Intelligent operation and maintenance system and method for storage battery pack of thermal power plant based on machine learning

Abstract

The invention relates to the technical field of storage battery operation and maintenance, in particular to an intelligent operation and maintenance system and method for a storage battery pack of a thermal power plant based on machine learning, and provides the following scheme: constructing a three-dimensional virtual model of the storage battery pack, and arranging a monitoring equipment group to collect operation information of single batteries and the storage battery pack; the fault state of the single battery and the unbalance phenomenon in the battery pack are identified through clustering analysis and feature extraction; and detecting the abnormal state of the single battery and the battery pack based on the identification result, sending an early warning signal for a potential fault in the three-dimensional virtual model, and generating an operation and maintenance scheme including charging and discharging optimization and fault single body isolation. The method can improve the operation efficiency and safety of the storage battery pack, prolongs the service life of the storage battery pack, reduces the maintenance cost, has the advantages of high real-time performance, high adaptability and high intelligent degree, and is widely suitable for intelligent operation and maintenance of the storage battery pack of a thermal power plant.

Description

Technical Field

[0001] The present invention relates to the technical field of battery operation and maintenance, and in particular to an intelligent operation and maintenance system and method for battery groups in thermal power plants based on machine learning. Background Art

[0002] As an important energy storage device in thermal power plants, battery packs are widely used in backup power supply, frequency regulation and peak regulation, and stable power system. However, there are many shortcomings in the traditional operation and maintenance methods of battery packs, including the reliance on manual regular inspections for operation status monitoring, delayed fault identification, and the difficulty in effectively controlling the imbalance between batteries in the group. These shortcomings may lead to the degradation of battery pack performance, the spread of faults, and even safety accidents. In addition, with the increasing requirements of modern thermal power plants for system efficiency and safety, traditional operation and maintenance methods cannot meet the real-time monitoring and precise operation and maintenance needs of large-scale complex systems.

[0003] For example, the Chinese patent with authorization announcement number CN118655476B discloses an artificial intelligence-based battery pack operation and maintenance monitoring method and system, which belongs to the field of battery monitoring. The method includes: configuring the online monitoring equipment through the parameter configuration interface of the host computer; the online monitoring equipment collects the operating data and ambient temperature of the battery pack in real time to determine whether to issue an alarm, whether to perform active balancing, activation operations and online desulfurization; the host computer analyzes and processes the operating data and ambient temperature to obtain visual information; according to different battery pack pressure difference alarm thresholds and the corresponding battery pack pressure difference historical alarm frequencies, as well as the corresponding battery service life, the optimal battery pack pressure difference alarm threshold is obtained; based on the optimal battery pack pressure difference alarm threshold, the online monitoring equipment is configured. The present invention can greatly reduce the maintenance costs such as manpower and material resources, and improve the safety of battery use.

[0004] The above patents all have the problems raised by this background technology: it is difficult to effectively identify single-cell failures and imbalances within the group, resulting in low operation and maintenance efficiency, reduced battery performance and increased failure risks. To solve the above problems, this application designs an intelligent operation and maintenance system and method for battery groups in thermal power plants based on machine learning. Summary of the invention

[0005] The technical problem to be solved by the present invention is to address the deficiencies of the prior art and provide an intelligent operation and maintenance system and method for a battery pack in a thermal power plant based on machine learning. A three-dimensional virtual model of the battery pack is constructed and equipped with a monitoring equipment group to collect the operating information of battery cells and battery packs. The fault status of battery cells and the imbalance phenomenon in the battery pack are identified through cluster analysis and feature extraction. The abnormal status of the single cell and the battery pack is detected based on the identification results, and a warning signal is issued for potential faults in the three-dimensional virtual model, and an operation and maintenance plan including charge and discharge optimization and isolation of faulty cells is generated.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] An intelligent operation and maintenance method for a battery pack in a thermal power plant based on machine learning, the method comprising:

[0008] Construct a three-dimensional virtual model of the battery pack;

[0009] According to the three-dimensional virtual model of the battery pack, a monitoring equipment group is equipped, wherein the monitoring equipment group includes a battery cell monitoring equipment and a battery pack monitoring equipment;

[0010] Collecting battery cell operation information according to the battery cell monitoring device, and identifying the battery cell fault state according to the cell operation information;

[0011] Collecting battery pack operation information according to the battery pack monitoring device, and identifying the imbalance phenomenon between batteries according to the battery pack operation information;

[0012] Detecting abnormal conditions of single cells and battery packs based on the recognition results of the two devices, and issuing early warning signals for potential faults in the three-dimensional virtual model of the battery pack;

[0013] An operation and maintenance plan is generated through the operation and maintenance strategy, wherein the operation and maintenance plan includes a plan for adjusting the charging and discharging within the group and a plan for isolating a faulty single cell.

[0014] The method of constructing a three-dimensional virtual model of a battery pack includes:

[0015] Acquire design information and current structure information of each battery in the battery pack, wherein the design information includes technical drawings and specification information of the battery, and the current structure information includes a battery image;

[0016] Obtaining the battery main structure in the technical drawing, and visualizing the battery pack through modeling software according to the battery main structure and the specification information to obtain a battery pack main model;

[0017] Extracting features from the battery image by image processing technology, wherein the features include geometric features and connection features of the battery;

[0018] The geometric features and connection features are reconstructed through multi-scale convolution downsampling, compensated according to the reconstruction error function, spliced ​​through morphological operations, converted into point cloud data, and the point cloud data is imported into point cloud processing software for processing, mapped on the battery main body model, and output as a three-dimensional virtual model of the battery pack.

[0019] The identifying of the battery cell fault state includes:

[0020] Acquire battery standard operation information, perform cluster analysis on the battery monomer operation information and the battery standard operation information, and output a clustering result, wherein the clustering result includes an operation model, a charge and discharge behavior, and a dynamic characteristic group;

[0021] Calculating characteristic information of battery cell operation information according to the clustering results, wherein the characteristic information includes electrochemical characteristics, thermal characteristics and electrical energy characteristics;

[0022] A fault identification network is constructed, and the characteristic information is used as an input parameter of the fault identification network. The input parameter is trained through the fault identification network to output the fault status of the single cell.

[0023] The characteristic information of calculating the battery cell operation information includes:

[0024] Extract the voltage of the battery cell under different operating conditions according to the operating model, calculate the voltage deviation, calculate the internal resistance conversion rate by using Ohm's law through the dynamic current and voltage data in the charge and discharge behavior, analyze the charge and discharge curve offset of the battery cell during the charge and discharge process, evaluate the electrochemical efficiency of the battery, and thus calculate the electrochemical characteristics;

[0025] According to the dynamic characteristic group, the surface temperature change curve of the battery cell during operation is extracted, the temperature rise rate is calculated, and the time constant and heat diffusion rate of the temperature rise process are calculated to evaluate the heat dissipation characteristics of the battery, thereby calculating the thermal characteristic characteristics;

[0026] The charge and discharge energy ratios of battery cells in different operating models in the clustering results are extracted to evaluate the energy conversion efficiency. According to the dynamic characteristic group, the power output characteristics of the battery under short-term high-rate discharge are extracted. The capacity utilization and voltage platform attenuation trends at different discharge depths in the clustering results are analyzed to calculate the electric energy characteristics.

[0027] The fault identification network comprises:

[0028] An input layer, which establishes a feature sequence according to input parameters and calculates the fitting value of the feature sequence;

[0029] A hidden layer converts the fitting value of the feature sequence of the input layer into a high-dimensional space, and calculates the output value of the hidden layer in the high-dimensional space;

[0030] The output layer is used to output the fault status of the single battery of the fault identification network.

[0031] The identifying of the imbalance phenomenon between batteries includes:

[0032] Performing a residual energy test on the batteries in the battery pack, extracting a capacity loss characteristic parameter, and calculating the state of charge (SOC) of each battery according to the capacity loss characteristic parameter, and calculating the SOC balance according to the state of charge (SOC);

[0033] Calculating the operation balance of the battery pack according to the battery pack operation information, wherein the operation balance includes voltage balance, power balance and thermal balance;

[0034] The SOC balance and the operation balance are subjected to min-max normalization processing to generate characteristic parameters, an imbalance level is calculated according to the characteristic parameters, and an imbalance identification result of the battery pack is output.

[0035] The calculating of the state of charge SOC of each battery includes:

[0036] Under standard room temperature conditions, discharge the battery with a current of 1 / 4C. When the battery voltage is equal to the discharge cut-off voltage, stop discharging, record the data and leave it alone.

[0037] Under standard room temperature conditions, charge the battery with a current of 1 / 4C. When the battery voltage is equal to the charging cut-off voltage, switch to constant voltage charging and detect the charging current. When the charging current is equal to the charging cut-off current, stop charging, record the data and leave it alone.

[0038] Under standard room temperature conditions, discharge the fully charged battery at a current of 1 / 4C. When the battery voltage reaches the discharge cut-off voltage, stop discharging, record the data and leave it alone.

[0039] Under standard room temperature conditions, charge the discharged battery at a current of 1 / 4C. When the battery voltage is equal to the charging cut-off voltage, stop charging, record the data and leave it alone.

[0040] Perform curve fitting on the recorded data according to the pattern search nonlinear optimization algorithm, optimize the fitting results through the loss function, and output the capacity loss characteristic parameters;

[0041] A SOC evaluation model is constructed, and the core parameters and penalty factors of the SOC evaluation model are optimized by the moth optimization algorithm. The test set is evaluated according to the optimized core parameters and penalty factors, and the state of charge SOC of the battery is output.

[0042] The generating characteristic parameters include:

[0043] According to the simulated annealing algorithm, the boundary parameters of the SOC balance and operation balance after the standardization are optimized to determine the characteristic parameter boundaries;

[0044] According to the characteristic parameter boundary, SOC balance and operation balance, curve fitting is performed, the SOC balance and operation balance are converted into signals through wavelet transform, the signals are translated, the positions of discontinuities are changed, displayed in the curve, and characteristic parameters are extracted.

[0045] The specific calculation formula of the imbalance level is:

[0046]

[0047] Among them, Φ ib represents the imbalance level, n represents the total number of characteristic parameters, f represents a single characteristic parameter, ξ f represents the weight of the fth feature parameter, F f represents the normalized value of the fth feature parameter, Θ f represents the dynamic sensitivity coefficient of the fth characteristic parameter, g f represents the dynamic deviation of the fth characteristic parameter, ε represents the smoothing correction coefficient, and Y s represents the total battery operating time, d represents a single battery, m represents the total number of batteries, ΔS d represents the operating deviation value of the dth battery, ζd 表 shows the standard deviation of the dth battery during operation, Represents the ceiling function.

[0048] An intelligent operation and maintenance system for a battery pack in a thermal power plant based on machine learning, the system comprising a three-dimensional modeling module, a battery monitoring module, a data processing module, and a maintenance suggestion module;

[0049] The three-dimensional modeling module is used to obtain the design information and current structure information of each battery in the battery pack and build a three-dimensional virtual model of the battery pack;

[0050] The battery monitoring module is used to equip a monitoring equipment group according to the three-dimensional virtual model of the battery group, and collect battery operation information according to the monitoring equipment group;

[0051] The data processing module is used to identify the fault status of battery cells and the imbalance phenomenon between batteries;

[0052] The maintenance suggestion module is used to detect abnormal conditions of single cells and battery packs according to the identification results of the two devices, issue early warning signals for potential faults in the three-dimensional virtual model of the battery pack, and generate an operation and maintenance plan.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] 1. The present invention combines real-time monitoring, intelligent analysis and optimization algorithms to achieve accurate identification of battery cell failure states and battery pack imbalance phenomena, and generates targeted operation and maintenance solutions, significantly improving the operating efficiency and safety of the battery pack. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0056] Figure 1 This is a flow chart of an intelligent operation and maintenance method for a battery pack in a thermal power plant based on machine learning according to Embodiment 1 of the present invention;

[0057] Figure 2 This is a flow chart of cluster analysis of battery operation information in Example 1 of the present invention;

[0058] Figure 3 This is a fault identification network structure diagram of Example 1 of the present invention;

[0059] Figure 4 This is a module diagram of an intelligent operation and maintenance system for battery packs in a thermal power plant based on machine learning according to Example 2 of the present invention. DETAILED DESCRIPTION

[0060] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0061] Embodiment 1:

[0062] See also Figure 1 The present invention provides an embodiment: an intelligent operation and maintenance method for a battery pack in a thermal power plant based on machine learning, wherein the specific steps of the method are as follows:

[0063] S1: construct a three-dimensional virtual model of the battery pack;

[0064] In this step, the design information and current actual structure information of the battery pack are collected and combined with modeling technology to generate a three-dimensional virtual model. Specifically, point cloud processing and mapping technology are used to ensure the accuracy and visualization of the model, so that the operation and maintenance personnel can intuitively understand the physical structure and status of the battery pack.

[0065] S2: equipping a monitoring equipment group according to the three-dimensional virtual model of the battery group;

[0066] In this step, the key monitoring locations are determined by analyzing the three-dimensional virtual model and equipped with appropriate monitoring equipment groups. The monitoring equipment group includes battery cell monitoring equipment and battery pack monitoring equipment, which improves the scientific nature of the monitoring equipment layout, reduces redundancy and energy consumption, and improves the monitoring efficiency of the system.

[0067] S3: Collecting battery cell operation information and identifying battery cell fault status;

[0068] In this step, the battery cell monitoring equipment is used to obtain real-time operation information, including data such as voltage, current, internal resistance and temperature. The battery operation data is divided into different operation models (such as high-rate operation, low-rate operation and static state) through cluster analysis method. Further, the electrochemical characteristics, thermal characteristics and electrical energy characteristics of the battery are extracted to identify the fault status of the single cell, such as overcharge, over-discharge, excessive internal resistance or thermal runaway. Potential faults of battery cells can be discovered in advance to prevent the fault from spreading to the entire battery pack.

[0069] S4: Collecting battery pack operation information and identifying imbalance between batteries;

[0070] In this step, the battery pack monitoring equipment collects group operation data including voltage, current, temperature, etc., and combines SOC balance and operation balance (including voltage, power and thermal balance) to analyze the imbalance between batteries in the battery pack. Using standardization and optimization algorithms, the imbalance level results are further generated to locate the problematic battery cells. Avoid excessive load or early degradation of local batteries caused by imbalance in the group, and extend the overall life of the battery pack.

[0071] S5: Detecting abnormal conditions of single cells and battery packs, and issuing early warning signals for potential faults in the three-dimensional virtual model of the battery pack;

[0072] S6: Generate an operation and maintenance plan based on the operation and maintenance strategy;

[0073] In this step, the operation and maintenance plan includes the charge and discharge optimization strategy and the faulty cell isolation strategy. For example, the balanced charge and discharge strategy can be used to improve the performance difference of the batteries in the group, or the faulty cell can be isolated to prevent it from affecting the normal operation of other batteries.

[0074] The specific steps of S1 are as follows:

[0075] S1.1: Acquire design information and current structure information of each battery in the battery pack, wherein the design information includes technical drawings and specification information of the battery, and the current structure information includes a battery image;

[0076] Traditional battery pack modeling often relies only on design information, ignoring the changes that may occur in the battery during actual operation. By introducing current structural information, especially images and 3D scanning data, the actual state of the battery can be reflected, improving the accuracy and reliability of the model.

[0077] Specifically, design information includes technical drawings, specification information, and assembly logic of battery cells. These data can be extracted from technical information or related databases provided by manufacturers, covering the physical dimensions of the battery, the location of the connection interface, the internal structure description, and standard operating parameters. Current structural information is collected through industrial imaging equipment, mainly including the appearance image of the battery cell.

[0078] S1.2: Obtain the battery main structure in the technical drawing, visualize the battery pack through modeling software according to the battery main structure and the specification information, and obtain the battery pack main model;

[0079] The main structure includes the battery's appearance, terminal position and assembly sequence. Combined with specification information (such as rated voltage, capacity, etc.), the geometric structure and electrical connection relationship of the battery pack are visualized through modeling software to generate a preliminary battery pack main model.

[0080] S1.3: extracting features from the battery image by image processing technology, wherein the features include geometric features and connection features of the battery;

[0081] The geometric features include the appearance of the battery, and the connection features include the location of the battery terminals and the connector status. Through edge detection, the outline and geometric boundaries of the battery are extracted to correct the geometric deviation in the design model. Then the YOLO network is used to identify the location of the battery terminals and connectors to ensure the accuracy of the connection features.

[0082] S1.4: Reconstruct the geometric features and connection features through multi-scale convolution downsampling, compensate according to the reconstruction error function, splice through morphological operations, convert into point cloud data, import the point cloud data into point cloud processing software for processing, map it on the battery main body model, and output the three-dimensional virtual model of the battery pack.

[0083] Specifically, convolution kernels of different scales are used to perform multi-level extraction of feature data, retaining global structural features while highlighting local details. The point cloud data is repaired through morphological algorithms to fill the holes and missing areas in the sampling, and texture mapping is performed on the surface of the point cloud model. The battery image is accurately superimposed on the three-dimensional model to improve the authenticity of the model.

[0084] The specific steps of S3 are as follows:

[0085] S3.1: Obtaining battery standard operation information, performing cluster analysis on the battery cell operation information and the battery standard operation information, and outputting cluster results, wherein the cluster results include operation models, charge and discharge behaviors, and dynamic characteristic groups;

[0086] First, the real-time operating information of each battery cell is collected through monitoring equipment, including voltage, current, internal resistance and temperature parameters. At the same time, the design reference values ​​(such as rated voltage, rated capacity, charge and discharge rate limit, temperature range) are extracted from the standard operating information of the battery. The real-time operating information is compared and normalized with the standard operating information to eliminate the dimensional differences between the features and ensure the accuracy of the subsequent analysis results.

[0087] See also Figure 2 , a flow chart of cluster analysis of battery operation information according to an embodiment of the present invention, the specific steps of cluster analysis are as follows:

[0088] S3.1.1: Perform preliminary clustering of the battery standard operation information according to the DBSCAN clustering algorithm to obtain core points, and determine the number of cluster center clusters according to the number of core points;

[0089] S3.1.2: Use the core points after DBSCAN clustering as the initial center clusters of the K-means clustering algorithm, and calculate the Euclidean distance from the battery cell operation information to each initial center cluster;

[0090] S3.1.3: Perform clustering according to Euclidean distance, calculate the average clustering error of the initial center cluster, and update the center cluster according to the average clustering error;

[0091] S3.1.4: Define the convergence condition and determine whether the updated center cluster meets the convergence condition. If it does not meet the convergence condition, continue to update the center cluster. If it meets the convergence condition, output the clustering result.

[0092] Specifically, by combining the two clustering algorithms, DBSCAN and K-means, the standard working operation information of the battery is clustered and analyzed. By iteratively optimizing the central cluster, a more accurate clustering result is obtained, so as to better understand the working status of the single battery, which helps to improve the operating efficiency of the battery pack and reduce energy loss. DBSCAN is a density clustering algorithm used to identify closely connected data points in space and divide them into clusters. DBSCAN can be used for preliminary clustering to obtain core points. Core points refer to the core area containing at least a specified number of data points within a given radius. Since each core point may represent a cluster, the number of core points can be used to preliminarily determine the number of central clusters of clustering.

[0093] The kernel points obtained by preliminary clustering through DBSCAN are used as the initial center cluster of the K-means clustering algorithm, providing an initial cluster center position for the K-means algorithm; the Euclidean distance is used for clustering, and each working operation information is assigned to the nearest center cluster by calculating the Euclidean distance from the battery cell operation information to each initial center cluster. The clustering average error refers to the average value of the Euclidean distance from all the data points to the center. The clustering average error can be used to update the initial center cluster to improve the clustering effect. The position of the center cluster can be adjusted according to the calculated clustering average error, so as to update the center cluster, and the convergence condition is defined to determine whether the updated center cluster meets the condition to determine whether to end the iteration. If the convergence condition is met, the clustering result is output, including the operation model, charging and discharging behavior and dynamic characteristic group.

[0094] The operating model is used to describe the current operating state of the battery, such as static mode, high-rate charge and discharge mode, low-rate charge and discharge mode, etc., reflecting the characteristics of the battery under different working conditions. The charge and discharge behavior is used to describe the characteristics of the battery charge and discharge curve, such as current response time, voltage change curve slope, etc., to identify whether there are abnormalities in the charge and discharge process. The dynamic characteristic group is classified based on the characteristic changes of the battery in different time windows (such as temperature change trend, internal resistance growth rate, etc.) to identify potential degradation problems that may exist.

[0095] S3.2: Calculating characteristic information of battery cell operation information according to the clustering result, wherein the characteristic information includes electrochemical characteristics, thermal characteristics and electrical energy characteristics;

[0096] Specifically, the electrochemical characteristics include voltage characteristics, internal resistance characteristics, and charge and discharge offset characteristics. The voltage characteristics are used to extract the voltage platform values ​​(such as the charge voltage platform and the discharge voltage platform) under different operating conditions in the operating model group, calculate the voltage fluctuation amplitude and voltage recovery time, and evaluate the voltage stability of the battery. The internal resistance characteristics use Ohm's law to calculate the dynamic internal resistance value of the battery through dynamic current and voltage data, and further analyze the internal resistance growth rate and internal resistance fluctuation range as core indicators of battery aging and failure. The charge and discharge offset is specifically used to combine the data of the charge and discharge behavior group to extract the offset of the battery's charge and discharge curve and evaluate its energy conversion efficiency and electrochemical activity changes.

[0097] Thermal characteristics include temperature rise characteristics, thermal balance and thermal stability. The temperature rise characteristics are based on the dynamic characteristics group data, analyze the temperature change curve of the battery surface, calculate the temperature rise rate and thermal diffusion rate, and are used to evaluate the heat dissipation performance of the battery. Thermal balance extracts the maximum deviation and mean deviation of the temperature of each single battery in the group to analyze whether the battery has a risk of thermal runaway. Thermal stability calculates the thermal diffusion time constant through the temperature change curve over time to evaluate the thermal dynamic performance of the battery.

[0098] The electrical energy characteristics include energy conversion efficiency, power output characteristics and capacity utilization. The energy conversion efficiency extracts the charge and discharge energy ratio in the operating model group to evaluate the energy conversion efficiency of the battery at different rates. The power output characteristics are combined with the dynamic characteristics group to analyze the peak power output under short-term high-rate discharge conditions to evaluate the transient performance of the battery. The capacity utilization analyzes the capacity attenuation curve based on the discharge depth to identify the impact of deep discharge on the battery capacity.

[0099] The specific steps of feature information extraction are as follows:

[0100] S3.2.1: According to the operation model, the voltage of the battery cell under different operating conditions is extracted, the voltage deviation is calculated, the internal resistance conversion rate is calculated using Ohm's law through the dynamic current and voltage data in the charge and discharge behavior, the charge and discharge curve offset of the battery cell during the charge and discharge process is analyzed, and the electrochemical efficiency of the battery is evaluated, thereby calculating the electrochemical characteristics. The calculation formula of the electrochemical characteristics is:

[0101]

[0102] Where η represents the electrochemical characteristics, V o (·) represents the time variation function of the open circuit voltage of the battery during operation, I(·) represents the dynamic current curve of the battery during the charge and discharge process, R(·) represents the time variation function of the internal resistance, ΔV represents the voltage offset of the charge and discharge curves, which is calculated as the difference between the average values ​​of the charge curve and the discharge curve, V m represents the average value of the open circuit voltage, ΔR represents the change rate of the internal resistance, R m Indicates the average value of internal resistance, t 1 and t 2 Indicates the start and end time of the charge and discharge behavior;

[0103] Specifically, according to the operating model, real-time voltage data of battery cells under different operating conditions such as high rate, low rate, and static are collected. During the charging and discharging process, the curve of voltage change over time is recorded, and key parameters such as charging platform voltage, discharging platform voltage, peak voltage, and minimum voltage are extracted. The voltage deviation is calculated from these data, specifically the difference between the voltage of the single cell and the average voltage in the group. Using dynamic current and voltage data, the rate of change of internal resistance is calculated by Ohm's law, focusing on the nonlinear change of internal resistance under different current intensities. Further, the internal resistance conversion rate is combined with the voltage curve to analyze the degree of curve deviation during charging and discharging, such as voltage hysteresis. When evaluating the electrochemical efficiency of a battery cell, the ratio of the energy lost during charging and discharging to the total input energy is combined to determine whether the electrochemical performance meets the design requirements.

[0104] Battery voltage, internal resistance and curve deviation are the core parameters that reflect electrochemical characteristics. They can reveal the degree of battery aging, the uniformity of electrochemical reaction rates, and differences in charge and discharge efficiency. By analyzing voltage deviation and internal resistance conversion rate, battery cells with potential problems can be quickly identified, such as batteries with excessive polarization, effectively improving fault prediction capabilities.

[0105] S3.2.2: Extract the change curve of the surface temperature of the battery cell during operation according to the dynamic characteristic group, calculate the temperature rise rate, and calculate the time constant and heat diffusion rate of the temperature rise process, evaluate the heat dissipation characteristics of the battery, and then calculate the thermal characteristic characteristics. The calculation formula of the thermal characteristic characteristics is:

[0106]

[0107] Among them, H represents the thermal characteristic, ΔT represents the maximum temperature difference of the battery surface temperature rise, represents the rate of temperature rise, represents the Laplace operator of the temperature field, α represents the thermal diffusion coefficient of the battery, represents the second-order time derivative of temperature, C represents the specific heat capacity of the battery, ρ represents the battery density, ΔQ represents the heat generated by the battery, γ represents the heat source phase coefficient of the battery, β represents the temperature rise influence coefficient of the battery, τ represents the time constant of temperature rise in the thermal diffusion process, s 1 and 2 Indicates the start and end time of the battery temperature rise process;

[0108] The thermal characteristics of batteries during operation are directly related to their safety and long-term performance. Especially at high-rate discharge or high ambient temperature, the risk of thermal runaway increases significantly.

[0109] Specifically, the surface temperature data of the battery cell during operation is collected, and the curve of temperature change over time is recorded. Especially in the high-rate charge and discharge stage, the temperature rise rate, that is, the rate of temperature rise, is extracted. Combined with the temperature change data, the time constant of the temperature rise process is calculated to reflect the speed at which heat diffuses from the battery surface to the surrounding environment. Furthermore, by analyzing the distribution of temperature gradients, the heat diffusion rate is evaluated to identify whether there is uneven heat dissipation capacity inside the battery. Under specific environmental conditions (such as different heat dissipation designs or changes in external temperature), the temperature rise rate is combined with the fluctuation trend of the time constant to establish a thermal characteristic evaluation model for predicting the risk of thermal runaway.

[0110] Compared with existing methods that rely on peak analysis of surface temperature, this method can more accurately locate the source of thermal runaway risk (such as local hot spots or insufficient heat dissipation efficiency) through comprehensive calculation of time constants and heat diffusion rates, effectively identify battery cells with poor heat dissipation, thereby preventing battery aging or failure caused by heat accumulation and improving the operating safety of the battery pack.

[0111] S3.2.3: Extract the charge-discharge energy ratio of the battery cells in different operating models in the clustering results, evaluate the energy conversion efficiency, extract the power output characteristics of the battery under short-term high-rate discharge according to the dynamic characteristic group, analyze the capacity utilization and voltage platform attenuation trend under different discharge depths in the clustering results, and calculate the electric energy characteristics. The calculation formula of the electric energy characteristics is:

[0112]

[0113] Where E represents the electrical energy characteristic, T 1 and T 2 represents the start time and end time of the battery in the operation model, V(T) represents the voltage change function of the battery in the operation model, CO(T) represents the energy conversion function during the battery charging and discharging process, A D Indicates the influencing factor of discharge depth, ΔC indicates the instantaneous capacity attenuation of the battery, C r Indicates the rated capacity of the battery, T 3 and T 4 Indicates the time interval of short-term high-rate discharge of the battery, P p (T) represents the instantaneous peak value conversion function of the battery, σ R Represents the battery internal resistance sensitivity coefficient, R ce Represents the instantaneous resistance of the battery, P m represents the average power in the short-time discharge process, and θ represents the weight factor of the short-time power characteristic.

[0114] The charge-discharge energy ratio, voltage platform change and power characteristics are the core indicators for evaluating the energy performance and dynamic response capability of batteries. These indicators not only reflect the current performance of the battery, but also predict its future performance degradation.

[0115] Specifically, the energy conversion efficiency is calculated by analyzing the input energy and output energy of the battery under different operating models (such as high-rate charging and low-rate discharging) in the clustering results. Under short-term high-rate discharge, the instantaneous power output characteristics are extracted, and the peak power, response time and power decay curve of the battery are analyzed in detail to evaluate the transient performance of the battery. Furthermore, the clustering results are combined to analyze the impact of different discharge depths on capacity utilization and voltage platforms, such as extracting the rate of decrease of the voltage platform and the changing trend of capacity utilization, reflecting the aging process of the battery in long-term cycles.

[0116] Compared with the traditional method that only relies on a single power curve analysis, this method can more comprehensively reflect the long-term energy characteristics and dynamic response capabilities of the battery through a comprehensive evaluation of the charge-discharge ratio combined with the discharge depth. It can identify battery cells with low energy efficiency, provide a reliable basis for the charge and discharge optimization of the battery pack, and predict the capacity attenuation trend of the battery in long-term cycles.

[0117] S3.3: construct a fault identification network, and use the characteristic information as input parameters of the fault identification network, train the input parameters through the fault identification network, and output the fault status of the single battery.

[0118] See also Figure 3 , a structural diagram of a fault identification network according to an embodiment of the present invention, wherein the fault identification network comprises:

[0119] An input layer, which establishes a feature sequence according to input parameters and calculates the fitting value of the feature sequence;

[0120] The input layer collects the operating information of the battery cells, including electrochemical characteristics, thermal characteristics and electrical characteristics, and uses standardized processing methods to normalize the characteristic data of different dimensions to ensure that the numerical range of each feature is consistent and eliminate the deviation caused by the difference in the characteristic value range. At the same time, in order to improve the data fitting ability, the time series sliding window mechanism is introduced to take the dynamic changes of each feature as the time series feature to form a complete feature sequence.

[0121] After the feature sequence is normalized, it is further fitted by the least square method to generate a fitted value sequence. The fitted value is used to describe the change trend of each feature in a specific time period and provide an optimized data basis for the high-dimensional space calculation of the hidden layer.

[0122] A hidden layer converts the fitting value of the feature sequence of the input layer into a high-dimensional space, and calculates the output value of the hidden layer in the high-dimensional space;

[0123] The hidden layer receives the fitted value sequence of the input layer and maps it to a high-dimensional space through an activation function. The hidden layer adopts a stacked structure and contains multiple layers of nonlinear activation units. The specific activation function is ReLU (rectified linear unit), whose nonlinear characteristics can capture complex feature interactions while avoiding the gradient vanishing problem.

[0124] In high-dimensional space, the hidden layer performs deep feature mining on the input data, focusing on the coupling relationship between the following features:

[0125] Coupling of electrochemical characteristics and thermal properties: such as the effect of internal resistance growth rate on temperature rise rate.

[0126] Coupling of thermal and electrical characteristics: such as the dynamic impact of temperature changes on power output efficiency.

[0127] Joint analysis of cross-dimensional features: such as the correlation between voltage platform changes and energy conversion efficiency.

[0128] Specifically, the input features are weighted using the weight matrix, and the training efficiency and generalization ability of the model are improved through batch normalization technology. Finally, the output value of the hidden layer is used as the intermediate result of further calculation of the network, which contains the deep information of multi-dimensional features.

[0129] The output layer is used to output the fault status of the single battery of the fault identification network.

[0130] The output layer receives the output value of the hidden layer and maps the high-dimensional features to the fault status of the single battery through the classifier. The classifier uses the Softmax function to process the probability of each fault category and finally outputs the fault classification result.

[0131] Fault conditions include:

[0132] Normal status: The operating parameters of the single battery are within the normal range, with no obvious abnormalities.

[0133] Minor faults: such as slight overcharge, over-discharge or slight increase in internal resistance, have limited impact on battery performance.

[0134] Serious failures: such as a sharp increase in internal resistance, voltage collapse, or thermal runaway risk caused by overcharging or over-discharging.

[0135] To improve the robustness of fault identification, the output layer adds confidence analysis to the classification results of the fault state. The classification results with low confidence will be returned to the hidden layer for recalculation, and the weight of the activation function will be updated according to the confidence analysis results to ensure the reliability of the output results. At the same time, the output layer will perform correlation analysis on the fault classification results and feature sequences to generate an explanatory report on the fault cause and fault location.

[0136] Specifically, the fault identification network is provided with an input layer, a hidden layer and an output layer. The input layer is used to establish a feature sequence according to the input parameters of the fault identification network, accumulate the feature sequence, establish a grey differential equation, calculate the least squares parameter of the grey differential equation, and calculate the fitting value of the feature sequence by discretizing the least squares parameter solution. The hidden layer is used to transform the feature sequence fitting value of the input layer into a high-dimensional space through mapping according to the activation function between the input layer and the hidden layer, and sum the feature sequence fitting value in the high-dimensional space after full matrix nonlinear weighting to calculate the output value of the hidden layer. The output layer is used to obtain the output layer input, classify the state through the classifier and confidence analysis, and output it.

[0137] The specific steps of S4 are as follows:

[0138] S4.1: Performing a remaining energy test on the batteries in the battery pack, extracting a capacity loss characteristic parameter, and calculating a state of charge (SOC) of each battery according to the capacity loss characteristic parameter, and calculating SOC balance according to the state of charge (SOC);

[0139] Specifically, by testing the remaining energy of each battery cell, the actual capacity information during the discharge process is fully collected. The specific operation is to use a standard charge and discharge process: charge and discharge the battery with a fixed current, record the charge and discharge curve and actual output capacity of the battery, and compare the actual capacity with the rated capacity of the battery to extract the capacity loss characteristic parameters of the battery. At the same time, combined with the recorded discharge termination voltage, discharge depth and remaining power, a nonlinear optimization algorithm is used to curve fit these data to further optimize the accuracy of the capacity loss characteristic parameters. According to the extracted capacity loss characteristic parameters, the SOC of each battery cell is calculated by establishing a charge state evaluation model. The calculation of SOC not only considers the current battery charge and discharge curve characteristics, but also combines the battery's historical operating data and capacity attenuation trend to ensure the accuracy of the evaluation. Subsequently, the SOC of all single cells in the battery pack is statistically analyzed to calculate the SOC balance indicators, including the standard deviation, maximum deviation and mean deviation of SOC. These indicators comprehensively reflect the charge state distribution of each single cell in the battery pack and identify possible imbalance problems.

[0140] Compared with the existing technology, this method improves the accuracy of SOC calculation by combining capacity loss characteristic parameters and nonlinear optimization algorithms. It is especially suitable for battery packs with severe capacity decay after long-term operation, and solves the problem that traditional SOC evaluation methods are difficult to deal with decay errors.

[0141] The specific steps for calculating the state of charge SOC are as follows:

[0142] S4.1.1: At standard room temperature, discharge the battery at a current of 1 / 4C. When the battery voltage reaches the discharge cut-off voltage, stop discharging, record the data and leave it alone.

[0143] S4.1.2: Under standard room temperature conditions, charge the battery at a current of 1 / 4C. When the battery voltage is equal to the charge cut-off voltage, switch to constant voltage charging and detect the charging current. When the charging current is equal to the charge cut-off current, stop charging, record the data and leave it alone.

[0144] S4.1.3: At standard room temperature, discharge the fully charged battery at a current of 1 / 4C. When the battery voltage reaches the discharge cut-off voltage, stop discharging, record the data and leave it alone.

[0145] S4.:1.4: Under standard room temperature conditions, charge the discharged battery at a current of 1 / 4C. When the battery voltage is equal to the charging cut-off voltage, stop charging, record the data and leave it alone.

[0146] S4.1.5: Perform curve fitting on the recorded data according to the pattern search nonlinear optimization algorithm, optimize the fitting results through the loss function, and output the capacity loss characteristic parameters. The calculation formula of the loss function is:

[0147]

[0148] Where L represents the loss function, i represents the measurement time point in the residual energy test, J represents the total measurement time point in the residual energy test, G{·} represents the annealing genetic function, and K b represents the Boltzmann constant, y model [·] represents the pattern search output function, U p represents the positive electrode potential, p o represents the initial conversion rate of the positive electrode, l i represents the current at the i-th measurement time point, M p Indicates the initial capacity of the positive electrode, U n represents the negative electrode potential, n o Indicates the initial conversion rate of the negative electrode, M n represents the initial capacity of the negative electrode, represents the model parameter matrix, u i represents the voltage at the i-th measurement time point;

[0149] The collected voltage, current and capacity data are curve-fitted using a pattern search nonlinear optimization algorithm. The fitting curve describes the nonlinear relationship between SOC and voltage. The fitting result is optimized by designing a loss function to minimize the fitting error, and finally the capacity loss characteristic parameters of the battery (such as capacity attenuation rate and dynamic response characteristics) are obtained.

[0150] S4.1.6: Construct a SOC evaluation model, optimize the core parameters and penalty factors of the SOC evaluation model through the moth optimization algorithm, evaluate the test set according to the optimized core parameters and penalty factors, and output the battery's state of charge SOC.

[0151] Specifically, before evaluation, it is necessary to prepare training set and test set data. The data sources include the battery charge and discharge cycle test results and dynamic information such as voltage, current, temperature and time collected under different working conditions. The SOC evaluation model is a nonlinear regression model that uses the support vector regression method to capture the complex nonlinear relationship between input variables and SOC. The kernel function, penalty factor and kernel parameter are set in the model. The kernel function is used to map the input variables to the high-dimensional feature space. The penalty factor is used to balance the model complexity and fitting accuracy, control overfitting and underfitting phenomena, and the kernel parameters determine the feature mapping effect in the high-dimensional space and affect the prediction ability of the model. The moth optimization algorithm is a population-based intelligent optimization algorithm that simulates the behavior of moths looking for flames and is used to optimize the kernel parameters and penalty factors of the SOC evaluation model. First, a set of solutions is randomly generated, each solution represents a set of kernel parameters and penalty factors. For each moth parameter combination, a SOC evaluation model is constructed, and the MSE value is calculated on the training set and the validation set as the fitness. According to the position of the flame (i.e., the current optimal solution), the next flight direction of the moth is adjusted, and the moth position is repeatedly updated until the set number of iterations or error threshold is reached, and the optimal kernel parameters and penalty factors are output.

[0152] S4.2: Calculating the operation balance of the battery pack according to the battery pack operation information, wherein the operation balance includes voltage balance, power balance and thermal balance;

[0153] The battery pack monitoring equipment collects the operating information of the battery pack in real time, including the real-time voltage, charge and discharge current, temperature and output power of each battery cell. Based on these collected data, the multi-dimensional balance indicators of the battery pack are calculated.

[0154] Voltage balance is used to analyze the real-time voltage data of all battery cells in the group, calculate the maximum, minimum and mean voltage of the battery group, and further calculate the standard deviation and deviation range of the voltage. Voltage balance can reflect the balance of voltage distribution in the battery group. Large voltage deviation may cause some battery cells to be overcharged or over-discharged.

[0155] Power balance calculates the output power of each battery cell through real-time current and voltage data, and evaluates the power balance of the battery pack based on the unevenness of power distribution (such as power deviation rate and maximum power difference). Uneven power distribution may cause some batteries to be overloaded and accelerate their aging.

[0156] Thermal balance obtains the surface temperature data of each battery cell and calculates the maximum temperature difference, temperature rise rate and heat diffusion rate of the battery pack. Thermal balance is used to identify heat dissipation problems within the battery pack and prevent local overheating from causing thermal runaway.

[0157] Specifically, the balance of the operating state of the battery pack is comprehensively evaluated through the balance calculation of the three dimensions of voltage, power and heat. The dynamic change characteristics of thermal balance are taken into account, and the early signals of thermal runaway can be effectively captured through the calculation of temperature rise rate and heat diffusion rate, thereby improving the safety assessment capability of the battery pack operating state.

[0158] S4.3: performing min-max normalization processing on the SOC balance and the operation balance to generate characteristic parameters, calculating the imbalance level according to the characteristic parameters, and outputting the imbalance identification result of the battery pack.

[0159] The generating characteristic parameters include:

[0160] According to the simulated annealing algorithm, the boundary parameters of the SOC balance and operation balance after the standardization are optimized to determine the characteristic parameter boundaries;

[0161] The simulated annealing algorithm is a global optimization algorithm that gradually finds the global optimal solution by simulating the energy changes in the physical annealing process. In this embodiment, according to the distribution characteristics of SOC balance and operation balance, the upper and lower limits of the feature parameters are adaptively adjusted to dynamically define the valid range of the data. This optimization method can avoid the misjudgment problem caused by fixed boundary parameters. By iteratively adjusting the boundary parameters, the influence of noise and outliers is filtered out, and the robustness and adaptability of feature extraction are improved.

[0162] Specifically, compared with the traditional fixed boundary parameter setting method, the simulated annealing algorithm can dynamically adjust the boundary range according to real-time data, significantly improving the accuracy of characteristic parameter calculation, especially when dealing with complex and irregular data distribution. In addition, through the combination of standardization and boundary optimization, the differences in the operating status of battery packs can be better captured.

[0163] According to the characteristic parameter boundary, SOC balance and operation balance, curve fitting is performed, the SOC balance and operation balance are converted into signals through wavelet transform, the signals are translated, the positions of discontinuities are changed, displayed in the curve, and characteristic parameters are extracted.

[0164] The fitting process uses a high-order polynomial or piecewise linear fitting method to convert the discrete raw data into a continuous curve expression. The main purpose of curve fitting is to eliminate the nonlinear characteristics and local discontinuities in the data, so as to more intuitively reflect the changes in the balance state of the battery pack. At the same time, in order to further analyze the correlation between SOC balance and operation balance, the wavelet transform technology is used to convert the fitted curve into a signal form. Wavelet transform is a mathematical tool that decomposes a signal into wavelet bases of multiple scales, which can effectively capture local features and mutation points in the signal. In this embodiment, local mutation points and fluctuation characteristics in the SOC balance and operation balance curves are obtained. These characteristics are important manifestations of the imbalance phenomenon of the battery pack. The signal after wavelet decomposition is translated, and the position of the discontinuous point is adjusted so that it is presented in a significant way in the curve. This processing method can enhance the visual expression of the unbalanced state. After curve fitting and wavelet transform processing, key feature parameters are extracted from the signal. The specific extraction content includes:

[0165] Amplitude characteristics: Extract the maximum value, minimum value and difference in the SOC balance and operation balance signals as the main indicators reflecting the difference in balance status.

[0166] Frequency domain features: Perform frequency domain analysis on the signal to extract the energy distribution within a specific frequency band to identify the periodic fluctuation characteristics in the balanced state of the battery pack.

[0167] Time domain characteristics: Calculate the time distribution characteristics of the signal, including the mean, variance, and skewness coefficient, to quantify the overall balance level of the battery pack operating status.

[0168] The specific calculation formula of the imbalance level is:

[0169]

[0170] Among them, Φ ib represents the imbalance level, n represents the total number of characteristic parameters, f represents a single characteristic parameter, ξ f represents the weight of the fth feature parameter, F f represents the normalized value of the fth feature parameter, Θ f represents the dynamic sensitivity coefficient of the fth characteristic parameter, g f represents the dynamic deviation of the fth characteristic parameter, ε represents the smoothing correction coefficient, and Y s represents the total battery operating time, d represents a single battery, m represents the total number of batteries, ΔS d represents the running deviation value of the dth battery, ζ d represents the standard deviation of the dth battery during operation, Represents the ceiling function.

[0171] Specifically, the calculation results are divided into normal, slightly unbalanced and severely unbalanced levels by pre-set thresholds. By outputting the imbalance identification results of the battery pack, a detailed analysis report is generated, including the overall imbalance level of the battery pack, the specific values ​​of the SOC deviation, voltage deviation and temperature deviation of each single cell, and the main source of the imbalance problem.

[0172] The operation and maintenance strategy is based on the above-collected single cell operation information and battery pack operation information, and generates an operation and maintenance plan through intelligent analysis and optimization algorithm. The operation and maintenance plan includes adjusting the charging and discharging strategy within the group and the faulty single cell isolation plan, specifically including:

[0173] By real-time monitoring and historical data analysis of key parameters such as SOC (state of charge), voltage, internal resistance, and temperature of each single cell in the battery pack, key factors that cause imbalance in the pack, such as differences in charge and discharge rates, uneven temperature rise, and mismatched internal resistance, can be identified;

[0174] Specifically, it includes charging phase optimization and discharging phase optimization;

[0175] For detected faulty single cells (such as cells with excessively high internal resistance, rapid capacity decay, or a high risk of thermal runaway), a combination of physical and electronic isolation is used to ensure that the faulty cells do not affect the normal operation of the battery pack.

[0176] After generating the charging and discharging optimization strategy and fault isolation plan, the plan is intuitively presented to the operation and maintenance personnel by combining the three-dimensional virtual model and the intelligent operation and maintenance platform, and specific implementation guidance is provided.

[0177] Embodiment 2:

[0178] See also Figure 4 , the present invention provides an embodiment: an intelligent operation and maintenance system for a battery pack in a thermal power plant based on machine learning, the system comprising a three-dimensional modeling module, a battery monitoring module, a data processing module and a maintenance suggestion module;

[0179] The three-dimensional modeling module is used to obtain the design information and current structure information of each battery in the battery pack and build a three-dimensional virtual model of the battery pack;

[0180] The battery monitoring module is used to equip a monitoring equipment group according to the three-dimensional virtual model of the battery group, and collect battery operation information according to the monitoring equipment group;

[0181] The data processing module is used to identify the fault status of battery cells and the imbalance phenomenon between batteries;

[0182] The maintenance suggestion module is used to detect abnormal conditions of single cells and battery packs according to the identification results of the two devices, issue early warning signals for potential faults in the three-dimensional virtual model of the battery pack, and generate an operation and maintenance plan.

[0183] The three-dimensional modeling module comprises:

[0184] An information acquisition unit, used to obtain battery design information and current structure information;

[0185] A feature extraction unit, used to process the battery image and extract the geometric features and connection features of the battery;

[0186] The modeling unit is used to reconstruct the battery features using multi-scale convolution and morphological processing to generate a three-dimensional model of the battery cell.

[0187] The data processing module comprises:

[0188] A fault identification unit is used to analyze the operating data of the battery cells and identify abnormal features through clustering algorithms and feature extraction techniques;

[0189] Balance analysis unit, used to calculate SOC balance and analyze the operation balance of the battery pack;

[0190] The abnormal classification unit is used to output the battery cell fault status and battery pack imbalance level according to the fault characteristics and balance analysis results.

[0191] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. Intelligent operation and maintenance method of battery packs in thermal power plants based on machine learning, characterized in that: The method comprises: Construct a three-dimensional virtual model of the battery pack; According to the three-dimensional virtual model of the battery pack, a monitoring equipment group is equipped, wherein the monitoring equipment group includes a battery cell monitoring equipment and a battery pack monitoring equipment; Collecting battery cell operation information according to the battery cell monitoring device, and identifying the battery cell fault state according to the cell operation information; Collecting battery pack operation information according to the battery pack monitoring device, and identifying the imbalance phenomenon between batteries according to the battery pack operation information; Detecting abnormal conditions of single cells and battery packs based on the recognition results of the two devices, and issuing early warning signals for potential faults in the three-dimensional virtual model of the battery pack; An operation and maintenance plan is generated through the operation and maintenance strategy, wherein the operation and maintenance plan includes an intra-group charge and discharge adjustment plan and a faulty single-cell isolation plan.

2. The intelligent operation and maintenance method of battery packs in thermal power plants based on machine learning according to claim 1 is characterized in that: The method of constructing a three-dimensional virtual model of a battery pack includes: Acquire design information and current structure information of each battery in the battery pack, wherein the design information includes technical drawings and specification information of the battery, and the current structure information includes a battery image; Obtaining the battery main structure in the technical drawing, and visualizing the battery pack through modeling software according to the battery main structure and the specification information to obtain a battery pack main model; Extracting features from the battery image by image processing technology, wherein the features include geometric features and connection features of the battery; The geometric features and connection features are reconstructed through multi-scale convolution downsampling, compensated according to the reconstruction error function, spliced ​​through morphological operations, converted into point cloud data, and the point cloud data is imported into point cloud processing software for processing, mapped on the battery main body model, and output as a three-dimensional virtual model of the battery pack.

3. The intelligent operation and maintenance method of battery packs in thermal power plants based on machine learning according to claim 1 is characterized in that: The identifying of the battery cell fault state includes: Acquire battery standard operation information, perform cluster analysis on the battery monomer operation information and the battery standard operation information, and output a clustering result, wherein the clustering result includes an operation model, a charge and discharge behavior, and a dynamic characteristic group; Calculating characteristic information of battery cell operation information according to the clustering results, wherein the characteristic information includes electrochemical characteristics, thermal characteristics and electrical energy characteristics; A fault identification network is constructed, the characteristic information is used as an input parameter of the fault identification network, the input parameter is trained through the fault identification network, and the fault status of the single cell is output.

4. The intelligent operation and maintenance method of a thermal power plant battery pack based on machine learning according to claim 3 is characterized in that: The characteristic information of calculating the battery cell operation information includes: Extract the voltage of the battery cell under different operating conditions according to the operating model, calculate the voltage deviation, calculate the internal resistance conversion rate by using Ohm's law through the dynamic current and voltage data in the charge and discharge behavior, analyze the charge and discharge curve offset of the battery cell during the charge and discharge process, evaluate the electrochemical efficiency of the battery, and thus calculate the electrochemical characteristics; According to the dynamic characteristic group, the surface temperature change curve of the battery cell during operation is extracted, the temperature rise rate is calculated, and the time constant and heat diffusion rate of the temperature rise process are calculated to evaluate the heat dissipation characteristics of the battery, thereby calculating the thermal characteristic characteristics; The charge and discharge energy ratios of battery cells in different operating models in the clustering results are extracted to evaluate the energy conversion efficiency. According to the dynamic characteristic group, the power output characteristics of the battery under short-term high-rate discharge are extracted. The capacity utilization and voltage platform attenuation trends at different discharge depths in the clustering results are analyzed to calculate the electric energy characteristics.

5. The intelligent operation and maintenance method of battery packs in thermal power plants based on machine learning according to claim 4 is characterized in that: The fault identification network comprises: An input layer, which establishes a feature sequence according to input parameters and calculates the fitting value of the feature sequence; A hidden layer converts the fitting value of the feature sequence of the input layer into a high-dimensional space, and calculates the output value of the hidden layer in the high-dimensional space; The output layer is used to output the fault status of the single battery of the fault identification network.

6. The intelligent operation and maintenance method of battery packs in thermal power plants based on machine learning according to claim 1 is characterized in that: The identifying of the imbalance phenomenon between batteries includes: Performing a residual energy test on the batteries in the battery pack, extracting a capacity loss characteristic parameter, and calculating the state of charge (SOC) of each battery according to the capacity loss characteristic parameter, and calculating the SOC balance according to the state of charge (SOC); Calculating the operation balance of the battery pack according to the battery pack operation information, wherein the operation balance includes voltage balance, power balance and thermal balance; The SOC balance and the operation balance are subjected to min-max normalization processing to generate characteristic parameters, an imbalance level is calculated according to the characteristic parameters, and an imbalance identification result of the battery pack is output.

7. The intelligent operation and maintenance method of battery packs in thermal power plants based on machine learning according to claim 6 is characterized in that: The calculating of the state of charge SOC of each battery includes: Under standard room temperature conditions, discharge the battery with a current of 1 / 4C. When the battery voltage is equal to the discharge cut-off voltage, stop discharging, record the data and leave it alone. Under standard room temperature conditions, charge the battery with a current of 1 / 4C. When the battery voltage is equal to the charging cut-off voltage, switch to constant voltage charging and detect the charging current. When the charging current is equal to the charging cut-off current, stop charging, record the data and leave it alone. Under standard room temperature conditions, discharge the fully charged battery at a current of 1 / 4C. When the battery voltage reaches the discharge cut-off voltage, stop discharging, record the data and leave it alone. Under standard room temperature conditions, charge the discharged battery at a current of 1 / 4C. When the battery voltage is equal to the charging cut-off voltage, stop charging, record the data and leave it alone. Perform curve fitting on the recorded data according to the pattern search nonlinear optimization algorithm, optimize the fitting results through the loss function, and output the capacity loss characteristic parameters; A SOC evaluation model is constructed, and the core parameters and penalty factors of the SOC evaluation model are optimized by the moth optimization algorithm. The test set is evaluated according to the optimized core parameters and penalty factors, and the state of charge SOC of the battery is output.

8. The intelligent operation and maintenance method of battery packs in thermal power plants based on machine learning according to claim 7 is characterized in that: The generating characteristic parameters include: According to the simulated annealing algorithm, the boundary parameters of the SOC balance and operation balance after the standardization are optimized to determine the characteristic parameter boundaries; According to the characteristic parameter boundary, SOC balance and operation balance, curve fitting is performed, the SOC balance and operation balance are converted into signals through wavelet transform, the signals are translated, the positions of discontinuities are changed, displayed in the curve, and characteristic parameters are extracted.

9. The intelligent operation and maintenance method of battery packs in thermal power plants based on machine learning according to claim 8 is characterized in that: The specific calculation formula of the imbalance level is: Among them, Φ ib represents the imbalance level, n represents the total number of characteristic parameters, f represents a single characteristic parameter, ξ f represents the weight of the fth feature parameter, F f represents the normalized value of the fth feature parameter, Θ f represents the dynamic sensitivity coefficient of the fth characteristic parameter, g f represents the dynamic deviation of the fth characteristic parameter, ε represents the smoothing correction coefficient, and Y s represents the total battery operating time, d represents a single battery, m represents the total number of batteries, ΔS d represents the running deviation value of the dth battery, ζ d represents the standard deviation of the dth battery during operation, Represents the ceiling function.

10. A thermal power plant battery pack intelligent operation and maintenance system based on machine learning, which is implemented based on the thermal power plant battery pack intelligent operation and maintenance method based on machine learning as claimed in any one of claims 1 to 9, characterized in that: The system includes a three-dimensional modeling module, a battery monitoring module, a data processing module and a maintenance suggestion module; The three-dimensional modeling module is used to obtain the design information and current structure information of each battery in the battery pack and build a three-dimensional virtual model of the battery pack; The battery monitoring module is used to equip a monitoring equipment group according to the three-dimensional virtual model of the battery group, and collect battery operation information according to the monitoring equipment group; The data processing module is used to identify the fault status of battery cells and the imbalance phenomenon between batteries; The maintenance suggestion module is used to detect abnormal conditions of single cells and battery packs according to the identification results of the two devices, issue early warning signals for potential faults in the three-dimensional virtual model of the battery pack, and generate an operation and maintenance plan.

Images