Battery health condition determination method and device, equipment and storage medium

Abstract

The embodiment of the invention provides a battery health condition determination method and device, equipment and a storage medium. The method comprises the following steps: acquiring first monitoring data of a power distribution automation terminal battery in a first duration period and second monitoring data at the current time; determining a target monitoring data feature vector according to the first monitoring data and the second monitoring data; and inputting the target monitoring data feature vector into a battery health condition evaluation model to obtain a battery health condition evaluation result and a floating charge life prediction result. The method is used for achieving the purpose of finding potential problems of the battery in real time so as to assist operation and maintenance personnel in making a reasonable battery maintenance and replacement plan, and the use reliability of the battery and the overall stability of the power distribution automation terminal are remarkably improved.

Description

Technical Field

[0001] This application relates to the field of battery management technology for power distribution automation terminals, and in particular to a method, apparatus, device, and storage medium for determining battery health status. Background Technology

[0002] In power distribution automation terminals, batteries are usually in a float charging state. After long-term operation, the battery capacity and health status will degrade. Timely and accurate assessment of battery health status and prediction of the remaining float charging life of the battery are of great significance to ensuring the reliability of the power distribution terminal.

[0003] Currently, battery health status testing in the middle stage of power distribution automation usually needs to be carried out offline, that is, the battery is removed and tested with specialized equipment, so as to determine whether the battery needs to be replaced or continued to be used based on the battery life prediction results.

[0004] However, this detection method is not real-time, so the equipment operation is often hindered due to untimely detection. Summary of the Invention

[0005] This application provides a method, apparatus, device, and storage medium for determining battery health status, so as to achieve timely and accurate assessment of battery health status.

[0006] In a first aspect, embodiments of this application provide a method for determining battery health status, including:

[0007] The system acquires first monitoring data of the distribution automation terminal battery within a first time period and second monitoring data at the current time. The first monitoring data includes: battery terminal voltage, float charging current, battery internal resistance, ambient temperature, battery surface temperature, and float charging duration. The second monitoring data includes: first battery capacity. The first time period is a preset time period prior to the current time.

[0008] Based on the first monitoring data and the second monitoring data, determine the target monitoring data feature vector;

[0009] The target monitoring data feature vector is input into the battery health status assessment model to obtain the battery health status assessment result and the float charge life prediction result. The battery health status assessment model includes a health status assessment network and a float charge life prediction network set in parallel.

[0010] In one or more embodiments, acquiring the first monitoring data of the distribution automation terminal battery within a first time period includes:

[0011] The system acquires and transmits first monitoring data collected and transmitted by a monitoring device embedded or externally located in the terminal corresponding to the battery of the power distribution automation terminal. The monitoring device includes a battery monitoring system, a temperature sensor, and an internal resistance tester.

[0012] In one or more embodiments, acquiring the second monitoring data at the current time includes:

[0013] In response to a request to determine battery health status, obtain the battery capacity collected and transmitted back by the online capacity assessment device;

[0014] The determination request is based on user input instructions or is generated based on a preset time interval.

[0015] In one or more embodiments, determining the target monitoring data feature vector based on the first monitoring data and the second monitoring data includes:

[0016] Based on the first monitoring data within the first time period, a digital signal is determined, which includes: the voltage fluctuation range and average voltage of the battery terminal voltage, the current fluctuation range and average current of the float charge current, the internal resistance value and internal resistance growth data of the battery internal resistance, the maximum temperature difference between the battery surface temperature and the ambient temperature at all sampling times, and the float charge duration.

[0017] The battery capacity decay rate is determined based on the first battery capacity and the second battery capacity of the power distribution automation terminal battery before the first duration period.

[0018] The target monitoring data feature vector is determined based on the digital signal, the battery capacity decay rate, and the first battery capacity.

[0019] In one or more embodiments, before inputting the target monitoring data feature vector into the battery health status assessment model to obtain the battery health status assessment result and float charge life prediction result, the method further includes:

[0020] Based on the feature vectors of historical monitoring data and the actual battery health status results corresponding to the feature vectors of historical monitoring data, an artificial neural network (ANN) is trained to obtain the health status assessment network in the battery health status assessment model.

[0021] Based on the feature vectors of historical monitoring data and the actual float life results corresponding to the feature vectors of historical monitoring data, the Long Short-Term Memory (LSTM) network is trained to obtain the float life prediction network in the battery health status assessment model.

[0022] In one or more embodiments, the health status assessment network is built on an ANN; the ANN includes: a first input layer, a first hidden layer, a second hidden layer, and a first output layer;

[0023] The float charge lifetime prediction network is built on LSTM; the LSTM includes: a second input layer, a first convolutional layer, a second convolutional layer, a first LSTM layer, and a second output layer;

[0024] The total output of the battery health status assessment model is connected to the first output layer and the second output layer, respectively.

[0025] In one or more embodiments, the float charge lifetime prediction network includes: a first float charge lifetime prediction network and a second float charge lifetime prediction network;

[0026] The first float charge lifetime prediction network is built on LSTM and includes: a second input layer, a first convolutional layer, a second convolutional layer, a first LSTM layer, and a second output layer.

[0027] The second float-charge lifetime prediction network is built on LSTM and includes: a first buffer layer, a third input layer, a third convolutional layer, a fourth convolutional layer, a second LSTM layer, a third LSTM layer, and a third output layer.

[0028] Secondly, embodiments of this application provide a device for determining battery health status, comprising:

[0029] The acquisition module is used to acquire first monitoring data of the distribution automation terminal battery within a first time period and second monitoring data at the current time. The first monitoring data includes: battery terminal voltage, float charging current, battery internal resistance, ambient temperature, battery surface temperature, and float charging duration. The second monitoring data includes: first battery capacity. The first time period is a preset time period before the current time.

[0030] The determination module is used to determine the target monitoring data feature vector based on the first monitoring data and the second monitoring data;

[0031] The processing module is used to input the feature vector of the target monitoring data into the battery health status assessment model to obtain the battery health status assessment result and the float charge life prediction result. The battery health status assessment model includes a health status assessment network and a float charge life prediction network set in parallel.

[0032] In one or more embodiments, the acquisition module acquires first monitoring data of the distribution automation terminal battery within a first time period, specifically for:

[0033] The system acquires and transmits first monitoring data collected and transmitted by a monitoring device embedded or externally located in the terminal corresponding to the battery of the power distribution automation terminal. The monitoring device includes a battery monitoring system, a temperature sensor, and an internal resistance tester.

[0034] In one or more embodiments, the acquisition module acquires the second monitoring data at the current time, specifically for:

[0035] In response to a request to determine battery health status, obtain the battery capacity collected and transmitted back by the online capacity assessment device;

[0036] The determination request is based on user input instructions or is generated based on a preset time interval.

[0037] In one or more embodiments, the determining module is specifically used for:

[0038] Based on the first monitoring data within the first time period, a digital signal is determined, which includes: the voltage fluctuation range and average voltage of the battery terminal voltage, the current fluctuation range and average current of the float charge current, the internal resistance value and internal resistance growth data of the battery internal resistance, the maximum temperature difference between the battery surface temperature and the ambient temperature at all sampling times, and the float charge duration.

[0039] The battery capacity decay rate is determined based on the first battery capacity and the second battery capacity of the power distribution automation terminal battery before the first duration period.

[0040] The target monitoring data feature vector is determined based on the digital signal, the battery capacity decay rate, and the first battery capacity.

[0041] In one or more embodiments, before inputting the target monitoring data feature vector into the battery health status assessment model to obtain the battery health status assessment result and the float charge life prediction result, the determining module is further configured to:

[0042] Based on the feature vectors of historical monitoring data and the actual battery health status results corresponding to the feature vectors of historical monitoring data, an artificial neural network (ANN) is trained to obtain the health status assessment network in the battery health status assessment model.

[0043] Based on the feature vectors of historical monitoring data and the actual float life results corresponding to the feature vectors of historical monitoring data, a Long Short-Term Memory (LSTM) network is trained to obtain the float life prediction network in the battery health status assessment model.

[0044] In one or more embodiments, the health status assessment network is built on an ANN; the ANN includes: a first input layer, a first hidden layer, a second hidden layer, and a first output layer;

[0045] The float charge lifetime prediction network is built on LSTM; the LSTM includes: a second input layer, a first convolutional layer, a second convolutional layer, a first LSTM layer, and a second output layer;

[0046] The total output of the battery health status assessment model is connected to the first output layer and the second output layer, respectively.

[0047] In one or more embodiments, the float charge lifetime prediction network includes: a first float charge lifetime prediction network and a second float charge lifetime prediction network;

[0048] The first float charge lifetime prediction network is built on LSTM and includes: a second input layer, a first convolutional layer, a second convolutional layer, a first LSTM layer, and a second output layer.

[0049] The second float-charge lifetime prediction network is built on LSTM and includes: a first buffer layer, a third input layer, a third convolutional layer, a fourth convolutional layer, a second LSTM layer, a third LSTM layer, and a third output layer.

[0050] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0051] The memory stores computer-executed instructions;

[0052] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0053] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0054] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0055] The battery health status determination method, apparatus, device, and storage medium provided in this application embodiment acquire first monitoring data of the distribution automation terminal battery within a first time period and second monitoring data at the current time. The first monitoring data includes: battery terminal voltage, float charge current, battery internal resistance, ambient temperature, battery surface temperature, and float charge duration. The second monitoring data includes: first battery capacity. The first time period is a preset time period prior to the current time. Based on the first and second monitoring data, a target monitoring data feature vector is determined. The target monitoring data feature vector is input into a battery health status assessment model to obtain battery health status assessment results and float charge life prediction results. The battery health status assessment model includes a health status assessment network and a float charge life prediction network set in parallel. In this technical solution, by collecting multi-dimensional operating parameters and current capacity data of the battery in real time, a comprehensive feature vector is constructed, and a parallel dual-network model is used for synchronous analysis, thereby achieving accurate assessment of the health status of the distribution automation terminal battery and scientific prediction of its remaining life. Its technical advantages are reflected in: 1) Comprehensiveness: Integrating dynamic operating parameters and static characteristic parameters to construct a more complete battery status profile; 2) Real-time performance: Combining data from the first time period with the current battery capacity for dynamic calibration to improve assessment accuracy; 3) Predictiveness: Through independent and parallel health assessment networks and life prediction networks, assessment results are output synchronously, avoiding the limitations of traditional single models; 4) Early warning: Based on multi-parameter fusion analysis, the battery performance degradation trend can be detected in real time, providing data support for maintenance decisions. Attached Figure Description

[0056] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0057] Figure 1 A flowchart illustrating the method for determining battery health status provided in this application embodiment. Figure 1 ;

[0058] Figure 2 A flowchart illustrating the method for determining battery health status provided in this application embodiment. Figure 2 ;

[0059] Figure 3 A flowchart illustrating the method for determining battery health status provided in this application embodiment. Figure 3 ;

[0060] Figure 4 A schematic diagram of the structure of the battery health status determination device provided in the embodiments of this application;

[0061] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0062] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0063] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0064] In power distribution automation terminals, batteries are usually in a float charging state. After long-term operation, the battery capacity and health status will degrade. Timely and accurate assessment of battery health status and prediction of the remaining float charging life of the battery are of great significance to ensuring the reliability of the power distribution terminal.

[0065] Currently, battery health status testing in the middle stage of power distribution automation usually needs to be carried out offline, that is, the battery is removed and tested with specialized equipment, so as to determine whether the battery needs to be replaced or continued to be used based on the battery life prediction results.

[0066] However, this detection method is not real-time, so the equipment operation is often hindered due to untimely detection.

[0067] Based on the above-mentioned technical problems, the inventors' technical concept is as follows: From the perspective of multi-dimensional dynamic monitoring, combined with the actual operation scenario of the battery, a comprehensive feature vector can be constructed by collecting multi-parameter data (voltage, current, internal resistance, temperature, etc.) from historical cycles and the current real-time battery capacity to fully capture the changing patterns of battery status; secondly, parallel dual-network models are introduced, focusing on health assessment and lifespan prediction respectively, avoiding the reduction in accuracy of a single model due to task conflicts. This not only solves the shortcomings of traditional methods in terms of real-time performance, but also improves the accuracy and foresight of predictions through data fusion and model collaboration, ultimately realizing intelligent and real-time monitoring of battery status.

[0068] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0069] Figure 1 A flowchart illustrating the method for determining battery health status provided in this application embodiment. Figure 1 ,like Figure 1 As shown, the method includes:

[0070] Step 11: Obtain the first monitoring data of the distribution automation terminal battery within the first time period and the second monitoring data at the current time;

[0071] The first monitoring data includes: battery terminal voltage, float charging current, battery internal resistance, ambient temperature, battery surface temperature, and float charging duration; the second monitoring data includes: first battery capacity, and the first time period is a preset time period prior to the current time.

[0072] In this step, the battery terminal voltage, float charging current, continuous float charging time, ambient temperature, battery surface temperature, battery capacity, and battery internal resistance all have a direct impact on the battery's health and lifespan.

[0073] Optionally, the preset duration period is 100 minutes. If the current time is 06:40, then the first duration period is 05:00-06:40. The collection frequency can be 10 minutes / time. The preset duration period can be 1 day, 7 days or 14 days.

[0074] In one possible implementation, Table 1 is a schematic table of the first monitoring data, as shown in Table 1:

[0075] Table 1

[0076]

[0077] Among them, the unit of battery terminal voltage V(t) is V, the unit of float charging current I(t) is A, the unit of battery internal resistance R(t) is mΩ, the unit of ambient temperature Tenv(t) is °C, the unit of battery surface temperature Tbat(t) is °C, and the unit of float charging duration tfloat(t) is min.

[0078] Optionally, one possible implementation for acquiring the first monitoring data of the distribution automation terminal battery within the first time period is as follows:

[0079] The system acquires and transmits first monitoring data collected and transmitted by a monitoring device embedded or externally located in the corresponding terminal of the power distribution automation terminal battery. The monitoring device includes a battery monitoring system, a temperature sensor, and an internal resistance tester.

[0080] In this implementation, the battery's current, voltage, and other data can be obtained through the battery monitoring system in the monitoring device; data related to ambient temperature and battery temperature can be obtained through the temperature sensor in the monitoring device; the battery's internal resistance data can be obtained through the internal resistance tester in the monitoring device; and the current continuous float charging time can be recorded through the intelligent device of the power distribution automation terminal in the monitoring device.

[0081] Through continuous monitoring data recording, multi-dimensional monitoring data tables can be accurately organized and recorded, laying the foundation for further feature extraction.

[0082] Optionally, one possible implementation for obtaining the second monitoring data at the current time is as follows:

[0083] In response to a request to determine battery health status, obtain the battery capacity collected and transmitted back by the online capacity assessment device;

[0084] The determination request is based on user input instructions or is triggered by a preset time interval.

[0085] In this implementation, when a battery health status assessment needs to be initiated, the battery capacity C collected and transmitted back by the online capacity assessment device is obtained. online By automatically acquiring real-time online estimates of battery capacity through an online capacity assessment device, accurate battery capacity estimation data can be obtained while ensuring normal battery operation.

[0086] In one possible implementation, the online capacity assessment device applies a brief, small current pulse while ensuring the normal operation of the distribution automation terminal, and records the terminal voltage response and internal resistance changes. Based on information such as internal resistance changes, terminal voltage drop / rise, and known nominal capacity, the actual battery capacity, i.e., the battery capacity, is estimated.

[0087] Correspondingly, a longer detection cycle (i.e., a preset time interval) can be set, or the battery capacity detection can be initiated again as needed to avoid the detection frequency of the capacity measuring device on the electromagnetic capacity being too frequent.

[0088] Step 12: Determine the target monitoring data feature vector based on the first monitoring data and the second monitoring data;

[0089] In this step, after obtaining the battery terminal voltage, float charging current, battery internal resistance, ambient temperature, battery surface temperature, float charging duration, and first battery capacity, feature extraction processing is performed based on the battery terminal voltage, float charging current, battery internal resistance, ambient temperature, battery surface temperature, float charging duration, and first battery capacity to construct the target monitoring data feature vector.

[0090] Step 13: Input the target monitoring data feature vector into the battery health status assessment model to obtain the battery health status assessment results and float charging life prediction results;

[0091] The battery health status assessment model includes a health status assessment network and a float charge life prediction network set up in parallel.

[0092] In this step, the target monitoring data feature vector is input into the health status assessment network, and the battery health status assessment result is output; the target monitoring data feature vector is input into the float charge life prediction network, and the float charge life prediction result is output.

[0093] The battery health status determination method provided in this application embodiment acquires first monitoring data of the battery in the distribution automation terminal within a first time period and second monitoring data at the current time. The first monitoring data includes: battery terminal voltage, float charge current, battery internal resistance, ambient temperature, battery surface temperature, and float charge duration. The second monitoring data includes: first battery capacity. The first time period is a preset time period prior to the current time. Based on the first and second monitoring data, a target monitoring data feature vector is determined. The target monitoring data feature vector is input into a battery health status assessment model to obtain a battery health status assessment result and a float charge life prediction result. The battery health status assessment model includes a health status assessment network and a float charge life prediction network set in parallel. This technical solution extracts key features and constructs feature vectors based on monitoring data acquired over a period of time, which helps improve the feature diversity of the feature vectors. Finally, based on the trained battery health status assessment model, the battery health status and float charge life are accurately assessed based on the feature vectors. This helps to promptly identify potential battery problems, assists maintenance personnel in formulating reasonable maintenance and replacement plans, and significantly improves the reliability of battery use and the overall stability of the distribution automation terminal.

[0094] Based on the above embodiments, Figure 2 A flowchart illustrating the method for determining battery health status provided in this application embodiment. Figure 2 ,like Figure 2 As shown, step 12 may include:

[0095] It should be understood that there are no restrictions on the execution order of steps 21 and 22.

[0096] Step 21: Determine the digital signal based on the first monitoring data within the first time period;

[0097] The digital signals include: the voltage fluctuation range and average voltage of the battery terminal voltage, the current fluctuation range and average current of the float charge current, the internal resistance value and internal resistance growth data of the battery internal resistance, the maximum temperature difference between the battery surface temperature and the ambient temperature at all sampling times, and the float charge duration.

[0098] Optionally, the feature extraction methods for the voltage fluctuation range and average voltage of the battery terminal voltage can be:

[0099] For the battery terminal voltage V(t), where t=1,2,…,T; T represents the total number of sampling points, the maximum value V of the battery terminal voltage is extracted for each sampling point. max and minimum value V min The voltage fluctuation range Vb = V is obtained. max -V min ;

[0100] The average voltage is further calculated based on the voltage values ​​at each sampling point.

[0101] Optionally, the current fluctuation range and average current of the float charge current can be characterized by the following methods:

[0102] For the floating charge current I(t), where t = 1, 2, ..., T; T represents the total number of sampling points, the maximum value I of the floating charge current is extracted for each sampling point. max and minimum value I min The floating charge current fluctuation range Ib = I is obtained. max -I min ;

[0103] The average float current is further calculated based on the float current values ​​at each sampling point.

[0104] Optionally, the feature extraction methods for the battery's internal resistance value and internal resistance growth data (growth rate) can be:

[0105] For the battery internal resistance R(t), where t=1,2,…,T; T represents the total number of sampling points, the growth rate ΔR of the battery internal resistance during the first time period is obtained. rate .

[0106] Optionally, the feature extraction method for the maximum temperature difference between the battery surface temperature and the ambient temperature across all sampling times can be:

[0107] For ambient temperature T env (t) and the surface temperature T of the current bat (t) Calculate the temperature difference T gap =T bat (t)-T env(t), and extract the maximum temperature difference T based on the temperature difference at each time point. gapmax .

[0108] Optionally, for the float charging duration, the float charging duration t at the time the battery health status was determined can be directly obtained. float_accum As a corresponding feature.

[0109] Furthermore, after step 21, filtering can be performed on the digital signal Y(t): EMD empirical mode decomposition can be performed on the digital signal Y(t) based on the feature data;

[0110] Where t = 1, 2, ..., T; T represents the total length of the digital signal; the n IMF components IMF1, IMF2, ..., IMFn and the margin Z of the digital signal are obtained;

[0111] Then, each IMF component is detected according to the time series t: the instantaneous frequency f of each IMF component is calculated respectively. k (t), where f k (t) represents the instantaneous frequency of the k-th IMF component;

[0112] Then, the obtained instantaneous frequency is compared with the set frequency threshold fTH:

[0113] When f k When (t)>fTH, then mark the current k-th IMF component as a high-frequency component; otherwise, when f k When (t)≤fTH, the current k-th IMF component is marked as a low-frequency component;

[0114] High-frequency noise detection is performed on the high-frequency components, and the high-frequency noise detection function used is:

[0115] Sech high (IMFk(t))=|H k (t)|-HTH

[0116]

[0117] Among them, Sech high (IMFk(t)) represents the high-frequency noise detection factor, when Sech high If H > 0, it indicates that the IMF component IMFk at time t is subject to high-frequency noise interference. k (t) represents the volatility coefficient, IMFk(t) represents the amplitude of the k-th IMF component at time t, and HTH represents the set volatility threshold.

[0118] When high-frequency noise interference is detected, the corresponding IMF component IMFk(t) is subjected to high-frequency filtering, where the high-frequency filtering function used is:

[0119]

[0120] Among them, IMFk ' f(t) represents the amplitude of the k-th IMF component at time t after high-frequency filtering, and IMFk(t) represents the amplitude of the k-th IMF component at time t. k (t) represents the instantaneous frequency of the k-th IMF component, fTH represents the set frequency threshold, and mm(t) represents the cumulative adjustment factor, the magnitude of which is the proportion of high-frequency noise interference moments contained in IMFk up to the current time t; H k (t) represents the volatility coefficient of the k-th IMF component at time t, and HTH represents the set volatility threshold.

[0121] Direct low-frequency filtering is performed on the low-frequency components, and the low-frequency filtering function used is as follows:

[0122]

[0123] Among them, IMFk ' (t) represents the amplitude of the k-th IMF component at time t after low-frequency filtering, IMFk(t) represents the amplitude of the k-th IMF component at time t, IMFk(v) represents the amplitude of the k-th IMF component at time v, where v∈[t-3,t+3], mean(*) represents the averaging function, and s1 and s2 represent the set higher-order correction factors.

[0124] After processing all time points sequentially, the data is reconstructed based on each IMF component (IMF1', IMF2', ..., IMFn') and the margin Z to obtain the digital signal of the filtered feature data.

[0125] In this implementation, EMD (Empirical Mode Decomposition) is first performed to obtain the decomposed IMF components, which are then used for noise detection and filtering. Taking into account the characteristics of environmental noise, the IMF components are first divided into high-frequency and low-frequency components based on instantaneous frequency. A progressive detection and filtering method is employed, which adaptively identifies the high- and low-frequency attributes of the IMF components, improving the adaptability of targeted high-frequency processing (avoiding the traditional method of using the overall signal as the criterion for high- and low-frequency IMF component division, which cannot adapt to situations where the high- and low-frequency characteristics of a particular IMF component change at a specific time point). For high-frequency IMF components, a high-frequency noise detection function is proposed, which can accurately identify impulse noise caused by the environment based on the instantaneous changes in the signal. This function can detect and process high-frequency noise with nonlinear characteristics of the characteristic data. Based on the detected noise points, a targeted high-frequency filtering function is proposed to eliminate the impulse noise. During high-frequency filtering, a cumulative adjustment factor is added to statistically analyze the cumulative noise interference received by the current IMF component, thereby adjusting the amplitude of subsequent high-frequency noise filtering and improving the noise elimination effect. Simultaneously, considering the potential low-frequency fluctuations in the battery during float charging, especially when battery health is poor, these fluctuations can affect the low-frequency signal. Therefore, further low-frequency filtering is applied to the low-frequency components. A low-frequency filtering function is proposed that can correct low-frequency components based on their trends, adaptively detecting and eliminating low-frequency noise, thereby improving the signal-to-noise ratio. Based on the propulsion detection and filtered IMF components, reconstruction is performed to obtain the filtered digital signals of the battery terminal voltage and float charging current. Further feature extraction is then performed, which helps improve the accuracy of subsequent feature extraction and feature vector set construction, indirectly improving the effectiveness and accuracy of subsequent battery health assessment and battery life prediction.

[0126] Step 22: Determine the battery capacity decay rate based on the first battery capacity and the second battery capacity of the distribution automation terminal battery before the first time period;

[0127] In this step, the battery capacity C at the time the battery health status determination is initiated is obtained. online Let C be the first battery capacity; and let C be the battery capacity of the distribution automation terminal battery before the first time period. T-1 This is recorded as the second battery capacity.

[0128] Furthermore, based on the difference between the capacity of the second battery and the capacity of the first battery, the battery capacity decay rate is determined.

[0129] Optional, battery capacity degradation rate ΔC rate The calculation formula is:

[0130] ΔC rate =(C T-1 -C online ) / C full

[0131] Among them, C full This is the maximum capacity of the battery.

[0132] Step 23: Determine the target monitoring data feature vector based on the digital signal, battery capacity decay rate, and first battery capacity.

[0133] In this step, based on the above, the current target monitoring data feature vector Φ can be constructed:

[0134] That is: Φ={C online ΔC rate , t float_accum , R(T), ΔR rate T gapmax Vb, V mean ,Ib,I mean}

[0135] The battery health status determination method provided in this application embodiment determines digital signals based on first monitoring data within a first time period. These digital signals include: the voltage fluctuation range and average voltage of the battery terminal voltage, the current fluctuation range and average current of the float charge current, the internal resistance value and internal resistance growth data of the battery internal resistance, the maximum temperature difference between the battery surface temperature and the ambient temperature at all sampling times, and the float charge duration. The battery capacity degradation rate is determined based on the first battery capacity and the second battery capacity of the distribution automation terminal battery before the first time period. A target monitoring data feature vector is determined based on the digital signals, the battery capacity degradation rate, and the first battery capacity. This technical solution extracts digital features and capacity degradation rates from historical battery operating data to construct a multi-dimensional target feature vector, which can more comprehensively quantify the battery performance degradation process. Furthermore, the structured feature vector is adapted to a parallel neural network model, synchronously outputting health assessment and lifespan prediction results, significantly improving the real-time performance of detection and the reliability of early warning.

[0136] Based on the above embodiments, Figure 3 A flowchart illustrating the method for determining battery health status provided in this application embodiment. Figure 3 ,like Figure 3 As shown, the steps preceding step 13 above may include:

[0137] It should be understood that there is no restriction on the execution order of steps 31 and 32.

[0138] Step 31: Train the ANN based on the feature vector of historical monitoring data and the actual battery health status results corresponding to the feature vector of historical monitoring data to obtain the health status assessment network in the battery health status assessment model.

[0139] In this step, after obtaining the historical monitoring data feature vector and the actual battery health status corresponding to the historical monitoring data feature vector, the ANN is trained. For example, the parameters of the ANN are adjusted based on the error between the actual battery health status and the battery health status prediction result output by the ANN, until the error between the actual battery health status and the battery health status prediction result output by the ANN converges. The converged ANN is then determined as the health status assessment network in the battery health status assessment model.

[0140] Optionally, the health assessment network is built on an ANN; the ANN includes a first input layer, a first hidden layer, a second hidden layer, and a first output layer.

[0141] In this implementation, the first input layer is used to obtain the monitoring data feature vector of the total input of the battery health status assessment model; the first hidden layer contains 128 neurons and uses the Rectified Linear Unit (ReLU) activation function; the second hidden layer contains 64 neurons and uses the ReLU activation function; the first output layer contains 1 neuron and outputs the battery health status assessment result.

[0142] The health status assessment network uses the mean squared error (MSE) loss function and the adaptive moment estimation (Adam) optimizer as the optimizer.

[0143] Step 32: Based on the feature vectors of historical monitoring data and the actual float life results corresponding to the feature vectors of historical monitoring data, train the LSTM to obtain the float life prediction network in the battery health status assessment model.

[0144] In this step, after obtaining the historical monitoring data feature vector and the corresponding actual float life results, the LSTM is trained. For example, the parameters of the LSTM are adjusted based on the error between the actual float life results and the float life prediction results output by the LSTM, until the error between the actual float life results and the float life prediction results output by the LSTM converges. The converged LSTM is then determined as the float life prediction network in the battery health status assessment model.

[0145] Optionally, the float charge lifetime prediction network is built on LSTM; the LSTM includes: a second input layer, a first convolutional layer, a second convolutional layer, a first LSTM layer, and a second output layer.

[0146] In this implementation, the second input layer is used to obtain the monitoring data feature vector of the total input of the battery health status assessment model; the first convolutional layer contains 64 convolutional kernels with a size of 3 and a stride of 1, and the activation function is ReLU; the second convolutional layer contains 32 convolutional kernels with a size of 3 and a stride of 1, and the activation function is ReLU; the first LSTM layer contains 128 LSTM units; the second output layer contains 1 neuron and outputs the float charge life prediction result.

[0147] The loss function used in the float charge lifetime prediction network is the MSE function, and the optimizer is the Adam optimizer.

[0148] Accordingly, the total input of the battery health status assessment model is connected to the first input layer and the second input layer, respectively; the total output of the battery health status assessment model is connected to the first output layer and the second output layer, respectively.

[0149] Optionally, the float charge lifetime prediction network includes: a first float charge lifetime prediction network and a second float charge lifetime prediction network.

[0150] 1) The first float charge lifetime prediction network is built on LSTM and includes: a second input layer, a first convolutional layer, a second convolutional layer, a first LSTM layer and a second output layer.

[0151] In this implementation, the second input layer is used to obtain the monitoring data feature vector of the total input of the battery health status assessment model; the first convolutional layer contains 64 convolutional kernels with a size of 3 and a stride of 1, and the activation function is ReLU; the second convolutional layer contains 32 convolutional kernels with a size of 3 and a stride of 1, and the activation function is ReLU; the first LSTM layer contains 128 LSTM units; the second output layer contains 1 neuron and outputs the float charge life prediction result.

[0152] The first float charge lifetime prediction network uses the MSE function as the loss function and the Adam optimizer as the optimizer.

[0153] 2) The second float-charge lifetime prediction network is built on LSTM and includes: a first buffer layer, a third input layer, a third convolutional layer, a fourth convolutional layer, a second LSTM layer, a third LSTM layer, and a third output layer.

[0154] In this implementation, the first caching layer stores the feature vector Φn-1 from the previous model input; the third input layer obtains the monitoring data feature vector Φn from the total input of the battery health assessment model, and constructs a new input vector Φ = {Φn-1, Φn} based on the feature vector Φn-1 and the feature vector Φn; the third convolutional layer contains 64 convolutional kernels of size 3 and stride 1, and uses the ReLU activation function; the fourth convolutional layer contains 32 convolutional kernels of size 3 and stride 1, and uses the ReLU activation function; the second LSTM layer and the third LSTM layer each contain 64 LSTM units; the third output layer contains one neuron, which outputs the second float life prediction result (actually Φ is a set);

[0155] The second float charge lifetime prediction network uses the MSE function as the loss function and the Adam optimizer as the optimizer.

[0156] Correspondingly, the total input of the battery health status assessment model is connected to the first input layer, the second input layer, and the third input layer, respectively; the total output of the battery health status assessment model is connected to the first output layer, the second output layer, and the third output layer, respectively; the first float life prediction result and the second float life prediction result obtained from the second output layer and the third output layer, respectively, are fused to obtain the final output float life prediction result.

[0157] The battery health status determination method provided in this application involves training an Artificial Neural Network (ANN) based on historical monitoring data feature vectors and the actual battery health status results corresponding to those feature vectors, thus obtaining a health status assessment network in the battery health status assessment model. Simultaneously, a Long Short-Term Memory (LSTM) network is trained based on historical monitoring data feature vectors and the actual float charge life results corresponding to those feature vectors, resulting in a float charge life prediction network in the same model. This technical solution constructs parallel health status assessment and float charge life prediction networks by training the ANN and LSTM separately. The ANN utilizes the mapping relationship between historical feature vectors and actual health results to learn the static degradation patterns of battery performance, making it suitable for real-time health scoring. The LSTM, on the other hand, captures the dynamic evolution characteristics of battery aging through time-series data analysis, improving the long-term accuracy of life prediction. The combination of the generalization ability of the ANN and the time-series modeling advantages of the LSTM avoids task conflicts associated with single models and comprehensively covers both the short-term state and long-term life of the battery through collaborative analysis. Continuous training with historical data allows for dynamic adaptation to different battery types or operating environments, significantly improving the reliability and practicality of the assessment results.

[0158] The following is a description of the device embodiments provided in this application.

[0159] Figure 4 This is a schematic diagram of the battery health status determination device provided in the embodiments of this application, as shown below. Figure 4 As shown, the battery health status determination device provided in this embodiment includes:

[0160] The acquisition module 41 is used to acquire the first monitoring data of the distribution automation terminal battery within a first time period and the second monitoring data at the current time. The first monitoring data includes: battery terminal voltage, float charging current, battery internal resistance, ambient temperature, battery surface temperature, and float charging duration. The second monitoring data includes: the first battery capacity. The first time period is a preset time period before the current time.

[0161] The determination module 42 is used to determine the target monitoring data feature vector based on the first monitoring data and the second monitoring data;

[0162] The processing module 43 is used to input the feature vector of the target monitoring data into the battery health status assessment model to obtain the battery health status assessment result and the float charge life prediction result. The battery health status assessment model includes a health status assessment network and a float charge life prediction network set in parallel.

[0163] In one or more embodiments, the acquisition module 41 acquires first monitoring data of the distribution automation terminal battery within a first time period, specifically for:

[0164] The system acquires and transmits first monitoring data collected and transmitted by a monitoring device embedded or externally located in the corresponding terminal of the power distribution automation terminal battery. The monitoring device includes a battery monitoring system, a temperature sensor, and an internal resistance tester.

[0165] In one or more embodiments, the acquisition module 41 acquires second monitoring data at the current time, specifically for:

[0166] In response to a request to determine battery health status, obtain the battery capacity collected and transmitted back by the online capacity assessment device;

[0167] The determination request is based on user input instructions or is triggered by a preset time interval.

[0168] In one or more embodiments, the determining module 42 is specifically used for:

[0169] Based on the first monitoring data within the first time period, digital signals are determined, including: the voltage fluctuation range and average voltage of the battery terminal voltage, the current fluctuation range and average current of the float charge current, the internal resistance value and internal resistance growth data of the battery internal resistance, the maximum temperature difference between the battery surface temperature and the ambient temperature at all sampling times, and the float charge duration.

[0170] The battery capacity decay rate is determined based on the first battery capacity and the second battery capacity of the distribution automation terminal battery before the first time period.

[0171] The target monitoring data feature vector is determined based on the digital signal, battery capacity decay rate, and first battery capacity.

[0172] In one or more embodiments, before inputting the target monitoring data feature vector into the battery health status assessment model to obtain the battery health status assessment result and float charge life prediction result, the determination module 42 is further configured to:

[0173] Based on the feature vectors of historical monitoring data and the actual battery health status results corresponding to the feature vectors of historical monitoring data, an artificial neural network (ANN) is trained to obtain the health status assessment network in the battery health status assessment model.

[0174] Based on the feature vectors of historical monitoring data and the actual float life results corresponding to the feature vectors of historical monitoring data, a Long Short-Term Memory (LSTM) network is trained to obtain the float life prediction network in the battery health status assessment model.

[0175] In one or more embodiments, the health status assessment network is built on an ANN; the ANN includes: a first input layer, a first hidden layer, a second hidden layer, and a first output layer;

[0176] The float charge lifetime prediction network is built on LSTM; the LSTM includes: a second input layer, a first convolutional layer, a second convolutional layer, a first LSTM layer, and a second output layer;

[0177] The total output of the battery health status assessment model is connected to the first output layer and the second output layer, respectively.

[0178] In one or more embodiments, the float charge lifetime prediction network includes: a first float charge lifetime prediction network and a second float charge lifetime prediction network.

[0179] The first float charge lifetime prediction network is built on LSTM and includes: a second input layer, a first convolutional layer, a second convolutional layer, a first LSTM layer, and a second output layer.

[0180] The second float-charge lifetime prediction network is built on LSTM and includes: a first buffer layer, a third input layer, a third convolutional layer, a fourth convolutional layer, a second LSTM layer, a third LSTM layer, and a third output layer.

[0181] The battery health status determination device provided in this embodiment can execute the battery health status determination method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0182] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 5 As shown, the electronic device provided in this embodiment includes:

[0183] At least one processor 51 and memory 52.

[0184] Optionally, the electronic device also includes a communication component 53. The processor 51, memory 52, and communication component 53 are connected via a bus 54.

[0185] In a specific implementation, at least one processor 51 executes computer execution instructions stored in memory 52, causing at least one processor 51 to perform the above-described method.

[0186] The specific implementation process of processor 51 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0187] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0188] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0189] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0190] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described method.

[0191] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0192] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0193] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0194] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0195] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0196] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0197] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0198] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0199] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A method for determining battery health status, characterized in that, include: The system acquires first monitoring data of the distribution automation terminal battery within a first time period and second monitoring data at the current time. The first monitoring data includes: battery terminal voltage, float charging current, battery internal resistance, ambient temperature, battery surface temperature, and float charging duration. The second monitoring data includes: first battery capacity. The first time period is a preset time period prior to the current time. Based on the first monitoring data and the second monitoring data, determine the target monitoring data feature vector; The target monitoring data feature vector is input into the battery health status assessment model to obtain the battery health status assessment result and the float charge life prediction result. The battery health status assessment model includes a health status assessment network and a float charge life prediction network set in parallel.

2. The method according to claim 1, characterized in that, The acquisition of the first monitoring data of the distribution automation terminal battery within the first time period includes: The system acquires and transmits first monitoring data collected and transmitted by a monitoring device embedded or externally located in the terminal corresponding to the battery of the power distribution automation terminal. The monitoring device includes a battery monitoring system, a temperature sensor, and an internal resistance tester.

3. The method according to claim 1, characterized in that, Obtaining the second monitoring data at the current time includes: In response to a request to determine battery health status, obtain the battery capacity collected and transmitted back by the online capacity assessment device; The determination request is based on user input instructions or is generated based on a preset time interval.

4. The method according to any one of claims 1-3, characterized in that, The step of determining the target monitoring data feature vector based on the first monitoring data and the second monitoring data includes: Based on the first monitoring data within the first time period, a digital signal is determined, which includes: the voltage fluctuation range and average voltage of the battery terminal voltage, the current fluctuation range and average current of the float charge current, the internal resistance value and internal resistance growth data of the battery internal resistance, the maximum temperature difference between the battery surface temperature and the ambient temperature at all sampling times, and the float charge duration. The battery capacity decay rate is determined based on the first battery capacity and the second battery capacity of the power distribution automation terminal battery before the first duration period. The target monitoring data feature vector is determined based on the digital signal, the battery capacity decay rate, and the first battery capacity.

5. The method according to claim 1, characterized in that, Before inputting the target monitoring data feature vector into the battery health status assessment model to obtain the battery health status assessment result and float charge life prediction result, the method further includes: Based on the feature vectors of historical monitoring data and the actual battery health status results corresponding to the feature vectors of historical monitoring data, an artificial neural network (ANN) is trained to obtain the health status assessment network in the battery health status assessment model. Based on the feature vectors of historical monitoring data and the actual float life results corresponding to the feature vectors of historical monitoring data, the Long Short-Term Memory (LSTM) network is trained to obtain the float life prediction network in the battery health status assessment model.

6. The method according to claim 1 or 5, characterized in that, The health status assessment network is built on an ANN; the ANN includes: a first input layer, a first hidden layer, a second hidden layer, and a first output layer; The float charge lifetime prediction network is built on LSTM; the LSTM includes: a second input layer, a first convolutional layer, a second convolutional layer, a first LSTM layer, and a second output layer; The total output of the battery health status assessment model is connected to the first output layer and the second output layer, respectively.

7. The method according to claim 1 or 5, characterized in that, The float charge lifetime prediction network includes: a first float charge lifetime prediction network and a second float charge lifetime prediction network. The first float charge lifetime prediction network is built on LSTM and includes: a second input layer, a first convolutional layer, a second convolutional layer, a first LSTM layer, and a second output layer. The second float-charge lifetime prediction network is built on LSTM and includes: a first buffer layer, a third input layer, a third convolutional layer, a fourth convolutional layer, a second LSTM layer, a third LSTM layer, and a third output layer.

8. A device for determining battery health status, characterized in that, include: The acquisition module is used to acquire first monitoring data of the distribution automation terminal battery within a first time period and second monitoring data at the current time. The first monitoring data includes: battery terminal voltage, float charging current, battery internal resistance, ambient temperature, battery surface temperature, and float charging duration. The second monitoring data includes: first battery capacity. The first time period is a preset time period before the current time. The determination module is used to determine the target monitoring data feature vector based on the first monitoring data and the second monitoring data; The processing module is used to input the feature vector of the target monitoring data into the battery health status assessment model to obtain the battery health status assessment result and the float charge life prediction result. The battery health status assessment model includes a health status assessment network and a float charge life prediction network set in parallel.

9. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-7.

Images