Lithium battery health state cloud monitoring method and system

Abstract

The invention discloses a lithium battery health state cloud monitoring method and system, and relates to the field of lithium batteries, and the method comprises the steps: obtaining a performance index reflecting the health state of a lithium battery in advance, measuring the health state of the lithium battery in an experiment environment, and generating an input data set and an output data set, the built full-connection neural network model is added to carry out model training and then is uploaded to the cloud server, so that the real-time health state of the lithium battery can be output and returned through the cloud server according to the real-time performance index of the lithium battery acquired by the user side, and the performance indexes of the lithium battery in different external environments are continuously acquired subsequently. And optimizing the full-connection neural network model in real time, and dynamically updating the full-connection neural network model of the cloud server. According to the invention, lithium battery health state monitoring is clouded, the problem that accurate monitoring cannot be realized due to insufficient local computing power and readable storage space of a user side is solved, the detection model is continuously optimized and dynamically updated, and the generalization ability of lithium battery health state monitoring is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of lithium battery monitoring, and in particular to a cloud-based monitoring method and system for the health status of a lithium battery. Background Art

[0002] Lithium battery is a rechargeable battery that stores and releases energy based on the migration of lithium ions between positive and negative electrodes. It has the advantages of high energy density, long life and low self-discharge rate. It is widely used in various electronic devices and electric vehicles. The development of lithium batteries has gone through two stages: lithium metal batteries and lithium-ion batteries. The latter has become the mainstream due to its higher safety. In recent years, lithium battery technology has been continuously optimized and its application areas have been continuously expanded.

[0003] Thanks to the wide application of lithium batteries in electric vehicles, energy storage, consumer electronics and other fields, lithium battery health status monitoring technology is crucial to ensure battery safety and understand battery performance and life. This technology monitors key parameters such as battery voltage, current, temperature, capacity, etc., and uses data collection, transmission, processing and analysis to achieve real-time monitoring and prediction of the health status of lithium batteries to ensure their reliability during use.

[0004] However, on user-end devices such as electric vehicles and electronic products, due to the limitations of computing power and the size of readable storage space, it may not be possible to accurately monitor the health status of lithium batteries. On the other hand, the lithium battery health status detection method based on machine learning models will cause the pre-trained machine learning models to lose universality due to changes in the external environment of the lithium battery, and the data sets previously used to train the models will also lose their representativeness to a certain extent. Summary of the invention

[0005] In view of the above-mentioned technical deficiencies, the purpose of the present invention is to provide a cloud-based monitoring method and system for the health status of lithium batteries, aiming to solve the problem that the computing power and readable storage space on the user-side device are limited, making it impossible to accurately monitor the health status of lithium batteries. At the same time, the lithium battery health status detection method based on the machine learning model will cause the pre-trained machine learning model to lose universality due to changes in the external environment of the lithium battery. In addition, the data set previously used to train the model will also lose its representativeness to a certain extent.

[0006] In view of the above problems, the present application provides a cloud-based monitoring method and system for the health status of a lithium battery.

[0007] In a first aspect disclosed in the present application, a method for cloud monitoring of the health status of a lithium battery is provided, the method comprising the following steps: Step 1: Obtain a set of performance indicators reflecting the health status of the lithium battery from each charge and discharge cycle in the entire life cycle of the lithium battery in advance, and determine the health status of the lithium battery by AC impedance analysis and chemical analysis in each charge and discharge cycle in the entire life cycle of the lithium battery, wherein the performance indicators include the number of charge and discharge cycles, charging time, discharging time, voltage variance during charging, voltage variance during discharging, current curve during discharging, lithium battery temperature during charging, and lithium battery temperature during discharging, and the health status of the lithium battery is a percentage value; Step 2: Construct vector x i =[n i ,t i ,T i ,f i ,F i ,s i ,S i ], where n i is the number of charge and discharge cycles in the i-th group of performance indicators, t i is the charging time in the i-th group of performance indicators, T i is the discharge time in the i-th group of performance indicators, f i is the voltage variance during the charging process in the i-th group of performance indicators, F i is the voltage variance during the discharge process in the i-th group of performance indicators, s i is the lithium battery temperature during charging in the i-th group of performance indicators, S i is the temperature of the lithium battery during the discharge process in the i-th group of performance indicators, i=1,2,3, ...,N, N is the number of charge and discharge cycles in the entire life cycle of the lithium battery, and the matrix [x 1 T ,x 2 T ,x 3 T , ...,x N T ] T As input dataset; Step 3: Calculate the discharge capacity of the lithium battery in the ith charge and discharge cycle by the current curve during the discharge process in the ith group of performance indicators, and then add the weighted sum of the lithium battery health status measured by AC impedance analysis and chemical analysis in the ith charge and discharge cycle in the entire life cycle of the lithium battery to obtain the weighted lithium battery health status in the ith charge and discharge cycle, expressed as g i =0.3j i +0.3h i +0.4r i / R×100%, where g i represents the weighted lithium battery health status in the i-th charge and discharge cycle, j iand h i They represent the health status of lithium batteries measured by AC impedance analysis and chemical analysis in the i-th charge and discharge cycle, r i represents the discharge capacity of the lithium battery in the i-th charge and discharge cycle, R represents the rated capacity of the lithium battery, and the vector [g 1 ,g 2 ,g 3 , ... ,g N ] as the output data set, where i=1,2,3, ... ,N, N is the number of charge and discharge cycles in the life cycle of the lithium battery; Step 4: Build a fully connected neural network model, add the input data set and the output data set to the fully connected neural network model to perform model training, and upload the fully connected neural network model that has completed model training to the cloud server; Step 5: Upload the real-time performance indicators of the lithium battery collected by the user end to the cloud server, construct the real-time feature vector of the lithium battery e=[n, t, T, f, F, s, S], add the real-time feature vector of the lithium battery to the fully connected neural network model of the cloud server, output the real-time health status of the lithium battery, and return it to the user end, where n is the number of real-time charge and discharge cycles, t is the real-time charging time, T is the real-time discharge time, f is the voltage variance during the real-time charging process, F is the voltage variance during the real-time discharge process, s is the lithium battery temperature during the real-time charging process, and S is the lithium battery temperature during the real-time discharge process; Step 6: Continue to obtain performance indicators reflecting the health status of the lithium battery from each charge and discharge cycle throughout the life cycle of the lithium battery under different external environments, use the operation of step 2 to generate an additional input data set, and obtain the health status of the lithium battery in each charge and discharge cycle in this scenario through AC impedance analysis and chemical analysis, use the same operation in step 3 to calculate the weighted lithium battery health status, generate an additional output data set, perform real-time gradient descent on the fully connected neural network model of the cloud server through the additional input data set and the additional output data set, optimize the fully connected neural network model in real time, and update the fully connected neural network model of the cloud server to the optimized fully connected neural network model.

[0008] Preferably, the method for calculating the discharge capacity of the lithium battery is as follows: by performing a definite integral calculation on the current curve during the discharge process within the discharge time, the value of the discharge capacity of the lithium battery can be obtained, which is expressed as , where r represents the discharge capacity of the lithium battery, T represents the discharge time, represents the time variable, Represents the current curve during the discharge process.

[0009] Preferably, the steps of constructing the fully connected neural network model are: Create the input layer, the first hidden layer, the second hidden layer and the output layer in sequence, where the input layer includes 7 neurons, the first hidden layer includes 12 neurons, the second hidden layer includes 7 neurons, the output layer has 1 neuron, and the connection mode between neurons in each layer is fully connected; Randomly sample from a Gaussian distribution with a mathematical expectation of 0 and a standard deviation of 1, and assign initial values ​​to the weight parameters of each layer of full connection; Set the loss function to the mean square error function and the activation function to the sigmoid function.

[0010] Preferably, the input data set needs to be normalized before being added to the fully connected neural network model to perform model training.

[0011] Preferably, the lithium battery real-time feature vector needs to be normalized before being added to the fully connected neural network model of the cloud server.

[0012] Preferably, the normalization operation is maximum value normalization, specifically expressed as y m * =y m / max 1:M (y), where y m is the mth value to be normalized, max 1:M (y m ) is the maximum value in a column of M values, y m * for y m The value after maximum normalization is performed.

[0013] The second aspect disclosed in the present application provides a lithium battery health status cloud monitoring system, the system is used for the above-mentioned lithium battery health status cloud monitoring method, the system includes: A collection module, the collection module is used to obtain a set of performance indicators reflecting the health status of the lithium battery from each charge and discharge cycle in the whole life cycle of the lithium battery in advance, and determine the health status of the lithium battery by AC impedance analysis and chemical analysis in each charge and discharge cycle in the whole life cycle of the lithium battery, wherein the performance indicators include the number of charge and discharge cycles, charging time, discharging time, voltage variance during charging, voltage variance during discharging, current curve during discharging, lithium battery temperature during charging, and lithium battery temperature during discharging, and the health status of the lithium battery is a percentage value; Construct a vector module, the vector construction module is used to construct a vector x i =[n i ,t i ,T i ,f i ,F i ,si ,S i ], where n i is the number of charge and discharge cycles in the i-th group of performance indicators, t i is the charging time in the i-th group of performance indicators, T i is the discharge time in the i-th group of performance indicators, f i is the voltage variance during charging in the i-th group of performance indicators, F i is the voltage variance during the discharge process in the i-th group of performance indicators, s i is the lithium battery temperature during charging in the i-th group of performance indicators, S i is the temperature of the lithium battery during the discharge process in the i-th group of performance indicators, i=1,2,3, ...,N, N is the number of charge and discharge cycles in the entire life cycle of the lithium battery, and the matrix [x 1 T ,x 2 T ,x 3 T , ... ,x N T ] T As input dataset; The determination module is used to calculate the discharge capacity of the lithium battery in the i-th charge and discharge cycle through the current curve during the discharge process in the i-th group of performance indicators, and then perform a weighted summation with the health status of the lithium battery determined by the AC impedance analysis method and the chemical analysis method in the i-th charge and discharge cycle within the life cycle of the lithium battery to obtain the weighted health status of the lithium battery in the i-th charge and discharge cycle, expressed as g i =0.3j i +0.3h i +0.4r i / R×100%, where g i represents the weighted lithium battery health status in the i-th charge and discharge cycle, j i and h i They represent the health status of lithium batteries measured by AC impedance analysis and chemical analysis in the i-th charge and discharge cycle, r i represents the discharge capacity of the lithium battery in the i-th charge and discharge cycle, R represents the rated capacity of the lithium battery, and the vector [g 1 ,g 2 ,g 3 , ... ,g N ] as the output data set, where i=1,2,3, ... ,N, N is the number of charge and discharge cycles in the life cycle of the lithium battery; A training module, wherein the training module is used to construct a fully connected neural network model, add an input data set and an output data set to the fully connected neural network model to perform model training, and upload the fully connected neural network model that has completed model training to a cloud server; A live module, which is used to upload the real-time performance indicators of the lithium battery collected by the user end to the cloud server, construct a real-time feature vector of the lithium battery e=[n, t, T, f, F, s, S], add the real-time feature vector of the lithium battery to the fully connected neural network model of the cloud server, output the real-time health status of the lithium battery, and return it to the user end, wherein n is the number of real-time charge and discharge cycles, t is the real-time charging time, T is the real-time discharge time, f is the voltage variance during the real-time charging process, F is the voltage variance during the real-time discharge process, s is the temperature of the lithium battery during the real-time charging process, and S is the temperature of the lithium battery during the real-time discharge process; A real-time optimization module, wherein the real-time optimization module is used to continue to obtain performance indicators reflecting the health status of the lithium battery from each charge and discharge cycle in the entire life cycle of the lithium battery under different external environments, use the operation of step 2 to generate an additional input data set, and obtain the health status of the lithium battery in each charge and discharge cycle in this scenario through AC impedance analysis and chemical analysis, use the same operation in step 3 to calculate the weighted health status of the lithium battery, generate an additional output data set, perform real-time gradient descent on the fully connected neural network model of the cloud server through the additional input data set and the additional output data set, optimize the fully connected neural network model in real time, and update the fully connected neural network model of the cloud server to the optimized fully connected neural network model.

[0014] The third aspect disclosed in the present application provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-mentioned cloud-based monitoring method for the health status of a lithium battery when executing the computer program.

[0015] The beneficial effects of the present invention are: (1) The core computing part of lithium battery health status monitoring is transferred to the cloud server. The user side only needs to pay attention to uploading the real-time performance indicators of the lithium battery and receiving the real-time health status of the lithium battery returned, which solves the problem that the computing power and readable storage space on the user-side device are limited and the health status of the lithium battery cannot be accurately monitored.

[0016] (2) Continuously obtain the performance indicators of lithium batteries under different external environments, optimize the fully connected neural network model in real time, and dynamically update the fully connected neural network model of the cloud server, so as to reduce the problem of pre-trained models losing universality due to changes in the external environment of lithium batteries, improve the generalization ability of subsequent lithium battery health status monitoring, and make the monitoring results more and more accurate.

[0017] (3) Two hidden layers are used in the fully connected neural network model, which improves the model's expressiveness and fitting capabilities. It can capture more complex and subtle changes in data features, making lithium battery health status monitoring more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 The figure is an overall flow chart of the cloud-based monitoring method for the health status of lithium batteries.

[0020] Figure 2 This is the overall structure diagram of the fully connected neural network model.

[0021] Figure 3 This is the overall structure diagram of the lithium battery health status cloud monitoring system.

[0022] Explanation of reference numerals: input layer 10 , first hidden layer 20 , second hidden layer 30 , output layer 40 . DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0024] like Figure 1 As shown, the embodiment of the present application provides a method for cloud monitoring of the health status of a lithium battery, the method comprising the following steps: Step 1: Obtain a set of performance indicators reflecting the health status of the lithium battery from each charge and discharge cycle in the entire life cycle of the lithium battery in advance, and determine the health status of the lithium battery by AC impedance analysis and chemical analysis in each charge and discharge cycle in the entire life cycle of the lithium battery, wherein the performance indicators include the number of charge and discharge cycles, charging time, discharging time, voltage variance during charging, voltage variance during discharging, current curve during discharging, lithium battery temperature during charging, and lithium battery temperature during discharging, and the health status of the lithium battery is a percentage value; Step 2: Construct vector x i =[n i ,t i ,Ti ,f i ,F i ,s i ,S i ], where n i is the number of charge and discharge cycles in the i-th group of performance indicators, t i is the charging time in the i-th group of performance indicators, T i is the discharge time in the i-th group of performance indicators, f i is the voltage variance during the charging process in the i-th group of performance indicators, F i is the voltage variance during the discharge process in the i-th group of performance indicators, s i is the lithium battery temperature during charging in the i-th group of performance indicators, S i is the temperature of the lithium battery during the discharge process in the i-th group of performance indicators, i=1,2,3, ...,N, N is the number of charge and discharge cycles in the entire life cycle of the lithium battery, and the matrix [x 1 T ,x 2 T ,x 3 T , ...,x N T ] T As input dataset; Step 3: Calculate the discharge capacity of the lithium battery in the ith charge and discharge cycle by the current curve during the discharge process in the ith group of performance indicators, and then add the weighted sum of the lithium battery health status measured by AC impedance analysis and chemical analysis in the ith charge and discharge cycle in the entire life cycle of the lithium battery to obtain the weighted lithium battery health status in the ith charge and discharge cycle, expressed as g i =0.3j i +0.3h i +0.4r i / R×100%, where g i represents the weighted lithium battery health status in the i-th charge and discharge cycle, j i and h i They represent the health status of lithium batteries measured by AC impedance analysis and chemical analysis in the i-th charge and discharge cycle, r i represents the discharge capacity of the lithium battery in the i-th charge and discharge cycle, R represents the rated capacity of the lithium battery, and the vector [g 1 ,g 2 ,g 3 , ... ,g N ] as the output data set, where i=1,2,3, ... ,N, where N is the number of charge and discharge cycles in the entire life cycle of the lithium battery; Step 4: Build a fully connected neural network model. The model structure is as follows Figure 2As shown, the input data set and the output data set are added to the fully connected neural network model to perform model training, and the fully connected neural network model that has completed model training is uploaded to the cloud server; Step 5: Upload the real-time performance indicators of the lithium battery collected by the user end to the cloud server, construct the real-time feature vector of the lithium battery e=[n, t, T, f, F, s, S], add the real-time feature vector of the lithium battery to the fully connected neural network model of the cloud server, output the real-time health status of the lithium battery, and return it to the user end, where n is the number of real-time charge and discharge cycles, t is the real-time charging time, T is the real-time discharge time, f is the voltage variance during the real-time charging process, F is the voltage variance during the real-time discharge process, s is the lithium battery temperature during the real-time charging process, and S is the lithium battery temperature during the real-time discharge process; Step 6: Continue to obtain performance indicators reflecting the health status of the lithium battery from each charge and discharge cycle throughout the life cycle of the lithium battery under different external environments, use the operation of step 2 to generate an additional input data set, and obtain the health status of the lithium battery in each charge and discharge cycle in this scenario through AC impedance analysis and chemical analysis, use the same operation in step 3 to calculate the weighted lithium battery health status, generate an additional output data set, perform real-time gradient descent on the fully connected neural network model of the cloud server through the additional input data set and the additional output data set, optimize the fully connected neural network model in real time, and update the fully connected neural network model of the cloud server to the optimized fully connected neural network model.

[0025] Specifically, continuously acquiring the performance indicators of lithium batteries under different external environments, optimizing the fully connected neural network model in real time, and dynamically updating the fully connected neural network model of the cloud server can reduce the problem of pre-trained models losing universality due to changes in the external environment of lithium batteries, improve the generalization ability of subsequent lithium battery health status monitoring, and make the monitoring results more and more accurate.

[0026] Furthermore, the calculation method of the lithium battery discharge capacity is: by calculating the definite integral of the current curve during the discharge process within the discharge time, the value of the lithium battery discharge capacity can be obtained, which is expressed as , where r represents the discharge capacity of the lithium battery, T represents the discharge time, represents the time variable, Represents the current curve during the discharge process; Specifically, since the current in the discharge process changes with the work power of the lithium battery, it is usually not constant, but an irregular function curve that changes with time. The discharge capacity of the lithium battery is the definite integral of the function curve within the discharge time interval.

[0027] Furthermore, the steps of constructing the fully connected neural network model are: like Figure 2 As shown, an input layer 10, a first hidden layer 20, a second hidden layer 30 and an output layer 40 are created in sequence, wherein the input layer 10 includes 7 neurons, the first hidden layer 20 includes 12 neurons, the second hidden layer 30 includes 7 neurons, and the output layer 40 is 1 neuron. The connection mode of neurons between each layer is full connection, as shown in FIG. Figure 2 As shown; Specifically, the fully connected neural network model uses a four-layer structure. In addition to the input layer 10 and the output layer 20, two hidden layers 20 and 30 are used, which improves the expression and fitting capabilities of the model and can capture more complex and subtle changes in data features; On the other hand, the number of neurons in the hidden layer is usually determined using an empirical formula, which can be expressed as: the number of neurons in the hidden layer < 2× the number of neurons in the input layer, so the number of neurons in the first hidden layer 20 and the second hidden layer 30 should both be less than 16.

[0028] Randomly sample from a Gaussian distribution with a mathematical expectation of 0 and a standard deviation of 1, and assign initial values ​​to the weight parameters of each layer of full connection; Specifically, randomly sampling from a Gaussian distribution to initialize the weight parameters can ensure that the weight parameters have a small initial value, which helps to stabilize the training of the model.

[0029] Set the loss function to mean square error function and the activation function to sigmoid function; Specifically, the mean square error function calculates the square of the difference between the weighted lithium battery health status value and the lithium battery health status value predicted by the model, and performs back propagation and gradient descent by performing partial differential calculations on the mean square error function. The sigmoid function is a classic activation function in the form of f ( x )=1 / (1+ exp (- x )), its output is between 0 and 1, which allows it to be used in the output layer to directly represent the percentage value of the health status of the lithium battery.

[0030] Furthermore, the input data set needs to be normalized before being added to the fully connected neural network model to perform model training.

[0031] Furthermore, the lithium battery actual feature vector needs to be normalized before being added to the fully connected neural network model of the cloud server.

[0032] Furthermore, the normalization operation is maximum value normalization, specifically expressed as y m * =y m / max 1:M (y), where ym is the mth value to be normalized, max 1:M (y m ) is the maximum value in a column of M values, y m * for y m The value after performing maximum normalization; Specifically, maximum normalization scales all data to between -1 and 1, which has the advantages of simple calculation and insensitivity to outliers, while avoiding the gradient vanishing problem caused by the use of the sigmoid activation function. Although numerical instability may occur when the maximum value in the data set changes greatly, the performance indicators of the health status of lithium batteries are stable and will not fluctuate greatly, thus avoiding this problem.

[0033] In summary, the cloud-based monitoring method for the health status of a lithium battery provided in the embodiment of the present application has the following technical effects: (1) The core computing part of lithium battery health status monitoring is transferred to the cloud server. The user side only needs to pay attention to uploading the real-time performance indicators of the lithium battery and receiving the real-time health status of the lithium battery returned, which solves the problem that the computing power and readable storage space on the user-side device are limited and the health status of the lithium battery cannot be accurately monitored.

[0034] (2) Continuously obtain the performance indicators of lithium batteries under different external environments, optimize the fully connected neural network model in real time, and dynamically update the fully connected neural network model of the cloud server, so as to reduce the problem of pre-trained models losing universality due to changes in the external environment of lithium batteries, improve the generalization ability of subsequent lithium battery health status monitoring, and make the monitoring results more and more accurate.

[0035] (3) Two hidden layers are used in the fully connected neural network model, which improves the model's expressiveness and fitting capabilities. It can capture more complex and subtle changes in data features, making lithium battery health status monitoring more accurate.

[0036] Based on the same inventive concept as the cloud monitoring method for the health status of a lithium battery in the aforementioned embodiment, Figure 3 As shown, the present application provides a lithium battery health status cloud monitoring system, the system comprising: A collection module, the collection module is used to obtain a set of performance indicators reflecting the health status of the lithium battery from each charge and discharge cycle in the whole life cycle of the lithium battery in advance, and determine the health status of the lithium battery by AC impedance analysis and chemical analysis in each charge and discharge cycle in the whole life cycle of the lithium battery, wherein the performance indicators include the number of charge and discharge cycles, charging time, discharging time, voltage variance during charging, voltage variance during discharging, current curve during discharging, lithium battery temperature during charging, and lithium battery temperature during discharging, and the health status of the lithium battery is a percentage value; Construct a vector module, the vector construction module is used to construct a vector x i =[n i ,t i ,T i ,f i ,F i ,s i ,S i ], where n i is the number of charge and discharge cycles in the i-th group of performance indicators, t i is the charging time in the i-th group of performance indicators, T i is the discharge time in the i-th group of performance indicators, f i is the voltage variance during charging in the i-th group of performance indicators, F i is the voltage variance during the discharge process in the i-th group of performance indicators, s i is the lithium battery temperature during charging in the i-th group of performance indicators, S i is the temperature of the lithium battery during the discharge process in the i-th group of performance indicators, i=1,2,3, ...,N, N is the number of charge and discharge cycles in the entire life cycle of the lithium battery, and the matrix [x 1 T ,x 2 T ,x 3 T , ... ,x N T ] T As input dataset; The determination module is used to calculate the discharge capacity of the lithium battery in the i-th charge and discharge cycle through the current curve during the discharge process in the i-th group of performance indicators, and then perform a weighted summation with the health status of the lithium battery determined by the AC impedance analysis method and the chemical analysis method in the i-th charge and discharge cycle within the life cycle of the lithium battery to obtain the weighted health status of the lithium battery in the i-th charge and discharge cycle, expressed as g i =0.3j i +0.3h i +0.4r i / R×100%, where g irepresents the weighted lithium battery health status in the i-th charge and discharge cycle, j i and h i They represent the health status of lithium batteries measured by AC impedance analysis and chemical analysis in the i-th charge and discharge cycle, r i represents the discharge capacity of the lithium battery in the i-th charge and discharge cycle, R represents the rated capacity of the lithium battery, and the vector [g 1 ,g 2 ,g 3 , ... ,g N ] as the output data set, where i=1,2,3, ... ,N, where N is the number of charge and discharge cycles in the entire life cycle of the lithium battery; A training module, wherein the training module is used to construct a fully connected neural network model, add an input data set and an output data set to the fully connected neural network model to perform model training, and upload the fully connected neural network model that has completed model training to a cloud server; A live module, which is used to upload the real-time performance indicators of the lithium battery collected by the user end to the cloud server, construct a real-time feature vector of the lithium battery e=[n, t, T, f, F, s, S], add the real-time feature vector of the lithium battery to the fully connected neural network model of the cloud server, output the real-time health status of the lithium battery, and return it to the user end, wherein n is the number of real-time charge and discharge cycles, t is the real-time charging time, T is the real-time discharge time, f is the voltage variance during the real-time charging process, F is the voltage variance during the real-time discharge process, s is the temperature of the lithium battery during the real-time charging process, and S is the temperature of the lithium battery during the real-time discharge process; A real-time optimization module, wherein the real-time optimization module is used to continue to obtain performance indicators reflecting the health status of the lithium battery from each charge and discharge cycle in the entire life cycle of the lithium battery under different external environments, use the operation of step 2 to generate an additional input data set, and obtain the health status of the lithium battery in each charge and discharge cycle in this scenario through AC impedance analysis and chemical analysis, use the same operation in step 3 to calculate the weighted health status of the lithium battery, generate an additional output data set, perform real-time gradient descent on the fully connected neural network model of the cloud server through the additional input data set and the additional output data set, optimize the fully connected neural network model in real time, and update the fully connected neural network model of the cloud server to the optimized fully connected neural network model.

[0037] Through the above-mentioned detailed description of a cloud-based monitoring method for the health status of a lithium battery, this specification allows those skilled in the art to clearly understand a cloud-based monitoring system for the health status of a lithium battery in this embodiment. Since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method section.

[0038] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the steps of the above-mentioned cloud-based monitoring method for the health status of a lithium battery are implemented.

[0039] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0040] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A cloud-based monitoring method for the health status of a lithium battery, characterized in that: The following steps are involved: Step 1: Obtain a set of performance indicators reflecting the health status of the lithium battery from each charge and discharge cycle in the entire life cycle of the lithium battery in advance, and determine the health status of the lithium battery by AC impedance analysis and chemical analysis in each charge and discharge cycle in the entire life cycle of the lithium battery, wherein the performance indicators include the number of charge and discharge cycles, charging time, discharging time, voltage variance during charging, voltage variance during discharging, current curve during discharging, lithium battery temperature during charging, and lithium battery temperature during discharging, and the health status of the lithium battery is a percentage value; Step 2: Construct vector x i =[n i ,t i ,T i ,f i ,F i ,s i ,S i ], where n i is the number of charge and discharge cycles in the i-th group of performance indicators, t i is the charging time in the i-th group of performance indicators, T i is the discharge time in the i-th group of performance indicators, f i is the voltage variance during the charging process in the i-th group of performance indicators, F i is the voltage variance during the discharge process in the i-th group of performance indicators, s i is the lithium battery temperature during charging in the i-th group of performance indicators, S i is the temperature of the lithium battery during the discharge process in the i-th group of performance indicators, i=1,2,3, ...,N, N is the number of charge and discharge cycles in the entire life cycle of the lithium battery, and the matrix [x1 T ,x2 T ,x3 T , ... ,x N T ] T As input dataset; Step 3: Calculate the discharge capacity of the lithium battery in the ith charge and discharge cycle by the current curve during the discharge process in the ith group of performance indicators, and then add the weighted sum of the lithium battery health status measured by AC impedance analysis and chemical analysis in the ith charge and discharge cycle in the entire life cycle of the lithium battery to obtain the weighted lithium battery health status in the ith charge and discharge cycle, expressed as g i =0.3j i +0.3h i +0.4r i / R×100%, where g i represents the weighted lithium battery health status in the i-th charge and discharge cycle, j i and h i They represent the health status of lithium batteries measured by AC impedance analysis and chemical analysis in the i-th charge and discharge cycle, r i represents the discharge capacity of the lithium battery in the i-th charge and discharge cycle, R represents the rated capacity of the lithium battery, and the vector [g1,g2,g3,... ,g N ] as the output data set, where i=1,2,3, ... ,N, where N is the number of charge and discharge cycles in the entire life cycle of the lithium battery; Step 4: Build a fully connected neural network model, add the input data set and the output data set to the fully connected neural network model to perform model training, and upload the fully connected neural network model that has completed model training to the cloud server; Step 5: Upload the real-time performance indicators of the lithium battery collected by the user end to the cloud server, construct the real-time feature vector of the lithium battery e=[n, t, T, f, F, s, S], add the real-time feature vector of the lithium battery to the fully connected neural network model of the cloud server, output the real-time health status of the lithium battery, and return it to the user end, where n is the number of real-time charge and discharge cycles, t is the real-time charging time, T is the real-time discharge time, f is the voltage variance during the real-time charging process, F is the voltage variance during the real-time discharge process, s is the lithium battery temperature during the real-time charging process, and S is the lithium battery temperature during the real-time discharge process; Step 6: Continue to obtain performance indicators reflecting the health status of the lithium battery from each charge and discharge cycle throughout the life cycle of the lithium battery under different external environments, use the operation of step 2 to generate an additional input data set, and obtain the health status of the lithium battery in each charge and discharge cycle in this scenario through AC impedance analysis and chemical analysis, use the same operation in step 3 to calculate the weighted lithium battery health status, generate an additional output data set, perform real-time gradient descent on the fully connected neural network model of the cloud server through the additional input data set and the additional output data set, optimize the fully connected neural network model in real time, and update the fully connected neural network model of the cloud server to the optimized fully connected neural network model.

2. A lithium battery health status cloud monitoring method as claimed in claim 1, characterized in that: The calculation method of the lithium battery discharge capacity is as follows: by calculating the definite integral of the current curve during the discharge process within the discharge time, the value of the lithium battery discharge capacity can be obtained, which is expressed as , where r represents the discharge capacity of the lithium battery, P represents the discharge time, represents the time variable, Represents the current curve during the discharge process.

3. A lithium battery health status cloud monitoring method as claimed in claim 1, characterized in that: The steps for constructing the fully connected neural network model are: Create the input layer, the first hidden layer, the second hidden layer and the output layer in sequence, where the input layer includes 7 neurons, the first hidden layer includes 12 neurons, the second hidden layer includes 7 neurons, the output layer has 1 neuron, and the connection mode between neurons in each layer is fully connected; Randomly sample from a Gaussian distribution with a mathematical expectation of 0 and a standard deviation of 1, and assign initial values ​​to the weight parameters of each layer of full connection; Set the loss function to the mean square error function and the activation function to the sigmoid function.

4. A method for cloud monitoring of lithium battery health status as claimed in claim 1, characterized in that: The input data set also needs to be normalized before being added to the fully connected neural network model to perform model training.

5. A lithium battery health status cloud monitoring method as claimed in claim 1, characterized in that: The lithium battery actual feature vector also needs to be normalized before being added to the fully connected neural network model of the cloud server.

6. A method for cloud monitoring of lithium battery health status as claimed in claim 4 or 5, characterized in that: The normalization operation is maximum value normalization, specifically expressed as y m * =y m / max 1:M (y), where y m is the mth value to be normalized, max 1:M (y m ) is the maximum value in a column of M values, y m * for y m The value after maximum normalization is performed.

7. A lithium battery health status cloud monitoring system, the system comprising: A collection module, the collection module is used to obtain a set of performance indicators reflecting the health status of the lithium battery from each charge and discharge cycle in the whole life cycle of the lithium battery in advance, and determine the health status of the lithium battery by AC impedance analysis and chemical analysis in each charge and discharge cycle in the whole life cycle of the lithium battery, wherein the performance indicators include the number of charge and discharge cycles, charging time, discharging time, voltage variance during charging, voltage variance during discharging, current curve during discharging, lithium battery temperature during charging, and lithium battery temperature during discharging, and the health status of the lithium battery is a percentage value; Construct a vector module, the vector construction module is used to construct a vector x i =[n i ,t i ,T i ,f i ,F i ,s i ,S i ], where n i is the number of charge and discharge cycles in the i-th group of performance indicators, t i is the charging time in the i-th group of performance indicators, T i is the discharge time in the i-th group of performance indicators, f i is the voltage variance during the charging process in the i-th group of performance indicators, F i is the voltage variance during the discharge process in the i-th group of performance indicators, s i is the lithium battery temperature during charging in the i-th group of performance indicators, S i is the temperature of the lithium battery during the discharge process in the i-th group of performance indicators, i=1,2,3, ...,N, N is the number of charge and discharge cycles in the entire life cycle of the lithium battery, and the matrix [x1 T ,x2 T ,x3 T , ... ,x N T ] T As input dataset; The determination module is used to calculate the discharge capacity of the lithium battery in the i-th charge and discharge cycle through the current curve during the discharge process in the i-th group of performance indicators, and then perform a weighted summation with the health status of the lithium battery determined by the AC impedance analysis method and the chemical analysis method in the i-th charge and discharge cycle within the life cycle of the lithium battery to obtain the weighted health status of the lithium battery in the i-th charge and discharge cycle, expressed as g i =0.3j i +0.3h i +0.4r i / R×100%, where g i represents the weighted lithium battery health status in the i-th charge and discharge cycle, j i and h i They represent the health status of lithium batteries measured by AC impedance analysis and chemical analysis in the i-th charge and discharge cycle, r i represents the discharge capacity of the lithium battery in the i-th charge and discharge cycle, R represents the rated capacity of the lithium battery, and the vector [g1,g2,g3, ... ,g N ] as the output data set, where i=1,2,3, ... ,N, where N is the number of charge and discharge cycles in the entire life cycle of the lithium battery; A training module, wherein the training module is used to construct a fully connected neural network model, add an input data set and an output data set to the fully connected neural network model to perform model training, and upload the fully connected neural network model that has completed model training to a cloud server; A live module, which is used to upload the real-time performance indicators of the lithium battery collected by the user end to the cloud server, construct a real-time feature vector of the lithium battery e=[n, t, T, f, F, s, S], add the real-time feature vector of the lithium battery to the fully connected neural network model of the cloud server, output the real-time health status of the lithium battery, and return it to the user end, wherein n is the number of real-time charge and discharge cycles, t is the real-time charging time, T is the real-time discharge time, f is the voltage variance during the real-time charging process, F is the voltage variance during the real-time discharge process, s is the temperature of the lithium battery during the real-time charging process, and S is the temperature of the lithium battery during the real-time discharge process; A real-time optimization module, wherein the real-time optimization module is used to continue to obtain performance indicators reflecting the health status of the lithium battery from each charge and discharge cycle in the entire life cycle of the lithium battery under different external environments, use the operation of step 2 to generate an additional input data set, and obtain the health status of the lithium battery in each charge and discharge cycle in this scenario through AC impedance analysis and chemical analysis, use the same operation in step 3 to calculate the weighted health status of the lithium battery, generate an additional output data set, perform real-time gradient descent on the fully connected neural network model of the cloud server through the additional input data set and the additional output data set, optimize the fully connected neural network model in real time, and update the fully connected neural network model of the cloud server to the optimized fully connected neural network model.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of a cloud-based monitoring method for the health status of a lithium battery as described in any one of claims 1 to 6 are implemented.

Images