New energy automobile battery state monitoring system

Abstract

The invention provides a new energy automobile battery state monitoring system, which relates to the technical field of new energy automobiles, and comprises the steps of collecting original battery state data of a new energy automobile end through a battery induction sensor, generating key battery state data after preprocessing, carrying out feature extraction on the key battery state data to obtain key battery state features, and sending the key battery state features to the new energy automobile end. Constructing a battery health state prediction model and a battery remaining capacity prediction model of the new energy vehicle end, inputting the key battery state characteristics into the battery health state prediction model and the battery remaining capacity prediction model respectively to obtain a predicted battery state category and a remaining capacity predicted value, obtaining a remaining capacity actual value, and combining the predicted value to obtain a new energy vehicle state prediction model; according to the method, the battery remaining capacity prediction model is optimized, abnormity monitoring is performed on the predicted battery state category and the remaining capacity prediction value, and the battery state control strategy of the new energy vehicle end is adjusted, so that comprehensive monitoring and control of the battery state can be realized, and the battery performance and the cruising ability of the new energy vehicle are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy vehicles, and in particular to a battery status monitoring system for new energy vehicles. Background Art

[0002] With the rapid development of the new energy vehicle industry, batteries, as core components of new energy vehicles, their performance, safety and stability directly affect the vehicle's endurance and overall performance.

[0003] In the existing technology, traditional battery status monitoring systems often have problems such as insufficient accuracy, slow response speed, and inaccurate data. They cannot achieve real-time estimation of battery health status and remaining power, cannot guarantee the endurance and safety of new energy vehicles, and cannot meet the high requirements of modern new energy vehicles for battery management. In addition, the existing monitoring system lacks a complete safety protection function and cannot effectively prevent the battery from being damaged, which seriously shortens the battery life.

[0004] Therefore, it is necessary to provide a new energy vehicle battery status monitoring system to solve the above technical problems. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a new energy vehicle battery status monitoring system for solving the problems that traditional battery status monitoring systems often have insufficient accuracy, slow response speed, and inaccurate data, and cannot achieve real-time estimation of battery health status and remaining power, and cannot guarantee the endurance and safety of new energy vehicles. In addition, the system lacks complete safety protection functions and cannot effectively prevent the battery from being damaged, which seriously shortens the battery life.

[0006] The present invention provides a new energy vehicle battery status monitoring system, the monitoring system comprising:

[0007] The data acquisition module is used to collect the original battery status data of the new energy vehicle through the battery sensing sensor;

[0008] A feature extraction module, used to perform a preprocessing operation on the original battery status data to generate corresponding key battery status data, and perform feature extraction on the key battery status data to obtain key battery status features;

[0009] A model prediction module, used to construct a battery health status prediction model and a battery remaining power prediction model corresponding to the new energy vehicle end, input the key battery status features into the battery health status prediction model to obtain a predicted battery status category, and input the key battery status features into the battery remaining power prediction model to obtain a remaining power prediction value;

[0010] An optimization and adjustment module, used to obtain the actual value of the remaining power of the new energy vehicle, and optimize and adjust the battery remaining power prediction model in combination with the remaining power prediction value;

[0011] The abnormality monitoring module is used to perform abnormality monitoring on the predicted battery status category and the predicted remaining power value, and automatically adjust the real-time battery status control strategy of the new energy vehicle according to the abnormality monitoring results.

[0012] Preferably, the battery sensing sensor includes a voltage sensing sensor, a current sensing sensor and a temperature sensing sensor, and the voltage sensing sensor is used to collect the original battery voltage data of the new energy vehicle end; the current sensing sensor is used to collect the original battery current data of the new energy vehicle end; the temperature sensing sensor is used to collect the original battery temperature data of the new energy vehicle end;

[0013] The original battery voltage data, the original battery current data and the original battery temperature data of the new energy vehicle are aggregated to generate the original battery status data.

[0014] Preferably, the preprocessing operation on the original battery status data to generate corresponding key battery status data specifically includes:

[0015] Performing denoising processing on the original battery status data to generate first battery status data;

[0016] The first battery status data is normalized to generate second battery status data. The corresponding calculation formula is as follows:

[0017]

[0018] Wherein, x' represents the normalized second battery status data; x represents the first battery status data; x min represents the minimum value of the first battery status data; x max Indicates the maximum value of the first battery status data;

[0019] The second battery status data is filtered based on a low-pass filter to obtain the key battery status data. The corresponding calculation formula is as follows:

[0020]

[0021] Wherein, x" (s) represents the Laplace transform of the key battery status data, i.e., the key battery status data after filtering; s represents the complex frequency filter variable; ω0 represents the initial filter frequency; ω crepresents the cut-off frequency, which is used to determine the highest frequency allowed to pass by the low-pass filter; x'(s) represents the Laplace transform of the second battery status data, that is, the second battery status data before filtering.

[0022] Preferably, the step of inputting the key battery status feature into the battery health status prediction model to obtain a predicted battery status category specifically includes:

[0023] The key battery status features are input into the battery health status prediction model, and the key battery status features are classified. The corresponding optimization problem is as follows:

[0024]

[0025] In the formula, w u represents the weight vector of the u-th support vector machine, which is used to determine the direction of the classification hyperplane; b u represents the bias term of the u-th support vector machine, which is used to determine the distance between the classification hyperplane and the origin; ξ uv represents the slack variable when the vth key battery state feature is linearly inseparable in the uth support vector machine; C represents the regularization parameter, which is used to control the trade-off between classification interval and classification error; R represents the total number of key battery state features; min represents the minimum operation; ∑ represents the summation symbol.

[0026] Preferably, the constraints corresponding to the optimization problem are as follows:

[0027] LBBQ v (w u *DCTZ v +b u )≥1-ξ uv ,LBBQ v =u

[0028] LBBQ v (w u *DCTZ v +b u )≤ξ uv -1,LBBQ v ≠u

[0029]

[0030] Where, LBBQ v Indicates the feature category label corresponding to the vth key battery status feature; DCTZ v represents the vth key battery status feature; w u represents the weight vector of the u-th support vector machine, which is used to determine the direction of the classification hyperplane; b urepresents the bias term of the u-th support vector machine, which is used to determine the distance between the classification hyperplane and the origin; ξ uv represents the slack variable when the vth key battery state feature is linearly inseparable in the uth support vector machine; R represents the total number of key battery state features;

[0031] Based on the feature classification results of the key battery status features, the predicted battery status category corresponding to the new energy vehicle is determined.

[0032] Preferably, the step of inputting the key battery status feature into the battery remaining capacity prediction model to obtain a remaining capacity prediction value specifically includes:

[0033] The key battery status feature is used as the input of the output layer neuron corresponding to the battery remaining power prediction model. The input expression of the output layer neuron is as follows:

[0034] z l =W l a l-1 +b l

[0035] In the formula, z l represents the input of the lth layer of neurons, that is, the input of the output layer neurons; W l represents the weight matrix of the neurons in the lth layer, that is, the weight matrix of the neurons in the output layer; a l-1 represents the input of the l-1th layer of neurons, that is, the activation value of the previous layer of neurons in the output layer; b l Represents the bias vector of the lth layer of neurons, that is, the bias vector of the output layer neurons;

[0036] The corresponding activation values ​​of the output layer neurons are as follows:

[0037] a l =σ(z l )

[0038] In the formula, a l represents the activation value of the neurons in the lth layer, that is, the activation value of the neurons in the output layer; σ represents the activation function corresponding to the battery remaining power prediction model; z l Represents the input of the lth layer of neurons, that is, the input of the output layer neurons;

[0039] Based on the activation value of the output layer neuron, the remaining power prediction value corresponding to the new energy vehicle is determined.

[0040] Preferably, the optimization and adjustment module is used to obtain the actual value of the remaining power of the new energy vehicle, and optimize and adjust the battery remaining power prediction model in combination with the remaining power prediction value, specifically including:

[0041] Based on the actual value of the remaining power and the predicted value of the remaining power at the new energy vehicle end, the loss function corresponding to the battery remaining power prediction model is calculated as follows:

[0042]

[0043] Where L represents the loss function corresponding to the battery remaining power prediction model; m represents the number of data sets corresponding to the actual value of the remaining power and the predicted value of the remaining power at the new energy vehicle end; y i represents the actual value of the remaining power of the ith item; y i ' represents the i-th remaining power prediction value; ∑ represents the summation symbol.

[0044] Preferably, the weight matrix and bias vector of the output layer neurons corresponding to the battery remaining capacity prediction model are updated based on the gradient descent method, and the weight matrix of the output layer neurons after the update is as follows:

[0045]

[0046] Where W l′ represents the updated weight matrix of the lth layer neurons, that is, the updated weight matrix of the output layer neurons; W l represents the weight matrix of the l-th layer of neurons, that is, the weight matrix of the output layer neurons; α represents the learning rate corresponding to the battery remaining power prediction model; L represents the loss function corresponding to the battery remaining power prediction model; Represents the partial derivative of the loss function corresponding to the battery remaining power prediction model with respect to the weight matrix of the output layer neurons;

[0047] The updated bias vector of the output layer neuron is as follows:

[0048]

[0049] Where b l′ represents the updated bias vector of the lth layer of neurons, that is, the updated bias vector of the output layer neurons; b l represents the bias vector of the l-th layer neuron, that is, the bias vector of the output layer neuron; α represents the learning rate corresponding to the battery remaining power prediction model; L represents the loss function corresponding to the battery remaining power prediction model; Represents the partial derivative of the loss function corresponding to the battery remaining capacity prediction model with respect to the bias vector of the output layer neuron.

[0050] Compared with the related art, the new energy vehicle battery status monitoring system provided by the present invention has the following beneficial effects:

[0051] The present invention can collect original battery status data of the new energy vehicle end through a battery sensing sensor through a data acquisition module, a feature extraction module, a model prediction module, an optimization and adjustment module and an abnormality monitoring module; perform preprocessing operations on the original battery status data to generate corresponding key battery status data, and perform feature extraction on the key battery status data to obtain key battery status features; construct a battery health status prediction model and a battery remaining power prediction model corresponding to the new energy vehicle end, input the key battery status features into the battery health status prediction model to obtain a predicted battery status category, and input the key battery status features into the battery remaining power prediction model to obtain a remaining power prediction value; obtain the actual remaining power value of the new energy vehicle end, and optimize and adjust the battery remaining power prediction model in combination with the remaining power prediction value; perform abnormal monitoring on the predicted battery status category and the remaining power prediction value, and automatically adjust the real-time battery status control strategy of the new energy vehicle end according to the abnormal monitoring results. The present invention can improve the monitoring accuracy and reliability of the battery status, realize accurate and real-time estimation of the battery health status and remaining power, improve the efficiency and accuracy of the monitoring system, ensure the endurance and safety of new energy vehicles, and reduce the risk of human intervention; and the system of the present invention has complete safety protection functions and abnormal warning mechanisms, which can effectively prevent the battery from being damaged, extend the service life of the battery, ensure the stability and safety of the battery during the charging and discharging process, and ensure the driving safety of new energy vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A system block diagram of a new energy vehicle battery status monitoring system of the present invention;

[0053] Figure 2 is a flow chart of a process for generating raw battery status data of the present invention;

[0054] Figure 3 The figure is a flow chart of the preprocessing process of the raw battery status data of the present invention. DETAILED DESCRIPTION

[0055] The present invention will be further described below in conjunction with the accompanying drawings and implementation modes.

[0056] Embodiment 1

[0057] like Figure 1 As shown, a new energy vehicle battery status monitoring system, the monitoring system comprises:

[0058] The data acquisition module is used to collect the original battery status data of the new energy vehicle through the battery sensing sensor;

[0059] A feature extraction module, used to perform a preprocessing operation on the original battery status data to generate corresponding key battery status data, and perform feature extraction on the key battery status data to obtain key battery status features;

[0060] A model prediction module, used to construct a battery health status prediction model and a battery remaining power prediction model corresponding to the new energy vehicle end, input the key battery status features into the battery health status prediction model to obtain a predicted battery status category, and input the key battery status features into the battery remaining power prediction model to obtain a remaining power prediction value;

[0061] An optimization and adjustment module, used to obtain the actual value of the remaining power of the new energy vehicle, and optimize and adjust the battery remaining power prediction model in combination with the remaining power prediction value;

[0062] The abnormality monitoring module is used to perform abnormality monitoring on the predicted battery status category and the predicted remaining power value, and automatically adjust the real-time battery status control strategy of the new energy vehicle according to the abnormality monitoring results.

[0063] Among them, the data acquisition module can use a high-precision battery sensing sensor network to capture the original status data of the power battery pack of new energy vehicles in real time, including key parameters such as voltage, current, temperature and internal resistance.

[0064] Next, the feature extraction module can pre-process the collected raw data, improve the data quality through filtering, denoising and other technical means, and generate a series of key battery status data. In addition, the module can use feature extraction algorithms to extract key features that reflect the battery health status and performance trends from these data, providing effective input data for the subsequent model prediction module.

[0065] The model prediction module can build a battery health status prediction model and a battery remaining power prediction model based on machine learning algorithms. By inputting the extracted key battery status features into these models, the battery health status category can be predicted, such as normal, early stage of decline, severe decline, etc., as well as the remaining power in the future.

[0066] The optimization and adjustment module obtains the actual value of the remaining power of new energy vehicles in real time, compares and analyzes it with the predicted value, and uses a feedback mechanism to iteratively optimize the battery remaining power prediction model to reduce the model prediction error and improve the model adaptability.

[0067] Finally, the abnormal monitoring module can continuously monitor and predict the battery status category and the remaining power prediction value. Once an abnormal situation is found, such as a sharp decline in battery health or a large deviation in the remaining power prediction, the module will immediately trigger the early warning mechanism and automatically adjust the real-time battery status control strategy of the new energy vehicle, such as limiting power output and recommending charging, to ensure the driving safety and battery life of the new energy vehicle.

[0068] In the specific implementation process, Figure 2 As shown, the battery sensing sensor includes a voltage sensing sensor, a current sensing sensor and a temperature sensing sensor, and the voltage sensing sensor is used to collect the original battery voltage data of the new energy vehicle end; the current sensing sensor is used to collect the original battery current data of the new energy vehicle end; the temperature sensing sensor is used to collect the original battery temperature data of the new energy vehicle end;

[0069] The original battery voltage data, the original battery current data and the original battery temperature data of the new energy vehicle are aggregated to generate the original battery status data.

[0070] It can be understood that the battery sensing sensor specifically includes a voltage sensing sensor, a current sensing sensor and a temperature sensing sensor.

[0071] Specifically, voltage sensing sensors are used to accurately collect the original voltage data of the power battery pack of new energy vehicles. Current sensing sensors are used to monitor and record the original current data flowing through the battery pack during the charging and discharging process of new energy vehicles, which can provide key input for the thermal management and efficiency optimization of the battery. The temperature sensing sensor is used to capture the original temperature data of the battery system in real time, which can effectively prevent the battery from being damaged, extend the battery life, ensure the stability and safety of the battery during the charging and discharging process, and ensure the driving safety of new energy vehicles.

[0072] Subsequently, the raw battery voltage data, raw battery current data, and raw battery temperature data can be summarized and analyzed to generate comprehensive raw battery status data, which in turn helps to improve the overall performance of new energy vehicles and extend battery life.

[0073] like Figure 3 As shown, the preprocessing operation on the original battery status data to generate corresponding key battery status data specifically includes:

[0074] Performing denoising processing on the original battery status data to generate first battery status data;

[0075] The first battery status data is normalized to generate second battery status data. The corresponding calculation formula is as follows:

[0076]

[0077] Wherein, x' represents the normalized second battery status data; x represents the first battery status data; x min represents the minimum value of the first battery status data; x max Indicates the maximum value of the first battery status data;

[0078] The second battery status data is filtered based on a low-pass filter to obtain the key battery status data. The corresponding calculation formula is as follows:

[0079]

[0080] Wherein, x" (s) represents the Laplace transform of the key battery status data, i.e., the key battery status data after filtering; s represents the complex frequency filter variable; ω0 represents the initial filter frequency; ω c represents the cut-off frequency, which is used to determine the highest frequency allowed to pass by the low-pass filter; x'(s) represents the Laplace transform of the second battery status data, that is, the second battery status data before filtering.

[0081] In practical applications, the original battery status data may first be subjected to denoising processing to eliminate noise caused by measurement errors or external environmental interference, thereby generating purer first battery status data.

[0082] Next, the first battery status data may be normalized and mapped into a standard numerical range through a specific mathematical transformation to generate second battery status data.

[0083] Finally, the second battery status data can be further smoothed based on the filtering process of the low-pass filter to obtain the key battery status data. Specifically, by controlling the frequency response of the low-pass filter, only signal components below the cutoff frequency are allowed to pass, which can effectively filter out high-frequency noise.

[0084] Among them, the initial filter frequency and cut-off frequency are key parameters in the design of the low-pass filter, which determine the frequency characteristics and performance of the low-pass filter.

[0085] The step of inputting the key battery status feature into the battery health status prediction model to obtain a predicted battery status category specifically includes:

[0086] The key battery status features are input into the battery health status prediction model, and the key battery status features are classified. The corresponding optimization problem is as follows:

[0087]

[0088] In the formula, w u represents the weight vector of the u-th support vector machine, which is used to determine the direction of the classification hyperplane; b u represents the bias term of the u-th support vector machine, which is used to determine the distance between the classification hyperplane and the origin; ξ uv represents the slack variable when the vth key battery state feature is linearly inseparable in the uth support vector machine; C represents the regularization parameter, which is used to control the trade-off between classification interval and classification error; R represents the total number of key battery state features; min represents the minimum operation; ∑ represents the summation symbol.

[0089] The constraints corresponding to the optimization problem are as follows:

[0090] LBBQ v (w u *DCTZ v +b u )≥1-ξ uv ,LBBQ v =u

[0091] LBBQ v 9w u *DCTZ v +b u )≤ξ uv -1,LBBQ v ≠u

[0092]

[0093] Where, LBBQ v Indicates the feature category label corresponding to the vth key battery status feature; DCTZ v represents the vth key battery status feature; w u represents the weight vector of the u-th support vector machine, which is used to determine the direction of the classification hyperplane; b u represents the bias term of the u-th support vector machine, which is used to determine the distance between the classification hyperplane and the origin; ξ uv represents the slack variable when the vth key battery state feature is linearly inseparable in the uth support vector machine; R represents the total number of key battery state features;

[0094] Among them, after the key battery status features are input into the battery health status prediction model, the battery status can be effectively classified and predicted. Specifically, by introducing the support vector machine, the input key battery status features can be feature classified. In this process, the weight vector and bias term of the support vector machine jointly determine the direction and position of the classification hyperplane, thereby effectively distinguishing the key battery status features of different categories.

[0095] In practical applications, there may be some linearly inseparable feature points. These special cases can be handled by introducing slack variables to enhance the flexibility and adaptability of the battery health status prediction model.

[0096] At the same time, the introduction of regularization parameters helps to find a reasonable balance between classification interval and classification error, ensuring that the battery health status prediction model does not overfit the training data while maintaining a high classification accuracy.

[0097] Under the constraints, the battery health status prediction model can ensure that each key battery status feature is correctly classified into the corresponding feature category label, thereby improving the accuracy of feature classification and facilitating the subsequent determination of the predicted battery status category of new energy vehicles.

[0098] The step of inputting the key battery status feature into the battery remaining capacity prediction model to obtain a remaining capacity prediction value specifically includes:

[0099] The key battery status feature is used as the input of the output layer neuron corresponding to the battery remaining power prediction model. The input expression of the output layer neuron is as follows:

[0100] z l =W l a l-1 +b l

[0101] In the formula, z l represents the input of the lth layer of neurons, that is, the input of the output layer neurons; W l represents the weight matrix of the neurons in the lth layer, that is, the weight matrix of the neurons in the output layer; a l-1 represents the input of the l-1th layer of neurons, that is, the activation value of the previous layer of neurons in the output layer; b l Represents the bias vector of the lth layer of neurons, that is, the bias vector of the output layer neurons;

[0102] The corresponding activation values ​​of the output layer neurons are as follows:

[0103] a l =σ(z l )

[0104] In the formula, a l represents the activation value of the neurons in the lth layer, that is, the activation value of the neurons in the output layer; σ represents the activation function corresponding to the battery remaining power prediction model; z l Represents the input of the lth layer of neurons, that is, the input of the output layer neurons;

[0105] Based on the activation value of the output layer neuron, the remaining power prediction value corresponding to the new energy vehicle is determined.

[0106] It should be noted that by using the key battery status features as the input of the output layer neurons corresponding to the battery remaining capacity prediction model, the battery status information can be accurately captured. The activation value of the output layer neurons is calculated by the activation function. Among them, the introduction of the activation function makes the battery remaining capacity prediction model more flexible and accurate when dealing with nonlinear problems.

[0107] Finally, based on the activation value of the neurons in the output layer, the corresponding remaining power prediction value of the new energy vehicle can be determined, and then the energy management strategy of the new energy vehicle can be optimized to extend the battery life.

[0108] The optimization and adjustment module is used to obtain the actual value of the remaining power of the new energy vehicle, and optimize and adjust the battery remaining power prediction model in combination with the remaining power prediction value, specifically including:

[0109] Based on the actual value of the remaining power and the predicted value of the remaining power at the new energy vehicle end, the loss function corresponding to the battery remaining power prediction model is calculated as follows:

[0110]

[0111] Where L represents the loss function corresponding to the battery remaining power prediction model; m represents the number of data sets corresponding to the actual value of the remaining power and the predicted value of the remaining power at the new energy vehicle end; y i Indicates the actual value of the remaining power of the ith item; y i ' represents the i-th remaining power prediction value; ∑ represents the summation symbol.

[0112] The weight matrix and bias vector of the output layer neurons corresponding to the battery remaining capacity prediction model are updated based on the gradient descent method. The updated weight matrix of the output layer neurons is as follows:

[0113]

[0114] Where W l′ represents the updated weight matrix of the lth layer neurons, that is, the updated weight matrix of the output layer neurons; W l represents the weight matrix of the l-th layer of neurons, that is, the weight matrix of the output layer neurons; α represents the learning rate corresponding to the battery remaining power prediction model; L represents the loss function corresponding to the battery remaining power prediction model; Represents the partial derivative of the loss function corresponding to the battery remaining capacity prediction model with respect to the weight matrix of the output layer neurons;

[0115] The updated bias vector of the output layer neuron is as follows:

[0116]

[0117] Where b l′ represents the updated bias vector of the lth layer of neurons, that is, the updated bias vector of the output layer neurons; b l represents the bias vector of the l-th layer neuron, that is, the bias vector of the output layer neuron; α represents the learning rate corresponding to the battery remaining power prediction model; L represents the loss function corresponding to the battery remaining power prediction model; Represents the partial derivative of the loss function corresponding to the battery remaining capacity prediction model with respect to the bias vector of the output layer neuron.

[0118] In practical applications, the optimization and adjustment module can first obtain the actual value and predicted value of the remaining power of the new energy vehicle, and use these data to calculate the loss function of the battery remaining power prediction model to quantify the difference between the model prediction value and the actual value.

[0119] By introducing the gradient descent method, the optimization and adjustment module can efficiently update the weight matrix and bias vector of the output layer neurons of the battery remaining power prediction model. This update strategy is based on the partial derivatives of the loss function with respect to the weight matrix and bias vector, which can ensure that the model is adjusted in the direction of reducing the prediction error. As the number of iterations increases, the prediction accuracy of the model will continue to improve, and ultimately achieve accurate prediction of the remaining battery power of new energy vehicles.

[0120] In addition, the optimization and adjustment module is also adaptive and flexible. Due to the application of loss function and gradient descent method, the model can dynamically adjust its prediction ability according to the real-time data of new energy vehicles, ensuring that reliable prediction results can be provided under different working conditions and driving conditions, thereby improving the endurance management ability of new energy vehicles, providing more accurate remaining power information, and facilitating the formulation of charging and driving decisions.

[0121] Through the introduction of the above embodiments, the present invention, through a new energy vehicle battery status monitoring system, can collect the original battery status data of the new energy vehicle end through a battery sensing sensor through a data acquisition module, a feature extraction module, a model prediction module, an optimization and adjustment module and an abnormality monitoring module; preprocess the original battery status data to generate corresponding key battery status data, and extract features from the key battery status data to obtain key battery status features; construct a battery health status prediction model and a battery remaining power prediction model corresponding to the new energy vehicle end, input the key battery status features into the battery health status prediction model to obtain a predicted battery status category, and input the key battery status features into the battery remaining power prediction model to obtain a remaining power prediction value; obtain the actual remaining power value of the new energy vehicle end, and optimize and adjust the battery remaining power prediction model in combination with the remaining power prediction value; perform abnormal monitoring on the predicted battery status category and the remaining power prediction value, and automatically adjust the real-time battery status control strategy of the new energy vehicle end according to the abnormal monitoring results. The present invention can improve the monitoring accuracy and reliability of the battery status, realize accurate and real-time estimation of the battery health status and remaining power, improve the efficiency and accuracy of the monitoring system, ensure the endurance and safety of new energy vehicles, and reduce the risk of human intervention; and the system of the present invention has complete safety protection functions and abnormal warning mechanisms, which can effectively prevent the battery from being damaged, extend the service life of the battery, ensure the stability and safety of the battery during the charging and discharging process, and ensure the driving safety of new energy vehicles.

[0122] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0123] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0124] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

Claims

1. A new energy vehicle battery status monitoring system, characterized in that: The monitoring system comprises: The data acquisition module is used to collect the original battery status data of the new energy vehicle through the battery sensing sensor; A feature extraction module, used to perform a preprocessing operation on the original battery status data to generate corresponding key battery status data, and perform feature extraction on the key battery status data to obtain key battery status features; A model prediction module, used to construct a battery health status prediction model and a battery remaining power prediction model corresponding to the new energy vehicle end, input the key battery status features into the battery health status prediction model to obtain a predicted battery status category, and input the key battery status features into the battery remaining power prediction model to obtain a remaining power prediction value; An optimization and adjustment module, used to obtain the actual value of the remaining power of the new energy vehicle, and optimize and adjust the battery remaining power prediction model in combination with the remaining power prediction value; The abnormality monitoring module is used to perform abnormality monitoring on the predicted battery status category and the predicted remaining power value, and automatically adjust the real-time battery status control strategy of the new energy vehicle according to the abnormality monitoring results.

2. A new energy vehicle battery status monitoring system according to claim 1, characterized in that: The battery sensing sensor includes a voltage sensing sensor, a current sensing sensor and a temperature sensing sensor, and the voltage sensing sensor is used to collect the original battery voltage data of the new energy vehicle end; the current sensing sensor is used to collect the original battery current data of the new energy vehicle end; the temperature sensing sensor is used to collect the original battery temperature data of the new energy vehicle end; The original battery voltage data, the original battery current data and the original battery temperature data of the new energy vehicle are aggregated to generate the original battery status data.

3. A new energy vehicle battery status monitoring system according to claim 1, characterized in that: The preprocessing operation on the original battery status data to generate corresponding key battery status data specifically includes: Performing denoising processing on the original battery status data to generate first battery status data; The first battery status data is normalized to generate second battery status data. The corresponding calculation formula is as follows: Wherein, x' represents the normalized second battery status data; x represents the first battery status data; x min represents the minimum value of the first battery status data; x max Indicates the maximum value of the first battery status data; The second battery status data is filtered based on a low-pass filter to obtain the key battery status data. The corresponding calculation formula is as follows: Wherein, x" (s) represents the Laplace transform of the key battery status data, i.e., the key battery status data after filtering; s represents the complex frequency filter variable; ω0 represents the initial filter frequency; ω c represents the cut-off frequency, which is used to determine the highest frequency allowed to pass by the low-pass filter; x'(s) represents the Laplace transform of the second battery status data, that is, the second battery status data before filtering.

4. A new energy vehicle battery status monitoring system according to claim 1, characterized in that: The step of inputting the key battery status feature into the battery health status prediction model to obtain a predicted battery status category specifically includes: The key battery status features are input into the battery health status prediction model, and the key battery status features are classified. The corresponding optimization problem is as follows: In the formula, w u represents the weight vector of the u-th support vector machine, which is used to determine the direction of the classification hyperplane; b u represents the bias term of the u-th support vector machine, which is used to determine the distance between the classification hyperplane and the origin; ξ uv represents the slack variable when the vth key battery state feature is linearly inseparable in the uth support vector machine; C represents the regularization parameter, which is used to control the trade-off between classification interval and classification error; R represents the total number of key battery state features; min represents the minimum operation; ∑ represents the summation symbol.

5. A new energy vehicle battery status monitoring system according to claim 4, characterized in that: The constraints corresponding to the optimization problem are as follows: LBBQ v (w u *DCTZ v +b u )≥1-ξ uv ,LBBQ v =u LBBQ v (w u *DCTZ v +b u )≤ξ uv -1,LBBQ v ≠u Where, LBBQ v Indicates the feature category label corresponding to the vth key battery status feature; DCTZ v represents the vth key battery status feature; w u represents the weight vector of the u-th support vector machine, which is used to determine the direction of the classification hyperplane; b u represents the bias term of the u-th support vector machine, which is used to determine the distance between the classification hyperplane and the origin; ξ uv represents the slack variable when the vth key battery state feature is linearly inseparable in the uth support vector machine; R represents the total number of key battery state features; Based on the feature classification results of the key battery status features, the predicted battery status category corresponding to the new energy vehicle is determined.

6. A new energy vehicle battery status monitoring system according to claim 1, characterized in that: The step of inputting the key battery status feature into the battery remaining capacity prediction model to obtain a remaining capacity prediction value specifically includes: The key battery status feature is used as the input of the output layer neuron corresponding to the battery remaining power prediction model. The input expression of the output layer neuron is as follows: z l =W l a l-1 +b l In the formula, z l represents the input of the lth layer of neurons, that is, the input of the output layer neurons; W l represents the weight matrix of the neurons in the lth layer, that is, the weight matrix of the neurons in the output layer; a l-1 represents the input of the l-1th layer of neurons, that is, the activation value of the previous layer of neurons in the output layer; b l Represents the bias vector of the lth layer of neurons, that is, the bias vector of the output layer neurons; The corresponding activation values ​​of the output layer neurons are as follows: a l =σ(z l ) In the formula, a l represents the activation value of the neurons in the lth layer, that is, the activation value of the neurons in the output layer; σ represents the activation function corresponding to the battery remaining power prediction model; z l Represents the input of the lth layer of neurons, that is, the input of the output layer neurons; Based on the activation value of the output layer neuron, the remaining power prediction value corresponding to the new energy vehicle is determined.

7. A new energy vehicle battery status monitoring system according to claim 1, characterized in that: The optimization and adjustment module is used to obtain the actual value of the remaining power of the new energy vehicle, and optimize and adjust the battery remaining power prediction model in combination with the remaining power prediction value, specifically including: Based on the actual value of the remaining power and the predicted value of the remaining power at the new energy vehicle end, the loss function corresponding to the battery remaining power prediction model is calculated as follows: Where L represents the loss function corresponding to the battery remaining power prediction model; m represents the number of data sets corresponding to the actual value of the remaining power and the predicted value of the remaining power at the new energy vehicle end; y i Indicates the actual value of the remaining power of the ith item; y i ' represents the i-th remaining power prediction value; ∑ represents the summation symbol.

8. A new energy vehicle battery status monitoring system according to claim 6 or 7, characterized in that: The weight matrix and bias vector of the output layer neurons corresponding to the battery remaining capacity prediction model are updated based on the gradient descent method. The updated weight matrix of the output layer neurons is as follows: Where W l' represents the updated weight matrix of the lth layer neurons, that is, the updated weight matrix of the output layer neurons; W l represents the weight matrix of the l-th layer of neurons, that is, the weight matrix of the output layer neurons; α represents the learning rate corresponding to the battery remaining power prediction model; L represents the loss function corresponding to the battery remaining power prediction model; Represents the partial derivative of the loss function corresponding to the battery remaining capacity prediction model with respect to the weight matrix of the output layer neurons; The updated bias vector of the output layer neuron is as follows: Where b l' represents the updated bias vector of the lth layer of neurons, that is, the updated bias vector of the output layer neurons; b l represents the bias vector of the l-th layer neuron, that is, the bias vector of the output layer neuron; α represents the learning rate corresponding to the battery remaining power prediction model; L represents the loss function corresponding to the battery remaining power prediction model; Represents the partial derivative of the loss function corresponding to the battery remaining capacity prediction model with respect to the bias vector of the output layer neuron.

Images