, the last estimated data, and the current collected data. The collected data includes operating condition data and environmental data. The time interval between the last collected data and the current collected data does not exceed a set threshold. The previously collected data, the previously estimated data, and the currently collected data are input into a pre-trained neural network model to obtain the currently estimated data. Get the loop count for this calculation, the health status for this calculation, and the time interval between the last data collection and this data collection. The step of inputting the previously collected data, the previously estimated data, and the currently collected data into a pre-trained neural network model to obtain the currently estimated data includes: The last collected data, the last estimated data, the current collected data, the number of loops in the current calculation, the health status in the current calculation, and the time interval between the last collection and the current collection are input into a pre-trained neural network model to obtain the current estimated data. The neural network model includes an input layer, a hidden layer, and an output layer; the activation function of the hidden layer is a monotonically differentiable sigmoid function, and the activation function of the output layer is a linear purelin function. Determine the training samples; wherein, the training samples are calculated from data collected within a preset acquisition period during a time period of stable and constant power and small current change during the energy storage discharge or charging process; The initial neural network model is trained based on the training samples to obtain a pre-trained neural network model; The determination of training samples includes: The true state of charge ΔSOC in the training samples is determined based on the following formula; where ΔSOC is determined as follows: ΔSOC=S-IT Where S is the previously calculated SOC value, and IT is the amount of electricity released during this time period; The hidden layer is determined based on the following formula (I): ; Where Pj is the output of the hidden layer, f is the activation function of the hidden layer, Xi is the input vector, i represents the i-th term in the X(m) vector; θj is the parameter, representing the adjustment factor of the j-th hidden layer; Wij is the weight, representing the weight factor of the i-th input vector in the j-th hidden layer. The output layer is determined based on the following formula (II): ; Where Y is the output of the output layer, Wj is the weight factor of the j-th hidden layer, Pj represents the output of the j-th hidden layer in formula (I), and g is the activation function of the output layer.

2. The method according to claim 1, characterized in that, The operating data includes voltage and current, and the environmental data includes temperature.

3. The apparatus for estimating the state of charge of a battery according to claim 1, characterized in that, include: The acquisition module is used to acquire the last collected data, the last estimated data, and the current collected data. The collected data includes operating condition data and environmental data. The time interval between the last collected data and the current collected data does not exceed a set threshold. The estimation module is used to input the previously collected data, the previously estimated data, and the currently collected data into a pre-trained neural network model to obtain the currently estimated data.

4. An electronic device, characterized in that, include: Processor and memory; The memory stores a computer program that, when executed by the processor, performs the method as described in any one of claims 1 to 2.

5. A computer-readable storage medium storing a computer program thereon, characterized in that, The computer program is executed by the processor to perform the method described in any one of claims 1 to 2.

Images