Safety early warning method and system for out-of-control power battery under storage condition

Abstract

The invention discloses a safety early warning method and system for out-of-control of a power battery under a storage condition. The method comprises the following steps: acquiring a safety feature prediction model of a to-be-tested power battery; safety feature data of the to-be-tested power battery are collected, and the safety feature data at least comprise warehouse temperature, warehouse humidity, battery internal resistance, battery current, battery voltage and battery temperature; inputting the security feature data into a security feature prediction model to obtain a security feature prediction value of the to-be-detected power battery after a preset duration; comparing the safety feature predicted value with a safety feature early warning threshold value to obtain a first comparison result; based on the first comparison result, a safety early warning strategy set of the power battery to be detected is generated, and the safety early warning strategy set at least comprises the steps of increasing the collection frequency of the safety feature data and outputting an early warning instruction. The technical problem of how to carry out safety early warning on the power battery in the storage state is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of power battery storage, and in particular to a safety early warning method and system for power battery runaway under storage conditions. Background Art

[0002] There can be several, dozens or even hundreds of power batteries stored in a warehouse. Since the state of power batteries is constantly changing, it brings complex problems to the management and maintenance of batteries. During storage, the battery is not static, and its chemical properties and physical characteristics will gradually change over time, causing some problems that will affect battery performance. The battery storage environment (such as temperature, humidity, etc.) will significantly affect its internal chemical reaction rate, thereby affecting the battery life and safety. In addition, power batteries should be stored in a suitable temperature space. Generally speaking, if there is no obvious bulge, it is impossible to judge the quality of the power battery.

[0003] Existing safety predictions and warnings for power batteries mainly use a factor of battery thermal runaway as an indicator to predict and issue warnings on the short-term safety of batteries when the vehicle is driving or charging. Currently, there is no effective safety warning plan for power batteries in storage. Summary of the invention

[0004] The embodiments of the present invention provide a method and system for providing a safety warning of a power battery out of control under storage conditions, so as to at least solve the technical problem of how to provide a safety warning for a power battery under storage conditions.

[0005] According to one aspect of an embodiment of the present invention, a safety warning method for power battery runaway under storage conditions is provided, comprising: obtaining a safety feature prediction model for a power battery to be tested; collecting safety feature data of the power battery to be tested, wherein the safety feature data at least includes: warehouse temperature, warehouse humidity, battery internal resistance, battery current, battery voltage and battery temperature; inputting the safety feature data into the safety feature prediction model to obtain a safety feature prediction value of the power battery to be tested after a preset period of time; comparing the safety feature prediction value with a safety feature warning threshold to obtain a first comparison result; based on the first comparison result, generating a safety warning strategy set for the power battery to be tested, the safety warning strategy set at least including: increasing the collection frequency of the safety feature data and outputting a warning indication.

[0006] Furthermore, based on the first comparison result, a safety warning strategy set for the power battery to be tested is generated, including: when the safety feature prediction value reaches a first safety feature warning threshold, a first target instruction of the safety warning strategy set is generated, and the first target instruction is used to control the collection frequency of the safety feature data to be increased to a first frequency.

[0007] Furthermore, based on the first comparison result, a safety warning strategy set for the power battery to be tested is generated, which also includes: when the safety feature prediction value reaches a second safety feature warning threshold, a second target instruction of the safety warning strategy set is generated, and the second target instruction is used to control the collection frequency of the safety feature data to be increased to a second frequency and output a low-level warning indication, wherein the second frequency is greater than the first frequency, and the second safety feature warning threshold is greater than the first safety feature warning threshold.

[0008] Furthermore, based on the first comparison result, a safety warning strategy set for the power battery to be tested is generated, which also includes: when the safety feature prediction value reaches the third safety feature warning threshold, a third target instruction of the safety warning strategy set is generated, and the third target instruction is used to control the collection frequency of the safety feature data to be increased to a third frequency and output an advanced warning indication, wherein the third frequency is greater than the second frequency, and the third safety feature warning threshold is greater than the second safety feature warning threshold.

[0009] Furthermore, obtaining a safety feature prediction model for the power battery to be tested includes constructing and training a safety feature prediction model, and constructing and training the safety feature prediction model includes: inducing thermal runaway on an experimental power battery to obtain safety feature experimental data of the experimental power battery, and dividing the safety feature experimental data into training data and test data, wherein the model of the experimental power battery is consistent with that of the power battery to be tested; constructing an LSTM neural network model; based on the training data, adjusting the hyperparameters of the LSTM neural network model to obtain a first prediction model; based on the test data, verifying the prediction accuracy of the first prediction model; and when the prediction accuracy of the first prediction model meets the preset accuracy requirements, using the first prediction model as the safety feature prediction model.

[0010] Furthermore, constructing and training the security feature prediction model also includes: optimizing the security feature prediction model in combination with the SCHO hyperbolic sine optimization algorithm.

[0011] Furthermore, the safety warning method also includes: obtaining the detection concentration of harmful gases in the area where the power battery to be tested is located; comparing the detection concentration with the concentration threshold to obtain a second comparison result; and outputting a warning prompt based on the second comparison result.

[0012] According to another aspect of an embodiment of the present invention, a safety warning system for power battery runaway under storage conditions is also provided, including: an acquisition module, the acquisition module is used to acquire a safety feature prediction model of the power battery to be tested; a collection module, the collection module is used to collect safety feature data of the power battery to be tested, wherein the safety feature data at least includes: warehouse temperature, warehouse humidity, battery internal resistance, battery current, battery voltage and battery temperature; a first generation module, the first generation module is used to input the safety feature data into the safety feature prediction model to obtain a safety feature prediction value of the power battery to be tested after a preset time period; a comparison module, the comparison module is used to compare the safety feature prediction value with the safety feature warning threshold to obtain a first comparison result; a second generation module, the second generation module is used to generate a safety warning strategy set for the power battery to be tested based on the first comparison result, the safety warning strategy set at least including: increasing the collection frequency of the safety feature data and outputting a warning indication.

[0013] According to another aspect of an embodiment of the present invention, a computer storage medium is provided. The computer storage medium includes a stored program, wherein when the program is running, the device where the computer storage medium is located is controlled to execute the above-mentioned security warning method.

[0014] According to another aspect of an embodiment of the present invention, a computer program product is provided, including a computer program, and the computer program is executed by a processor to execute the above-mentioned security warning method.

[0015] In an embodiment of the present invention, a safety feature prediction model for a power battery to be tested is constructed, safety feature data of the power battery to be tested is collected, the safety feature data is input into the safety feature prediction model, the safety feature prediction model is used to predict the changes in the safety features of the power battery to be tested in the short term in the future, and a safety feature prediction value is obtained, the safety feature prediction value is compared with the safety feature warning threshold, a first comparison result is obtained, and a corresponding safety warning strategy is generated according to the first comparison result. In the above scheme, the safety features of the power battery to be tested are predicted in the short term through the safety feature prediction model, and the safety warning of the power battery to be tested in the storage state is given according to the prediction result, thereby realizing the safe storage of the power battery. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0017] Figure 1 It is a hardware structure block diagram of a computer terminal (or mobile device) for implementing a safety warning method for power battery runaway under storage conditions according to an embodiment of the present invention;

[0018] Figure 2 It is a flow chart of the safety warning method for power battery out of control in the present invention;

[0019] Figure 3 It is a structural block diagram of the safety warning system for power battery runaway in the present invention. DETAILED DESCRIPTION

[0020] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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 should fall within the scope of protection of the present invention.

[0021] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0022] Example 1

[0023] According to an embodiment of the present invention, an embodiment of a safety warning method for power battery out of control under storage conditions 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 may be executed in an order different from that shown here.

[0024] Figure 1 1 is a hardware structure block diagram of a computer terminal (or mobile device) for implementing a safety warning method for power battery runaway under storage conditions according to an embodiment of the present invention. Figure 1As shown, the computer terminal (or mobile device) may include one or more processors 102 (the processor may include but is not limited to a central processing unit (CPU), a graphics processing unit (GPU), a digital signal processing (DSP) chip, a microprocessor (MCU), a field programmable logic device (FPGA), a neural network processor (NPU), a tensor processing unit (TPU), an artificial intelligence (AI) type processor, etc.) and a memory 104 for storing data. In addition, a transmission device 106 for communication functions, an input and output device 108, and a display 110 may also be included. It can be understood by those skilled in the art that Figure 1 The structure shown is for illustration only and does not limit the structure of the above-mentioned computer terminal (or mobile device). For example, the computer terminal may also include more or fewer components than the above-mentioned structure description, or have a configuration different from the above-mentioned structure description.

[0025] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the safety warning method for power battery out of control under storage conditions in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the safety warning method for power battery out of control under storage conditions mentioned above is realized. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the mobile terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0026] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0027] The display 110 may be a touch screen type liquid crystal display (LCD). The liquid crystal display may enable a user to interact with a user interface of the mobile terminal. In some embodiments, the mobile terminal has a graphical user interface (GUI), and a user may interact with the GUI by finger contact and / or gestures on a touch-sensitive surface, wherein the human-computer interaction functions here may optionally include the following interactions: creating web pages, drawing, word processing, making electronic documents, games, video conferencing, instant messaging, sending and receiving emails, call interfaces, playing digital videos, playing digital music and / or web browsing, etc. Executable instructions for performing the above human-computer interaction functions are configured / stored in a computer program product or readable storage medium executable by one or more processors.

[0028] Figure 2 is a flow chart of a safety warning method for power battery out of control under storage conditions according to one embodiment of the present invention. Figure 2 As shown, the process includes the following steps:

[0029] Step S1: Obtain a safety feature prediction model for the power battery to be tested.

[0030] Step S2: collecting safety characteristic data of the power battery to be tested, wherein the safety characteristic data at least includes: warehouse temperature, warehouse humidity, battery internal resistance, battery current, battery voltage and battery temperature.

[0031] Specifically, the power batteries to be tested in the warehouse are placed on battery shelves, and a whole row of shelves is divided into an area. An ambient temperature sensor and a humidity sensor are set in each area. The ambient temperature sensor is used to detect the warehouse temperature T C , humidity sensor is used to detect warehouse humidity H C .

[0032] The power batteries to be tested in each area are connected to the input terminal of the battery management system (BMS) in the prescribed order and polarity, and the battery internal resistance R of the power batteries to be tested is detected by the battery management system (BMS). B 、Battery current I B , battery voltage V B And the battery temperature W B .

[0033] The acquisition module is connected to the ambient temperature sensor, humidity sensor and battery management system (BMS) signal, and collects the safety characteristic data of the power battery to be tested at a preset frequency, and converts the collected warehouse temperature T C , Warehouse humidity H C 、Battery internal resistance H C 、Battery current I B , battery voltage V BAnd the battery temperature W B Preprocessing is performed and the collected security feature data is processed into a matrix form, namely D C =[T C ,H C ], D B =[I B ,V B ,C B ,W B ].

[0034] Step S3: inputting the safety feature data into a safety feature prediction model to obtain a safety feature prediction value of the power battery to be tested after a preset time period.

[0035] Step S4: Compare the safety feature prediction value with the safety feature warning threshold to obtain a first comparison result.

[0036] Step S5: Based on the first comparison result, a safety warning strategy set for the power battery to be tested is generated, where the safety warning strategy set at least includes: increasing the frequency of collecting safety feature data and outputting a warning indication.

[0037] In an embodiment of the present application, a safety feature prediction model for a power battery to be tested is constructed, safety feature data of the power battery to be tested is collected, the safety feature data is input into the safety feature prediction model, the safety feature prediction model is used to predict the changes in the safety features of the power battery to be tested in the short term in the future, and a safety feature prediction value is obtained, the safety feature prediction value is compared with the safety feature warning threshold, a first comparison result is obtained, and a corresponding safety warning strategy is generated according to the first comparison result. In the above scheme, the safety features of the power battery to be tested are predicted in the short term through the safety feature prediction model, and the safety warning of the power battery to be tested in the storage state is given according to the prediction result, thereby realizing the safe storage of the power battery.

[0038] In an exemplary implementation of the present application, in step S5, a safety warning strategy set for the power battery to be tested is generated based on the first comparison result, including the following situations:

[0039] Case 1: When the security feature prediction value reaches the first security feature warning threshold, a first target instruction of the security warning strategy set is generated, and the first target instruction is used to control the collection frequency of the security feature data to be increased to a first frequency.

[0040] It should be noted that when any data in the safety feature prediction value is equal to or greater than the corresponding first safety feature warning threshold, it is regarded that the current power battery to be tested has reached the warning level, and the sampling frequency of the acquisition module needs to be increased to shorten the prediction interval of the power battery to be tested, thereby accurately warning the power battery to be tested and eliminating safety hazards in a timely manner.

[0041] Case 2: When the predicted value of the security feature reaches the second security feature warning threshold, a second target instruction of the security warning strategy set is generated. The second target instruction is used to control the collection frequency of the security feature data to be increased to a second frequency and to output a low-level warning indication, wherein the second frequency is greater than the first frequency, and the second security feature warning threshold is greater than the first security feature warning threshold.

[0042] It should be noted that when any data in the safety feature prediction value is equal to or greater than the corresponding second safety feature warning threshold, it is considered that the current power battery to be tested has reached the warning level, and the sampling frequency of the acquisition module needs to be increased to shorten the prediction interval of the power battery to be tested, thereby accurately warning the power battery to be tested and eliminating safety hazards in a timely manner. At the same time, a low-level warning indication is output, that is, the yellow flash light in the corresponding area of ​​the power battery to be tested is turned on, and the sampling frequency is adjusted in time.

[0043] Case three: When the predicted value of the security feature reaches the third security feature warning threshold, a third target instruction of the security warning strategy set is generated. The third target instruction is used to control the collection frequency of the security feature data to be increased to a third frequency and to output an advanced warning indication, wherein the third frequency is greater than the second frequency, and the third security feature warning threshold is greater than the second security feature warning threshold.

[0044] It should be noted that when any data in the safety feature prediction value is equal to or greater than the corresponding third safety feature warning threshold, it is considered that the current power battery to be tested has reached the warning level, and the sampling frequency of the acquisition module needs to be increased to shorten the prediction interval of the power battery to be tested, thereby accurately warning the power battery to be tested and eliminating safety hazards in a timely manner. At the same time, an advanced warning indication is output, that is, the red flash light in the corresponding area of ​​the power battery to be tested is turned on, and the pre-set warning voice is played in a loop, and the sampling frequency is adjusted in time, and the power battery to be tested is ready to be removed from the warehouse at any time.

[0045] Among them, the method for obtaining the safety feature warning threshold is as follows: induce thermal runaway in the experimental power battery, record the safety feature value when the solid electrolyte interface (SEI) film inside the experimental power battery begins to decompose, and use the safety feature value as the safety feature warning threshold, wherein the model of the experimental power battery is consistent with that of the power battery to be tested, that is, the safety feature warning threshold measured in the experiment is also the safety feature warning threshold of the power battery to be tested.

[0046] In order to accurately warn the power batteries to be tested in storage and avoid potential safety hazards, the safety warning of the power batteries to be tested is set to three levels, and each level of safety warning corresponds to its own safety feature warning threshold. Specifically, the first safety feature warning threshold = C1*safety feature warning threshold, the second safety feature warning threshold = C2*safety feature warning threshold, and the third safety feature warning threshold = C3*safety feature warning threshold, where C1<C2<C3<1.

[0047] In an exemplary embodiment of the present application, obtaining the safety feature prediction model of the power battery to be tested in step S1 includes constructing and training the safety feature prediction model, and constructing and training the safety feature prediction model includes the following steps:

[0048] Step S11: inducing thermal runaway on the experimental power battery to obtain safety feature experimental data of the experimental power battery, and dividing the safety feature experimental data into training data and test data, wherein the model of the experimental power battery is consistent with that of the power battery to be tested.

[0049] The safety feature experimental data measured in the experiment are standardized. For example, there are k characteristic indicators (Y1, Y2, ..., Y k ), Y i ={x1,x2,…,x n}, the result after standardization is X1,X2,...,X k Specifically, according to the formula: Standardization is performed, where x ij is the actual value of the jth indicator on the i-th day, X ij is the standardized value of the j-th indicator on the i-th day.

[0050] The standardized multidimensional data is subjected to dimensionality reduction processing. The principal component analysis (PCA) is used to reduce the dimensionality of the experimental data. The preprocessed safety scores are arranged and combined into a matrix. The standardized matrix is ​​shown below:

[0051]

[0052] The covariance matrix is ​​obtained based on the standardized matrix. The covariance matrix is ​​as follows:

[0053]

[0054] in,

[0055] Calculate the eigenvalues ​​of R and arrange them in order of size, that is, λ1≥λ2≥…≥λ p ≥0, and The eigenvectors are:

[0056]

[0057] get When applied, the first, second, and third principal components corresponding to the top three eigenvalues ​​of the contribution rate are taken, and the i-th principal component: F i =a 1i X1+…+a pi X p .

[0058] Step S12: Construct an LSTM neural network model.

[0059] Step S13: Based on the training data, adjust the hyperparameters of the LSTM neural network model to obtain a first prediction model.

[0060] According to the standardized training data, the hyperparameters of the LSTM neural network model are adjusted, where the hyperparameters include: the number of hidden layer units, learning rate, time step, regularization coefficient and initial cell state.

[0061] Step S14: Based on the test data, verify the prediction accuracy of the first prediction model.

[0062] The prediction accuracy of the first prediction model is verified based on the standardized test data.

[0063] Step S15: When the prediction accuracy of the first prediction model meets the preset accuracy requirement, the first prediction model is used as the security feature prediction model.

[0064] Preferably, constructing and training the security feature prediction model further includes: optimizing the security feature prediction model in combination with the SCHO hyperbolic sine optimization algorithm.

[0065] Specifically, the LSTM hyperparameters are used as candidate solutions for the SCHO hyperbolic sine optimization algorithm and randomly initialized. The initialized candidate solutions are as follows:

[0066]

[0067] where the best candidate solution obtained during the iteration is considered to be near-optimal, X is the set of random candidate solutions obtained by using Eq. Xi,j represents the jth position of the ith solution, N represents the number of candidate solutions, and dim is the problem dimension.

[0068] The SCHO hyperbolic sine optimization algorithm is applied to optimize the hyperparameters of LSTM. The SCHO hyperbolic sine optimization algorithm includes an exploration phase and a development phase.

[0069] Exploration is divided into two phases in the optimization process (iteration) and should exist in later iterations to escape the local optimum. The value of switching between these two phases is determined using the formula, which is as follows:

[0070]

[0071] Among them, Max iteration represents the maximum value of iteration, floor is the rounding down function in MATLAB, and ct is the coefficient used to set the two-phase switching point, which is fixed to 3.6 here. In the first stage, the space near the search agent position will be explored in the early iterations, and then the search agent will gradually approach the optimal solution. The position update function of the exploration in the first stage is proposed as:

[0072]

[0073] Where t represents the current iteration, and denote the j-th position of the ith solution in the current iteration and the next iteration respectively. is the jth position of the optimal solution obtained so far, r1 / r2 is a random number in the interval [0, 1]. W1 is the number in the first exploration phase. The weight coefficient controls the candidate solution in the first stage to move away from itself and gradually explore the optimal solution. It uses the following formula:

[0074] W1=r3×a1×(coshr4+u×sinhr4-1).

[0075] where a1 is a monotonically decreasing function, which is obtained by using Calculation. Where m is the sensitivity coefficient that controls the exploration accuracy, and here it is taken as m = 0.45. r3 / r4 is a random number in [0, 1]. u is the sensitivity coefficient that controls the exploration accuracy in the first stage, and is fixed at 0.388. The candidate solution gradually moves away from itself in the first stage, and then explores the optimal solution:

[0076] In the second phase of exploration, the search agents are almost unaffected by the best solution, so they explore the next position without direction based on the current position. The position update function uses the following formula:

[0077]

[0078] Among them, ε represents a small positive number, which is taken as 0.003. W2 is the second stage of exploration The weight coefficient is calculated by using W2=r6×a2, where r6 is a random number in [0, 1] and a2 represents the weight coefficient by using The monotonically decreasing function calculated by the equation is: n represents the sensitivity coefficient that controls the detection accuracy in the second stage of exploration and is set to 0.5. Multiplying W2 by ε greatly weakens the influence of the optimal solution on the current solution, which will lead to Perform undirected random exploration of candidate solutions around it.

[0079] Determine the switching between the two phases in the SCHO optimization algorithm, that is, as the iteration progresses, it will gradually enter the second phase, the development phase. SCHO uses the switching method between "Sinh" and "Cosh" to switch between exploration and development, which is specified by the following formula, which is as follows:

[0080]

[0081] Among them, r 13 A is a random number in the interval [0, 1], and p and q are the balance coefficients used to control exploration and exploitation during iterations, which are equal to 10 and 9, respectively. When A>1, SCHO performs the exploration phase, and when A<1, SCHO starts to perform the exploitation phase.

[0082] The development phase is divided into two stages and is performed throughout the iteration. In the first local exploration phase, the neighboring space of X is explored, so the development formula is designed as:

[0083]

[0084] Among them, r7 / r8 are random numbers in the interval [0, 1]. W3 is the weight coefficient used in the first stage, and controls the candidate solution to use the surrounding search space from near to far, using W3 = r9 × a1 × (cosh r 10 +u×sinh r 10 ), where r9 / r10 are random numbers in [0, 1], a1 and u are the same as in the first global exploration phase, and u is fixed to 0.388. At this point, the candidate solution can use the space around it from near to far.

[0085] For the second stage of development, the candidate solution will be developed as close as possible to the optimal solution obtained so far, and the intensity of the optimal solution development will increase with the number of iterations. The position update function is:

[0086]

[0087] Where r11 / r12 are random numbers in the interval [0, 1] and W2 is calculated using the previous equation. W2 controls the degree of the exploitation phase of the local optimization. Its absolute value gradually increases in subsequent iterations, increasing the degree of exploitation. The coefficient in front of the absolute value in the above equation is used to maintain the diversity of candidate solutions.

[0088] In order to fully explore and utilize the potential search space, SCHO adopts a strategy similar to animal hunting in the later stage, called bounded search strategy, that is, to discover the potential search space by exploring the entire search area in early iterations. In order to fully explore and utilize the potential space, all candidate solutions are randomly generated in this potential space using Eq2, so that the space can be deeply explored and utilized. The strategy starts by using: To be determined.

[0089] Where k is a positive integer starting from 1, BS k Using the equation Calculate, where β controls the value at which the bounded search strategy is started and is set to 1.55, BS k:1 They represent the number of iterations for starting the current and next bounded search strategies, respectively. α is a sensitivity coefficient that controls the accuracy of deep exploration and development of the latent space. Here, α is taken as 4.6.

[0090] When using this optimizer, the upper and lower bounds are determined by the following formula:

[0091]

[0092] Among them, ub k and lb k denotes the upper and lower bounds of the potential search space, represents the jth position of the suboptimal solution.

[0093] Determine whether the SCHO hyperbolic sine optimization algorithm has iterated to the maximum number of times or has been updated to the optimal position. If one of the conditions is met, output it to obtain the optimal solution for each hyperparameter. Otherwise, increase the number of iterations by one until the global optimal position is output.

[0094] The parameter configuration optimized by the SCHO algorithm is used for the training of the LSTM model. The LSTM unit state update equation is as follows, including the input gate i t , forget gate f, output gate o t .i t =σ(W i ·[h t；1 ,x t ]+b i ), f t =σ(Wf ·[h t；1 ,x t ]+b f ), o t =σ(W o ·[h t；1 ,x t ]+b o ).

[0095] Among them, x t represents the input data at time step t. This is the feature vector input to the LSTM unit at the current moment. t；1 is the hidden state at time step t-1. The hidden state is the short-term memory of LSTM, which saves the key information of the previous moment as one of the inputs of the current moment. i is the weight matrix of the input gate, W f is the weight matrix of the forget gate, W o is the weight matrix of the output gate, the candidate cell state and the cell state update. The calculation formula of the cell state is as follows:

[0096]

[0097] Among them, the candidate cell state is the new information at the current moment, and its value will update the cell state under the action of the input gate.

[0098] The update formula for cell state and hidden state is:

[0099]

[0100] Among them, i t It is the activation value of the input gate, which determines the introduction of new information. is the candidate cell state, indicating the candidate information currently input, and the calculation formula is f t is the activation value of the forget gate, which controls the degree of retention of the cell state. t；1 is the cell state at the previous time step.

[0101] The cell state is the long-term memory of LSTM, which is determined by the cell state at the previous moment and the candidate state at the current moment.

[0102] The hidden state is the output of the LSTM and is used for the next time step or output layer. Its calculation formula is:

[0103] h t =o t ⊙tanh(C t ).

[0104] Among them, the output gate controls which parts of the current cell state will be output. The cell state is converted into a hidden state through a function. t Is the activation value of the output gate, which controls the output degree of the current hidden state. tanh(C t ) is the result of the cell state after passing through the hyperbolic tangent function, which limits the range of the output.

[0105] In each round of optimization, the prediction error of the LSTM model is evaluated according to the loss function, using the mean relative error, mean absolute error, and minimum mean square error, using the following formula.

[0106]

[0107] MRE is the mean relative error, MAE is the mean absolute error, MSE is the minimum mean square error, and N is the number of samples. t is the true value, is the predicted value.

[0108] During the optimization process, SCHO selects the parameter combination with better performance based on the fitness evaluation value and minimizes the loss of the model.

[0109] In an exemplary embodiment of the present application, the safety warning method further includes the following steps:

[0110] Obtain the detection concentration of harmful gases in the area where the power battery to be tested is located;

[0111] Comparing the detected concentration with the concentration threshold to obtain a second comparison result;

[0112] Based on the second comparison result, a warning prompt is output.

[0113] Specifically, a gas detector is installed in the area where the power battery to be tested is located. When the concentration of harmful gases in the air exceeds the set concentration threshold, the detector's built-in light alarm is directly turned on. When harmful gases appear, it proves that the battery has begun to be out of control. Therefore, when the gas detector detects that the harmful gas exceeds the threshold, it directly turns on the highest level warning.

[0114] According to another specific embodiment of the present application, a safety warning system for power battery out of control under storage conditions is also provided. Figure 3 As shown, the safety warning system includes: an acquisition module, a collection module, a first generation module, a comparison module and a second generation module.

[0115] The acquisition module is used to obtain the safety feature prediction model of the power battery to be tested.

[0116] The acquisition module is used to collect safety feature data of the power battery to be tested, wherein the safety feature data at least includes: warehouse temperature, warehouse humidity, battery internal resistance, battery current, battery voltage and battery temperature.

[0117] The first generating module is used to input the safety feature data into the safety feature prediction model to obtain the safety feature prediction value of the power battery to be tested after a preset time period.

[0118] The comparison module is used to compare the safety feature prediction value with the safety feature warning threshold to obtain a first comparison result.

[0119] The second generating module is used to generate a safety warning strategy set for the power battery to be tested based on the first comparison result, and the safety warning strategy set at least includes: increasing the collection frequency of safety feature data and outputting a warning indication.

[0120] In an embodiment of the present application, a safety feature prediction model for a power battery to be tested is constructed, safety feature data of the power battery to be tested is collected, the safety feature data is input into the safety feature prediction model, the safety feature prediction model is used to predict the changes in the safety features of the power battery to be tested in the short term in the future, and a safety feature prediction value is obtained, the safety feature prediction value is compared with the safety feature warning threshold, a first comparison result is obtained, and a corresponding safety warning strategy is generated according to the first comparison result. In the above scheme, the safety features of the power battery to be tested are predicted in the short term through the safety feature prediction model, and the safety warning of the power battery to be tested in the storage state is given according to the prediction result, thereby realizing the safe storage of the power battery.

[0121] According to another specific embodiment of the present application, a computer storage medium is also provided, and the computer storage medium includes a stored program, wherein when the program is running, the device where the computer storage medium is located is controlled to execute the above-mentioned security warning method.

[0122] In this embodiment, the storage medium may be configured to store a computer program for performing the following steps:

[0123] Step S1: Obtain a safety feature prediction model for the power battery to be tested.

[0124] Step S2: collecting safety characteristic data of the power battery to be tested, wherein the safety characteristic data at least includes: warehouse temperature, warehouse humidity, battery internal resistance, battery current, battery voltage and battery temperature.

[0125] Step S3: inputting the safety feature data into a safety feature prediction model to obtain a safety feature prediction value of the power battery to be tested after a preset time period.

[0126] Step S4: Compare the safety feature prediction value with the safety feature warning threshold to obtain a first comparison result.

[0127] Step S5: Based on the first comparison result, a safety warning strategy set for the power battery to be tested is generated, where the safety warning strategy set at least includes: increasing the frequency of collecting safety feature data and outputting a warning indication.

[0128] In an embodiment of the present application, a safety feature prediction model for a power battery to be tested is constructed, safety feature data of the power battery to be tested is collected, the safety feature data is input into the safety feature prediction model, the safety feature prediction model is used to predict the changes in the safety features of the power battery to be tested in the short term in the future, and a safety feature prediction value is obtained, the safety feature prediction value is compared with the safety feature warning threshold, a first comparison result is obtained, and a corresponding safety warning strategy is generated according to the first comparison result. In the above scheme, the safety features of the power battery to be tested are predicted in the short term through the safety feature prediction model, and the safety warning of the power battery to be tested in the storage state is given according to the prediction result, thereby realizing the safe storage of the power battery.

[0129] According to another specific embodiment of the present application, a computer program product is also provided, including a computer program, and the computer program is executed by a processor to perform the above-mentioned security warning method.

[0130] In this embodiment, the processor may be configured to perform the following steps through a computer program:

[0131] Step S1: Obtain a safety feature prediction model for the power battery to be tested.

[0132] Step S2: collecting safety characteristic data of the power battery to be tested, wherein the safety characteristic data at least includes: warehouse temperature, warehouse humidity, battery internal resistance, battery current, battery voltage and battery temperature.

[0133] Step S3: inputting the safety feature data into a safety feature prediction model to obtain a safety feature prediction value of the power battery to be tested after a preset time period.

[0134] Step S4: Compare the safety feature prediction value with the safety feature warning threshold to obtain a first comparison result.

[0135] Step S5: Based on the first comparison result, a safety warning strategy set for the power battery to be tested is generated, where the safety warning strategy set at least includes: increasing the frequency of collecting safety feature data and outputting a warning indication.

[0136] In an embodiment of the present application, a safety feature prediction model for a power battery to be tested is constructed, safety feature data of the power battery to be tested is collected, the safety feature data is input into the safety feature prediction model, the safety feature prediction model is used to predict the changes in the safety features of the power battery to be tested in the short term in the future, and a safety feature prediction value is obtained, the safety feature prediction value is compared with the safety feature warning threshold, a first comparison result is obtained, and a corresponding safety warning strategy is generated according to the first comparison result. In the above scheme, the safety features of the power battery to be tested are predicted in the short term through the safety feature prediction model, and the safety warning of the power battery to be tested in the storage state is given according to the prediction result, thereby realizing the safe storage of the power battery.

[0137] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0138] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0139] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0140] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0141] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0142] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0143] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A safety warning method for power battery out of control under storage conditions, characterized in that: include: Obtain a safety feature prediction model for the power battery to be tested; Collecting safety feature data of the power battery to be tested, wherein the safety feature data at least includes: warehouse temperature, warehouse humidity, battery internal resistance, battery current, battery voltage and battery temperature; Inputting the safety feature data into the safety feature prediction model to obtain a safety feature prediction value of the power battery to be tested after a preset time period; Comparing the safety feature prediction value with the safety feature warning threshold to obtain a first comparison result; Based on the first comparison result, a safety warning strategy set for the power battery to be tested is generated, and the safety warning strategy set at least includes: increasing the collection frequency of the safety feature data and outputting a warning indication.

2. The safety warning method according to claim 1, characterized in that: Based on the first comparison result, a safety warning strategy set for the power battery to be tested is generated, including: When the security feature prediction value reaches a first security feature warning threshold, a first target instruction of the security warning strategy set is generated, and the first target instruction is used to control the collection frequency of the security feature data to be increased to a first frequency.

3. The safety warning method according to claim 2, characterized in that: Generating a safety warning strategy set for the power battery to be tested based on the first comparison result also includes: When the predicted value of the security feature reaches a second security feature warning threshold, a second target instruction of the security warning strategy set is generated, and the second target instruction is used to control the collection frequency of the security feature data to be increased to a second frequency and output a low-level warning indication, wherein the second frequency is greater than the first frequency, and the second security feature warning threshold is greater than the first security feature warning threshold.

4. The safety warning method according to claim 3, characterized in that: Generating a safety warning strategy set for the power battery to be tested based on the first comparison result also includes: When the predicted value of the security feature reaches the third security feature warning threshold, a third target instruction of the security warning strategy set is generated, and the third target instruction is used to control the collection frequency of the security feature data to be increased to a third frequency and output an advanced warning indication, wherein the third frequency is greater than the second frequency, and the third security feature warning threshold is greater than the second security feature warning threshold.

5. The safety warning method according to claim 1, characterized in that: Acquiring the safety feature prediction model of the power battery to be tested includes constructing and training the safety feature prediction model, and constructing and training the safety feature prediction model includes: Inducing thermal runaway on an experimental power battery to obtain safety feature experimental data of the experimental power battery, and dividing the safety feature experimental data into training data and test data, wherein the experimental power battery is of the same model as the power battery to be tested; Build an LSTM neural network model; Based on the training data, adjusting the hyperparameters of the LSTM neural network model to obtain a first prediction model; Based on the test data, verifying the prediction accuracy of the first prediction model; When the prediction accuracy of the first prediction model meets the preset accuracy requirement, the first prediction model is used as the security feature prediction model.

6. The safety warning method according to claim 5, characterized in that: Building and training the security feature prediction model also includes: The safety feature prediction model is optimized by combining the SCHO hyperbolic sine optimization algorithm.

7. The safety early warning method according to claim 1, characterized in that: The safety early warning method also includes: Obtaining the detection concentration of harmful gases in the area where the power battery to be tested is located; Comparing the detected concentration with a concentration threshold to obtain a second comparison result; Based on the second comparison result, a warning prompt is output.

8. A safety warning system for power battery runaway under storage conditions, characterized in that: include: An acquisition module, the acquisition module is used to acquire a safety feature prediction model of the power battery to be tested; A collection module, the collection module is used to collect safety feature data of the power battery to be tested, wherein the safety feature data at least includes: warehouse temperature, warehouse humidity, battery internal resistance, battery current, battery voltage and battery temperature; A first generating module, the first generating module is used to input the safety feature data into the safety feature prediction model to obtain a safety feature prediction value of the power battery to be tested after a preset time period; A comparison module, the comparison module is used to compare the security feature prediction value with the security feature warning threshold to obtain a first comparison result; A second generation module, the second generation module is used to generate a safety warning strategy set for the power battery to be tested based on the first comparison result, and the safety warning strategy set at least includes: increasing the collection frequency of the safety feature data and outputting a warning indication.

9. A computer storage medium, characterized in that: The computer storage medium includes a stored program, wherein when the program is running, the device where the computer storage medium is located is controlled to execute the safety warning method described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that The computer program is executed by a processor to implement the safety warning method described in any one of claims 1-7.

Images