Battery swap station abnormity prediction method and system, computer equipment and storage medium
By using Kepler optimization algorithm to optimize the convolutional neural network model in the abnormal prediction of battery swap stations, combined with real-time data processing and feature extraction, the convolutional neural network model has low data processing efficiency and poor prediction accuracy in the abnormal prediction prediction of battery swap stations. The problem is to realize accurate monitoring and timely warning of the operating status of the power swap station, reducing operating costs.
Patent Information
- Application Number
- CN202411225805.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-03
- Publication Date
- 2024-12-20
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The convolutional neural network model has problems such as low data processing efficiency and poor prediction accuracy in the prediction of abnormality of battery swap stations, which is difficult to warn in time.
The Kepler optimization algorithm is used to optimize the initial convolutional neural network model, combined with real-time running data for preprocessing and feature extraction, used convolutional neural network for abnormal prediction, and generated detailed exception reports.
It improves the accuracy of abnormal prediction of battery swap stations and data processing efficiency, realizes accurate monitoring and timely warning of the operating status of battery swap stations, reduces operation costs, and improves service quality.
Smart Images
Figure CN119167256A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of battery swap stations, and in particular to a method, system, computer equipment and storage medium for predicting abnormality in a battery swap station. Background Art
[0002] With the development of technology, battery swap stations have been widely used; in order to ensure the normal operation of battery swap stations, it is necessary to predict abnormalities of battery swap stations. With the development of deep learning technology, especially the successful application of convolutional neural network models (CNN) in image processing and time series data analysis, more and more researchers have begun to explore its potential in predicting abnormalities of battery swap stations. However, the performance of convolutional neural network models is highly dependent on the selection of hyperparameters. The current convolutional neural network models generally have low data processing efficiency and difficulty in finding the global optimal solution, which leads to low prediction accuracy when predicting abnormalities of battery swap stations, and failure to provide timely warnings when abnormalities occur in battery swap stations. Summary of the invention
[0003] Based on this, it is necessary to provide a battery swap station abnormality prediction method, system, computer equipment and storage medium that can improve prediction accuracy, enhance data processing efficiency, and accurately monitor and promptly warn of battery swap station abnormalities in response to the above-mentioned technical problems.
[0004] In a first aspect, the present invention provides a method for predicting abnormality of a battery swap station, the method comprising:
[0005] Obtain real-time operation data of the battery swap station;
[0006] Preprocessing the real-time operation data;
[0007] Build an initial convolutional neural network model;
[0008] Optimizing the initial convolutional neural network model using a Kepler optimization algorithm to obtain a convolutional neural network model;
[0009] Inputting the preprocessed real-time operation data into the convolutional neural network to perform abnormal prediction on the battery swap station;
[0010] Generate exception reports based on prediction results.
[0011] Optionally, the using a Kepler optimization algorithm to optimize the initial convolutional neural network model to obtain a convolutional neural network model includes:
[0012] Determining a set of hyperparameters to be optimized for the initial convolutional neural network model;
[0013] Based on the hyperparameter set, generating a hyperparameter combination;
[0014] Obtaining the fitness of the initial convolutional neural network model according to the hyperparameter combination;
[0015] Acquire a running body of the hyperparameter combination, wherein the running body includes corresponding running data, and the running data includes position information, speed information, acceleration information, and quality information of the running body;
[0016] Performing planetary motion simulation update on the operating data of the operating body corresponding to the hyperparameter combination to update the parameters of the initial convolutional neural network model;
[0017] Iterative optimization is performed until the fitness converges or reaches a preset number of iterations, thereby obtaining the convolutional neural network model.
[0018] Optionally, performing planetary motion simulation update on the operating data of the operating body corresponding to the hyperparameter combination to update the parameters of the initial convolutional neural network model includes:
[0019] Performing planetary motion simulation based on the operating data of the operating body corresponding to the hyperparameter combination, and updating the operating data;
[0020] The corresponding hyperparameter combination is updated based on the updated operating data, and the parameters to be optimized corresponding to the updated hyperparameter combination are used as the parameters of the initial convolutional neural network model to complete the parameter update of the initial convolutional neural network model.
[0021] Optionally, the performing planetary motion simulation based on the operating data of the operating body corresponding to the hyperparameter combination and updating the operating data includes:
[0022] The position information of the running body corresponding to the hyperparameter combination is updated based on the following formula: Among them, p i is the position information of the running body before updating, p i+1 is the updated position information of the running body, v i is the current speed information of the running body, a i is the current acceleration information of the running body, △t is the time step, and i is an integer greater than or equal to 1;
[0023] The acceleration information of the running body corresponding to the hyperparameter combination is updated based on the following formula: Among them, a j is the updated acceleration information of the running body, G is the gravitational constant, m i is the current quality information of the running body, p iis the position information of the running body before updating, p j is the updated position information of the running body, i is an integer greater than or equal to 1, and j is an integer greater than i;
[0024] The speed information of the running body corresponding to the hyperparameter combination is updated based on the following formula: i+1 =v i +a i .△t, where V i+1 is the updated speed of the running body, v i is the current speed of the running body, a i is the current acceleration information of the running body, and △t is the time step.
[0025] Optionally, after inputting the preprocessed real-time operation data into the convolutional neural network and performing abnormal prediction on the battery swap station, and before generating an abnormal report based on the prediction result, the method further includes:
[0026] Determining whether the battery swap station has an abnormality based on the prediction result;
[0027] If so, the abnormal type of the battery swap station is identified.
[0028] Optionally, the abnormality report includes: the time when the abnormality occurred, the location where the abnormality occurred, the type of abnormality, the abnormal operating parameters, the severity of the abnormality, the cause of the abnormality and the treatment measures for the abnormality.
[0029] Optionally, after generating an abnormality report based on the prediction result, the method further includes:
[0030] Visually display the prediction results and the abnormal report;
[0031] An alarm message is issued based on the abnormal report.
[0032] In a second aspect, the present invention further provides a battery swap station abnormality prediction system, the battery swap station abnormality prediction system comprising:
[0033] An acquisition module is used to obtain real-time operation data of the battery swap station;
[0034] A preprocessing module, used for preprocessing the real-time operation data;
[0035] Model building module, used to build the initial convolutional neural network model;
[0036] A model optimization module, used to optimize the initial convolutional neural network model using a Kepler optimization algorithm to obtain a convolutional neural network model; the convolutional neural network model is used to predict abnormalities of the battery swap station based on the preprocessed real-time operation data;
[0037] The report generation module is used to generate anomaly reports based on the prediction results.
[0038] In a third aspect, the present invention further provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and wherein the processor implements the steps of the battery swap station abnormality prediction method as described in the first aspect when executing the computer program.
[0039] In a fourth aspect, the present invention further provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the steps of the battery swap station abnormality prediction method as described in the first aspect are implemented.
[0040] The above-mentioned battery swap station abnormality prediction method, system, computer equipment and storage medium of the present invention, by combining the powerful feature extraction capability of the convolutional neural network model and the efficient hyperparameter optimization capability of the Kepler optimization algorithm, can not only improve the prediction accuracy of battery swap station abnormalities, but also significantly improve data processing efficiency, thereby enabling accurate monitoring and timely warning of the operating status of the battery swap station, reducing operating costs and improving service quality. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] Figure 1 A flow chart of a method for predicting abnormality of a battery swap station is provided in one embodiment of the present invention.
[0042] Figure 2 It is a structural block diagram of a battery swap station abnormality prediction system provided in another embodiment of the present invention.
[0043] Figure 3 This is a diagram of the internal structure of a computer device provided in yet another embodiment of the present invention. DETAILED DESCRIPTION
[0044] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0045] In one embodiment, see Figure 1 The present invention provides a method for predicting abnormality of a battery swap station, and the method for predicting abnormality of a battery swap station includes the following steps: S10 to S60.
[0046] S10: Obtain real-time operation data of the battery swap station.
[0047] S20: Preprocessing the real-time operation data.
[0048] S30: Build an initial convolutional neural network model.
[0049] S40: Use the Kepler optimization algorithm to optimize the initial convolutional neural network model to obtain a convolutional neural network model.
[0050] S50: Inputting the preprocessed real-time operation data into the convolutional neural network to perform abnormal prediction on the battery swap station.
[0051] S60: Generate an exception report based on the prediction result.
[0052] In the battery swap station abnormality prediction method of the present invention, by combining the powerful feature extraction capability of the convolutional neural network model and the efficient hyperparameter optimization capability of the Kepler optimization algorithm, not only the prediction accuracy of the battery swap station abnormality can be improved, but also the data processing efficiency can be significantly improved, thereby enabling accurate monitoring and timely warning of the operating status of the battery swap station, reducing operating costs and improving service quality.
[0053] In step S10, refer to Figure 1 In step S10, real-time operation data of the battery swap station is obtained.
[0054] As an example, the real-time operation data of the battery swap station can be obtained through a variety of sensors; specifically, a voltage sensor can be installed on the main power supply line of the battery swap station to monitor the voltage changes in real time; a current sensor can be arranged in the current loop of the battery swap station to monitor the real-time changes of the current; a temperature sensor can be installed in the key equipment and environment of the battery swap station to record the temperature data in real time; according to actual needs, humidity sensors, vibration sensors, etc. can also be installed to monitor other environmental parameters such as humidity and vibration.
[0055] As an example, after acquiring the real-time data collected by each sensor, the real-time data can also be converted into a digital signal through a data collector, and filtered and denoised. The processed data can be transmitted through a data transmission module. Specifically, limited transmission can be performed through Ethernet, or wireless transmission can be performed using WiFi. By configuring the data transmission module, it can be ensured that the data can be transmitted stably and quickly.
[0056] As an example, the sampling rate of each sensor and data logger should be high enough to capture subtle changes in parameters such as voltage, current, and temperature. For example, the sampling rate of voltage and current can be set to thousands of times per second, and the sampling rate of temperature sensors can be set to tens of times per second.
[0057] As an example, the data collected by each sensor needs to be time synchronized to ensure that the data collected by different sensors can be analyzed at the same time point. GPS clock or network time protocol (NTP) can be used for synchronization.
[0058] As an example, in order to ensure the accuracy of the collected data, each sensor needs to be calibrated before installation, and performance testing and calibration updates must be performed regularly.
[0059] In step S20, refer to Figure 1 In S20, the real-time operation data is preprocessed.
[0060] As an example, in step S20, preprocessing the real-time operation data may include the following steps: S201 to S204.
[0061] S201: Clean the collected real-time operation data. Specifically, a filtering algorithm may be used to remove noise from the real-time operation data. For example, a moving average filter may be used to smooth the data. The window width of the smoothing process may be selected according to actual needs, such as 3, 4, or 5. The cleaned data may be processed for missing values. Specifically, an interpolation method or a filling method may be used to process missing values in the data. For example, a linear interpolation method may be used to process missing values in the data.
[0062] S202: Perform data normalization on the cleaned data. Specifically, the cleaned data may be scaled to a uniform range (eg, 0 to 1) to eliminate dimensional differences between different data. Commonly used data normalization methods may include Min-Max normalization.
[0063] S203: Extract features from the normalized data. Specifically, it may include time domain feature extraction and frequency domain feature extraction. The time domain feature extraction is to extract useful time domain features from time series data, such as mean, variance, peak value and skewness, etc. The data may be subjected to Fourier transformation to extract frequency domain features, such as main frequency and spectrum energy, etc.
[0064] S204: Convert the format of the extracted features. Specifically, the extracted features may be reshaped. More specifically, the extracted features may be reshaped into a format suitable for input into a convolutional neural network model; for example, one-dimensional time series data may be reshaped into two-dimensional matrix data to facilitate convolution operations.
[0065] In step S30, refer to Figure 1 In step S30, an initial convolutional neural network model is constructed.
[0066] As an example, the initial convolutional neural network model plays a core role in the abnormal prediction of the battery swap station, and the processed real-time operation data is used for abnormal prediction through a multi-layer convolutional neural network. The initial convolutional neural network model can include an input layer, several convolutional layers, a pooling layer, a fully connected layer, and an output layer.
[0067] As an example, the input layer may receive preprocessed running data, which is usually in the form of a two-dimensional matrix, such as [N,M], where N represents the time step or number of samples of the data, and M represents the number of features of each sample.
[0068] As an example, the number of the convolutional layers can be multiple, and the convolutional layers are used to extract features of the input data; each of the convolutional layers can be followed by an activation function to introduce nonlinearity and enhance the expressiveness of the convolutional neural network model. The convolutional layer can perform a convolution operation on the input data through a convolution kernel (which can be a filter) to extract local features. The activation function can be a nonlinear activation function, such as ReLU (Rectified Linear Unit).
[0069] As an example, the pooling layer can reduce the size of the feature map through maximum pooling or average pooling operations, retaining important features while reducing computational complexity. The formula for the maximum pooling layer can be:
[0070]
[0071] Among them, h i，j is the feature value of position (i, j) in the feature map after pooling, X i+m,j+n is the feature value of position (i+m, j+n) in the input feature map, It means to traverse all elements in the pooling window and take the maximum value, m represents the offset of the pooling window in the height direction, and n represents the offset of the pooling window in the width direction.
[0072] As an example, the fully connected layer can flatten the pooled feature map into a one-dimensional vector and integrate the features through a fully connected operation.
[0073] As an example, the output layer outputs a prediction structure according to the requirements of a specific task. For classification tasks, the output is converted into a probability distribution by using a Softmax activation function.
[0074] In one example, taking the data of a battery swap station as an example, the preprocessed operation data input by the input layer is assumed to be in the form of [100, 10], where 100 is the time step and 10 is the number of features. The first convolution layer in the convolution layer uses a 3*3 convolution kernel to extract local features; assuming that the number of convolution kernels is 32, the output shape is [98, 8, 32]; the ReLU activation function is applied to the convolution feature map to set negative values to zero. The first pooling layer in the pooling layer uses a 2*2 maximum pooling operation to reduce the size of the feature map by half, and the output shape is [49, 4, 32]. The fully connected layer flattens the pooled feature map into a one-dimensional vector with a shape of [49, 4, 32]; by integrating features through the fully connected layer, the output dimension can be 128. The output layer uses the Softmax activation function, assuming that there are 3 types of predictions, and the output shape is [3], indicating the prediction probability of each class.
[0075] In step S40, refer to Figure 1 In step S40, the initial convolutional neural network model is optimized using the Kepler optimization algorithm to obtain a convolutional neural network model.
[0076] As an example, in step S40, the use of the Kepler optimization algorithm to optimize the initial convolutional neural network model to obtain a convolutional neural network model may include the following steps: S401 to S406.
[0077] S401: Determine a set of hyperparameters to be optimized for the initial convolutional neural network model.
[0078] S402: Generate a hyperparameter combination based on the hyperparameter set.
[0079] S403: Obtaining the fitness of the initial convolutional neural network model according to the hyperparameter combination.
[0080] S404: Acquire a running body of the hyperparameter combination, wherein the running body includes corresponding running data, and the running data includes position information, speed information, acceleration information, and quality information of the running body.
[0081] S405: Perform planetary motion simulation update on the operating data of the operating body corresponding to the hyperparameter combination to update the parameters of the initial convolutional neural network model.
[0082] S406: Perform iterative optimization until the fitness converges or reaches a preset number of iterations, thereby obtaining the convolutional neural network model.
[0083] As an example, in step S401, the hyperparameter set may include a learning rate, a convolution kernel size, and a number of convolution layers, and may also include the number of convolution kernels, a pooling window size, a batch size, and the like.
[0084] As an example, in step S402, a group of hyperparameter combinations may be randomly generated based on the hyperparameter set, and each hyperparameter combination may be regarded as a "planet", and its position represents the value of the current hyperparameter.
[0085] As an example, in step S403, the initial convolutional neural network model can be trained for each of the hyperparameter combinations and its performance indicators, such as accuracy and recall, can be obtained; the performance indicators can be regarded as the fitness of the planet. The fitness function F can be:
[0086] F=α.Accuracy+β.Recall-γ.Complexity
[0087] Among them, α, β and γ are weight coefficients, which adjust the relative importance of each indicator; Accuracy represents accuracy, Recall represents recall rate, and Complexity represents complexity.
[0088] As an example, in step S405, performing planetary motion simulation update on the operating data of the operating body corresponding to the hyperparameter combination to update the parameters of the initial convolutional neural network model may include the following steps: S4051 to S4052.
[0089] S4051: Perform planetary motion simulation based on the operating data of the operating body corresponding to the hyperparameter combination, and update the operating data.
[0090] S4052: Update the corresponding hyperparameter combination based on the updated operating data, and use the parameters to be optimized corresponding to the updated hyperparameter combination as the parameters of the initial convolutional neural network model to complete the parameter update of the initial convolutional neural network model.
[0091] As an example, in step S4051, performing planetary motion simulation based on the operating data of the operating body corresponding to the hyperparameter combination and updating the operating data may include the following contents:
[0092] According to Kepler's laws of planetary motion, each of the hyperparameter combinations moves in the search space and updates its position; specifically, the position information of the running body corresponding to the hyperparameter combination can be updated based on the following formula: Among them, p i is the position information of the running body before updating, p i+1 is the updated position information of the running body, v i is the current speed information of the running body, a i is the current acceleration information of the running body, △t is the time step, and i is an integer greater than or equal to 1;
[0093] The acceleration information of the running body corresponding to the hyperparameter combination is updated based on the following formula: Among them, a j is the updated acceleration information of the running body, G is the gravitational constant, m i is the current quality information of the running body, p i is the position information of the running body before updating, p j is the updated position information of the running body, i is an integer greater than or equal to 1, and j is an integer greater than i;
[0094] The speed information of the running body corresponding to the hyperparameter combination is updated based on the following formula: i+1 =v i +a i .△t, where V i+1 is the updated speed of the running body, v i is the current speed of the running body, a i is the current acceleration information of the running body, and △t is the time step.
[0095] In step S50, refer to Figure 1 In step S50, the preprocessed real-time operation data is input into the convolutional neural network to perform abnormal prediction on the battery swap station.
[0096] As an example, after inputting the preprocessed real-time operation data into the convolutional neural network and performing abnormal prediction on the battery swap station, that is, after step S50, the following steps may also be included:
[0097] Determining whether the battery swap station has an abnormality based on the prediction result;
[0098] If so, the abnormal type of the battery swap station is identified.
[0099] As an example, prediction results may be received from the convolutional neural network, which prediction results are typically classification results regarding the operating status of the battery swap station, such as “normal” or “abnormal”.
[0100] As an example, the prediction results may be parsed to extract specific prediction information and corresponding timestamps.
[0101] As an example, if it is determined that the battery swap station has an abnormality, the abnormality type may be further classified, such as voltage abnormality, current abnormality, temperature abnormality, and so on.
[0102] In step S60, refer to Figure 1 In step S60, an exception report is generated based on the prediction result.
[0103] As an example, the exception report may include: the time when the exception occurred, the location where the exception occurred, the type of exception, the abnormal operating parameters (for example, voltage, current or temperature, etc.), the severity of the exception (minor, moderate or severe), the cause of the exception and the measures to deal with the exception.
[0104] As an example, after generating the abnormality report based on the prediction result, that is, after step S60, the following steps are also included: S70 to S80.
[0105] S70: Visually display the prediction results and the abnormality report.
[0106] S80: issuing an alarm message based on the abnormality report.
[0107] As an example, in step S70, the prediction results and the anomalies can be visualized in the form of charts or dashboards, etc., so that the operation and maintenance personnel can intuitively understand. Common visualization charts may include line charts, bar charts or pie charts, etc., to show the operation trend and abnormal conditions of the battery swap station.
[0108] As an example, in step S80, the alarm information can be sent by SMS, email or APP push to ensure timely response. The alarm information should include key information of the abnormal report so that the operation and maintenance personnel can quickly understand the situation.
[0109] As an example, in the abnormality prediction method of the battery swap station in the present invention, the convolutional neural network model can have the following advantages:
[0110] 1. Strong feature extraction capability: When processing high-dimensional data such as images and videos, the convolutional neural network model can automatically extract complex features without the need for manual feature design.
[0111] 2. Spatial invariance: Through convolution operations, the convolutional neural network model can effectively capture local spatial correlations and is suitable for data of various spatial structures.
[0112] 3. Parameter sharing: The parameter sharing mechanism of the convolutional layer greatly reduces the model parameters, improves the training efficiency and the generalization ability of the model.
[0113] The Kepler optimization algorithm may have the following advantages:
[0114] 1. Global optimization capability: Kepler optimization algorithm is a global optimization algorithm that can effectively avoid local optimality and find a better solution.
[0115] 2. Strong adaptability: The algorithm can adapt to various complex optimization problems, including nonlinear, multi-peak, multi-constraint and other problems.
[0116] 3. Efficiency: Kepler optimization algorithm performs well in dealing with large-scale optimization problems and can find the best solution in a short time.
[0117] Combining the convolutional neural network model with the Kepler optimization algorithm can have the following advantages:
[0118] 1. Improve model performance: Optimizing the hyperparameters of the convolutional neural network model (such as learning rate, convolution kernel size, number of layers, etc.) through the Kepler optimization algorithm can significantly improve the performance of the convolutional neural network model.
[0119] 2. Accelerate the training process: Kepler optimization algorithm can find the optimal training parameters, reduce training time and improve training efficiency.
[0120] 3. Enhance model robustness: Combined with the Kepler optimization algorithm, the convolutional neural network model can better cope with different types of data and tasks, improving the robustness and generalization ability of the model.
[0121] 4. Automated hyperparameter tuning: The Kepler optimization algorithm can automatically adjust the hyperparameters of the convolutional neural network model, reduce manual intervention, and make the model development process more efficient and intelligent.
[0122] In another embodiment, see Figure 2 The present invention also provides an abnormality prediction system for a battery swap station, which may include: an acquisition module 10, a preprocessing module 20, a model building module 30, a model optimization module 40 and a report generation module 50; wherein the acquisition module 10 is used to acquire real-time operation data of the battery swap station; the preprocessing module 20 is used to preprocess the real-time operation data; the model building module 30 is used to construct an initial convolutional neural network model; the model optimization module 40 is used to optimize the initial convolutional neural network model using a Kepler optimization algorithm to obtain a convolutional neural network model; the convolutional neural network model is used to predict abnormalities of the battery swap station based on the preprocessed real-time operation data; the report generation module 50 is used to generate an abnormality report based on the prediction results.
[0123] For the specific limitations of the battery swap station abnormality prediction system, please refer to the limitations of the fault prediction model training method and the lighting equipment fault prediction method mentioned above, which will not be repeated here. Each module in the above-mentioned battery swap station abnormality prediction system can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.
[0124] In another embodiment, the present invention further provides a computer device, which may be a server, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store historical work data, real-time work data and other data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it is implemented as follows Figure 1 And the battery swap station abnormality prediction method provided in the corresponding embodiment.
[0125] In another embodiment, the present invention further provides a computer device, which may be a terminal, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, it is implemented as follows Figure 1 The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covered on the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse.
[0126] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0127] In another embodiment of the present invention, the present invention further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following is implemented: Figure 1And the battery swap station abnormality prediction method provided in the corresponding embodiment.
[0128] In another embodiment, the present invention further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the following is achieved: Figure 1 And the battery swap station abnormality prediction method provided in the corresponding embodiment.
[0129] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
[0130] 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.
[0131] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.
Claims
1. A method for predicting abnormality of a battery swap station. It is characterized in that The abnormality prediction method of the battery swap station includes: Obtain real-time operation data of the battery swap station; Preprocessing the real-time operation data; Build an initial convolutional neural network model; Using the Kepler optimization algorithm to optimize the initial convolutional neural network model to obtain a convolutional neural network model; Inputting the preprocessed real-time operation data into the convolutional neural network to perform abnormal prediction on the battery swap station; Generate exception reports based on prediction results.
2. The abnormality prediction method of a battery swap station according to claim 1, It is characterized in that The using of the Kepler optimization algorithm to optimize the initial convolutional neural network model to obtain a convolutional neural network model includes: Determining a set of hyperparameters to be optimized for the initial convolutional neural network model; Based on the hyperparameter set, generating a hyperparameter combination; Obtaining the fitness of the initial convolutional neural network model according to the hyperparameter combination; Acquire a running body of the hyperparameter combination, wherein the running body includes corresponding running data, and the running data includes position information, speed information, acceleration information, and quality information of the running body; Performing planetary motion simulation update on the operating data of the operating body corresponding to the hyperparameter combination to update the parameters of the initial convolutional neural network model; Iterative optimization is performed until the fitness converges or reaches a preset number of iterations, thereby obtaining the convolutional neural network model.
3. The abnormality prediction method of a battery swap station according to claim 2, It is characterized in that The updating of the running data of the running body corresponding to the hyperparameter combination by planetary motion simulation to update the parameters of the initial convolutional neural network model includes: Performing planetary motion simulation based on the operating data of the operating body corresponding to the hyperparameter combination, and updating the operating data; The corresponding hyperparameter combination is updated based on the updated operating data, and the parameters to be optimized corresponding to the updated hyperparameter combination are used as the parameters of the initial convolutional neural network model to complete the parameter update of the initial convolutional neural network model.
4. The abnormality prediction method of a battery swap station according to claim 3, It is characterized in that The performing planetary motion simulation based on the operating data of the operating body corresponding to the hyperparameter combination and updating the operating data includes: The position information of the running body corresponding to the hyperparameter combination is updated based on the following formula: Among them, p i is the position information of the running body before updating, p i+1 is the updated position information of the running body, v i is the current speed information of the running body, a i is the current acceleration information of the running body, △t is the time step, and i is an integer greater than or equal to 1; The acceleration information of the running body corresponding to the hyperparameter combination is updated based on the following formula: Among them, a j is the updated acceleration information of the running body, G is the gravitational constant, m i is the current quality information of the running body, p i is the position information of the running body before updating, p j is the updated position information of the running body, i is an integer greater than or equal to 1, and j is an integer greater than i; The speed information of the running body corresponding to the hyperparameter combination is updated based on the following formula: i+1 =v i +a i .△t, where V i+1 is the updated speed of the running body, v i is the current speed of the running body, a i is the current acceleration information of the running body, and △t is the time step.
5. The abnormality prediction method of a battery swap station according to claim 4, It is characterized in that After inputting the preprocessed real-time operation data into the convolutional neural network and performing abnormal prediction on the battery swap station, and before generating an abnormal report based on the prediction result, the method further includes: Determining whether the battery swap station has an abnormality based on the prediction result; If so, the abnormal type of the battery swap station is identified.
6. The abnormality prediction method of a battery swap station according to claim 5, It is characterized in that The abnormality report includes: the time when the abnormality occurred, the location where the abnormality occurred, the type of abnormality, the abnormal operating parameters, the severity of the abnormality, the cause of the abnormality and the treatment measures for the abnormality.
7. The abnormality prediction method of a battery swap station according to claim 6, It is characterized in that After generating the abnormal report based on the prediction result, the method further includes: Visually display the prediction results and the abnormal report; An alarm message is issued based on the abnormal report.
8. A battery swap station abnormality prediction system, It is characterized in that The abnormality prediction system for a battery swap station includes: An acquisition module is used to obtain real-time operation data of the battery swap station; A preprocessing module, used for preprocessing the real-time operation data; Model building module, used to build the initial convolutional neural network model; A model optimization module, used to optimize the initial convolutional neural network model using a Kepler optimization algorithm to obtain a convolutional neural network model; the convolutional neural network model is used to predict abnormalities of the battery swap station based on the preprocessed real-time operation data; The report generation module is used to generate anomaly reports based on the prediction results.
9. A computer device comprising a memory and a processor, wherein the memory stores a computer program. It is characterized in that When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.
10. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.