Energy storage system operation evaluation and fault early warning method

Abstract

The invention provides an energy storage system operation evaluation and fault early warning method, and relates to the technical field of energy storage systems, and the method comprises the steps: carrying out the historical data back calculation of a platform database, and carrying out the summarization to obtain an energy storage system unuploaded signal data set; determining a battery management system evaluation index set, an energy storage converter evaluation index set and a thermal management system evaluation index set; daily change condition evaluation and summarization are carried out on the non-uploaded signal data set according to the evaluation index set, and energy storage system operation daily state data are obtained; establishing a data threshold value, and performing fault data marking and early warning on the energy storage system operation day state data based on the data threshold value; and establishing an operation trend prediction model according to the energy storage system operation day state data, and carrying out risk data prediction and early warning. The technical problem that the prior art mainly depends on real-time data and an alarm mechanism to carry out operation state monitoring and is difficult to effectively prevent potential faults is solved, and the operation safety of the energy storage system is improved.

Description

Technical Field

[0001] The present application relates to the technical field of energy storage systems, and in particular to an energy storage system operation evaluation and fault warning method. Background Art

[0002] As an important part of the new energy power system, the energy storage system can realize efficient energy storage and flexible scheduling, provide support for the stability of the power grid, and is widely used in power peak regulation, emergency backup and distributed energy management. With the increasing scale and complexity of energy storage systems, the evaluation of their operating status and fault warning are becoming more and more important. At present, energy storage systems usually use sub-modules such as battery management system (BMS), energy storage inverter and thermal management system to collect real-time data for operating status monitoring, and realize fault warning through preset rule base or static threshold. These methods are mainly based on real-time signals and alarm information for status assessment, lack the ability to predict risks, and most of them can only detect problems after the fault occurs. Potential abnormalities cannot be blocked in advance, which increases safety hazards and equipment loss risks. Summary of the invention

[0003] The present application provides a method for energy storage system operation evaluation and fault warning, which solves the technical problem that the existing technology mainly relies on real-time data and alarm mechanisms to monitor the operation status, resulting in abnormal problems usually only being passively handled after they occur or in the current state, and it is difficult to effectively prevent potential faults. It achieves the technical effect of actively identifying potential problems and thus improving the operational safety of the energy storage system.

[0004] In view of the above problems, the present application provides a method for energy storage system operation evaluation and fault warning, the method comprising: performing historical data back calculation on the platform database, and summarizing a data set of energy storage system signals that have not been uploaded, wherein the sources of the data set of energy storage system signals that have not been uploaded include a battery management system, an energy storage inverter, and a thermal management system; determining a battery management system evaluation index set, an energy storage inverter evaluation index set, and a thermal management system evaluation index set according to the battery management system, the energy storage inverter, and the thermal management system; performing daily change evaluation and summary on the data set of energy storage system signals that have not been uploaded according to the battery management system evaluation index set, the energy storage inverter evaluation index set, and the thermal management system evaluation index set, to obtain daily operation status data of the energy storage system; establishing a data threshold according to the daily operation status data of the energy storage system, and performing fault data marking warning on the daily operation status data of the energy storage system based on the data threshold; establishing an operation trend prediction model according to the daily operation status data of the energy storage system, and performing risk data prediction and warning based on the operation trend prediction model.

[0005] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0006] The platform database is back-calculated for historical data, and the energy storage system's unuploaded signal data set is summarized, and the sources of the energy storage system's unuploaded signal data set include the battery management system, the energy storage inverter, and the thermal management system. Through the back-calculation of historical data, data that is ignored in conventional evaluations can be mined, providing a comprehensive data basis for subsequent evaluations and early warnings. According to the battery management system, the energy storage inverter, and the thermal management system, the battery management system evaluation index set, the energy storage inverter evaluation index set, and the thermal management system evaluation index set are determined. Compared with the traditional overall evaluation index, this more targeted evaluation index set can more accurately reflect the operating status of each subsystem and improve the accuracy of the analysis of the operating status of each subsystem. According to the battery management system evaluation index set, the energy storage inverter evaluation index set, and the thermal management system evaluation index set, the daily changes of the energy storage system's unuploaded signal data set are evaluated and summarized to obtain the energy storage system's daily operating status data. Through the comprehensive evaluation of the daily change data, the operating status of the energy storage system is dynamically reflected, which is convenient for capturing the anomalies and trends that may exist in the short term, and providing data support for fault marking and trend prediction. According to the daily operation status data of the energy storage system, a data threshold is established, and based on the data threshold, a fault data marking warning is performed on the daily operation status data of the energy storage system, so as to realize real-time warning of potential problems of the energy storage system and solve the problems of false alarms and omissions in the static rule base. According to the daily operation status data of the energy storage system, an operation trend prediction model is established, and risk data prediction and warning are performed based on the operation trend prediction model. The operation trend prediction model can be used to analyze future trends, identify potential risks in advance, and realize the transformation from passive response to active prevention.

[0007] In summary, this application achieves a comprehensive assessment and accurate early warning of the operating status of the energy storage system through the back-calculation and completion of historical data and the modular design of multi-subsystem evaluation indicators, combined with daily change evaluation, dynamic data thresholds and operation trend prediction models. It not only improves the coverage and accuracy of fault detection, but also significantly enhances the ability to predict potential risks in advance, thereby greatly improving the operating safety and stability of the energy storage system.

[0008] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 A flowchart of an energy storage system operation evaluation and fault warning method provided in an embodiment of the present application.

[0010] Figure 2A schematic diagram of a flow chart for determining a battery management system evaluation index set, an energy storage converter evaluation index set, and a thermal management system evaluation index set in a method for evaluating the operation of an energy storage system and providing a fault warning in an embodiment of the present application.

[0011] Figure 3 A schematic diagram of a process for obtaining daily operating status data of an energy storage system in a method for energy storage system operation evaluation and fault warning provided in an embodiment of the present application. DETAILED DESCRIPTION

[0012] The embodiment of the present application provides a method for energy storage system operation evaluation and fault warning, which solves the technical problem that the prior art mainly relies on real-time data and alarm mechanisms to monitor the operation status, resulting in abnormal problems that can usually only be passively handled after they occur or in the current state, making it difficult to effectively prevent potential faults. The embodiment of the present application achieves the technical effect of actively identifying potential problems and thus improving the operational safety of the energy storage system.

[0013] like Figure 1 As shown, an embodiment of the present application provides an energy storage system operation evaluation and fault warning method, the method comprising:

[0014] Step S1: perform historical data back calculation on the platform database to obtain a data set of signals not uploaded by the energy storage system, where the data set of signals not uploaded by the energy storage system comes from a battery management system, an energy storage converter and a thermal management system.

[0015] Specifically, the platform database is a database that stores the historical operation data of the energy storage system, including real-time collected data, historical records and related calculation results. The unuploaded signal data set is a collection of signals that failed to be transmitted to the database due to communication delays, failures or other reasons during the operation of the energy storage system. These unuploaded signals include operating data generated or collected by subsystems such as the battery management system, energy storage inverter and thermal management system, such as battery voltage, temperature, inverter status, cooling system operating parameters, etc.

[0016] Historical data related to the energy storage system are extracted from the platform database, and the historical data are back-calculated to summarize the important signals in the energy storage system that have not been calculated and uploaded by each subsystem, such as the battery efficiency of the SOC at a specific time point of the energy storage system, the total DC internal resistance of the total DC side of the SOC system at a specific time point of the energy storage system, etc. that have not been uploaded by the battery management system; the target minimum outlet water temperature, the highest and lowest average ambient temperature, the lowest liquid supply pressure, etc. that have not been uploaded by the thermal management system; the grid voltage and standard voltage deviation value, the actual power and target power deviation value, the upper and lower half mother school voltage deviation value, etc. that have not been uploaded by the energy storage converter.

[0017] These unuploaded data may hide key information about the operation of the energy storage system. Collecting this data can provide a richer data source for subsequent comprehensive evaluation of the operating status of the energy storage system, avoiding incomplete evaluation due to missing data.

[0018] Step S2: According to the battery management system, the energy storage converter and the thermal management system, a battery management system evaluation index set, an energy storage converter evaluation index set and a thermal management system evaluation index set are determined.

[0019] Specifically, different subsystems have different functions and operating characteristics in the energy storage system. According to the functional characteristics of each subsystem, the corresponding evaluation index sets are determined for the battery management system, energy storage converter and thermal management system, so as to measure the operating status of each subsystem in a targeted manner and accurately locate or predict the specific faults of each subsystem. Specifically, it includes: establishing a battery management system evaluation index set for the battery management system, which is specifically used to measure the operating status of the battery management system. For example, energy efficiency, battery voltage, current, internal resistance, battery temperature, etc. Establishing an energy storage converter evaluation index set for the energy storage converter to evaluate the performance and operating status of the energy storage converter. For example, the maximum grid voltage deviation and maximum power deviation of the energy storage converter. Establishing a thermal management system evaluation index set for the thermal management system to evaluate the operating status of the thermal management system. For example, target water temperature, average ambient temperature, liquid supply pressure, etc.

[0020] Step S3: Evaluate and summarize the daily changes of the energy storage system's non-uploaded signal data set according to the battery management system evaluation index set, the energy storage converter evaluation index set, and the thermal management system evaluation index set to obtain the energy storage system's daily operating status data.

[0021] Specifically, the energy storage system's unuploaded signal data set is analyzed and synthesized according to the determined battery management system evaluation index set, energy storage inverter evaluation index set, and thermal management system evaluation index set according to a one-day time period, and the data is analyzed from the perspective of daily changes to obtain the energy storage system's daily operating status data. This daily evaluation summary can reflect the energy storage system's operating trends and status changes in the short term, providing a detailed data basis for subsequent early warnings and forecasts.

[0022] Step S4: establishing a data threshold according to the daily operation status data of the energy storage system, and performing a fault data marking warning on the daily operation status data of the energy storage system based on the data threshold.

[0023] Specifically, data thresholds are established based on the characteristics of the daily operating status data of the energy storage system and the normal operating range of each subsystem. The data thresholds are used to determine whether the parameters in the daily operating status data of the energy storage system are in a normal state. The data thresholds can be determined by referring to the technical specifications provided by the equipment manufacturer, industry practice standards, product performance parameters, etc. The data thresholds are used to check the daily operating status data of the energy storage system, find out the data that exceeds the data threshold, mark it as fault data and issue an early warning notification, such as displaying alarm information, sound and light signal warnings, etc. in the monitoring system. By marking and warning the fault data, problems that have occurred in the energy storage system can be discovered in a timely manner so that appropriate measures can be taken for repair and maintenance.

[0024] Step S5: establishing an operation trend prediction model according to the daily operation status data of the energy storage system, and performing risk data prediction and early warning based on the operation trend prediction model.

[0025] Specifically, the operation trend prediction model is a mathematical model built based on historical data (i.e., the daily operation status data of the energy storage system), which is used to predict the future operation status and possible risks of the energy storage system. For example, it can be an ARIMA model or a neural network model based on time series analysis. The machine learning algorithm is used to learn the trend of the daily operation status data of the energy storage system, and the operation trend prediction model is constructed. The current daily operation status data of the energy storage system is input into the model. The model will predict the future operation trend based on the input current and historical data. When the prediction result shows that a parameter is about to exceed the normal range or an abnormal trend occurs, a risk warning is issued. The warning method is similar to that in step S4.

[0026] By establishing an operation trend prediction model, starting from short-term operation status data, we can predict possible risks in the future and conduct forward-looking warnings. This allows us to take preventive measures before a failure occurs, improve the safety and stability of the energy storage system operation, and reduce downtime caused by failures.

[0027] Further, such as Figure 2 As shown, step S2 of the embodiment of the present application also includes:

[0028] Step S21: Establishing operation data analysis logic based on energy storage system testing theory.

[0029] Step S22: extracting important indicators of the battery management system, energy storage inverter and thermal management system in turn according to the operation data analysis logic to obtain an initial battery management system indicator set, an initial energy storage inverter indicator set and an initial thermal management system indicator set.

[0030] Step S23: performing dimensionality reduction analysis on the initial battery management system indicator set, the initial energy storage converter indicator set and the initial thermal management system indicator set respectively to determine the battery management system evaluation indicator set, the energy storage converter evaluation indicator set and the thermal management system evaluation indicator set.

[0031] Specifically, the energy storage system test theory covers the principles and methodologies for testing the performance and functions of various components of the energy storage system (such as the battery management system, energy storage inverter, and thermal management system) under different operating conditions, including the test objectives, test condition settings, and test data interpretation. These theories can be derived from relevant industry standard documents and test manuals provided by energy storage equipment manufacturers. Based on the energy storage system test theory, determine the ideas and rules for analyzing the operating data of the energy storage system, and construct the operating data analysis logic, including the type of data to be analyzed, data judgment criteria, and calculation basis for the analysis process.

[0032] According to the operation data analysis logic, important indicators with key significance are selected from the operation data related indicators of each subsystem (battery management system, energy storage inverter and thermal management system), and these indicators are summarized to obtain the initial battery management system indicator set, the initial energy storage inverter indicator set and the initial thermal management system indicator set. Exemplarily, the operation data analysis logic stipulates that the battery power related indicators should be paid attention to first, and then the energy efficiency, battery voltage, current, internal resistance, battery temperature and battery surface humidity and other indicators are extracted from the data of the battery management system to establish the initial battery management system indicator set. For the energy storage inverter, the operation data analysis logic focuses on the indicators related to the power conversion efficiency, and extracts the grid voltage and standard voltage deviation values, the actual power and target power deviation values ​​and other indicators to establish the initial energy storage inverter indicator set. These indicators reflect the key characteristics of the energy storage inverter in operation. For the thermal management system, the operation data analysis logic focuses on the temperature control related indicators, and extracts the minimum value of the target outlet water temperature, the highest and lowest average values ​​of the ambient temperature, the lowest value of the liquid supply pressure and other indicators to establish the initial thermal management system indicator set.

[0033] The principal component analysis is used to perform dimensionality reduction analysis on the above three initial indicator sets. Under the premise of retaining the main features of the data, the initial indicator sets of each system are simplified, and those indicators that have little impact on the evaluation of the operating status or have redundant information are removed, so as to reduce the dimension of the data (i.e., the number of indicators) and make the final evaluation indicator set more refined and effective. For example, the initial battery management system indicator set includes multiple indicators such as battery voltage, current, internal resistance, battery temperature, and battery surface humidity. Through principal component analysis, it is found that the battery surface humidity indicator has a low correlation with other indicators and contributes little to the overall evaluation of the operating status of the battery management system. The battery surface humidity indicator is removed from the indicator set, and the battery management system evaluation indicator set is finally determined. A similar method is also used for the initial energy storage inverter indicator set and the initial thermal management system indicator set. Through dimensionality reduction analysis, the initial indicator set is simplified, which not only ensures the reliability of the evaluation results, but also reduces the computational complexity.

[0034] Furthermore, step S23 further includes:

[0035] Step S231: respectively evaluate the importance of the initial battery management system indicator set, the initial energy storage converter indicator set and the initial thermal management system indicator set to obtain a battery management system indicator importance coefficient set, an energy storage converter indicator importance coefficient set and a thermal management system indicator importance coefficient set.

[0036] Step S232: Based on the battery management system indicator importance coefficient set, the energy storage inverter indicator importance coefficient set and the thermal management system indicator importance coefficient set, respectively perform dimensionality reduction analysis on the initial battery management system indicator set, the initial energy storage inverter indicator set and the initial thermal management system indicator set to determine the battery management system evaluation indicator set, the energy storage inverter evaluation indicator set and the thermal management system evaluation indicator set.

[0037] Specifically, the battery management system indicator importance coefficient set is a set of importance coefficients obtained by quantifying the importance of each indicator in the initial battery management system indicator set. These coefficients in this set represent the relative importance of each indicator in evaluating the operating status of the battery management system. The larger the importance coefficient, the more important the indicator. Similarly, the energy storage inverter indicator importance coefficient set is a set of importance coefficients obtained by quantifying the importance of each indicator in the initial energy storage inverter indicator set, reflecting the importance of each indicator in the evaluation of the operating status of the energy storage inverter. The thermal management system indicator importance coefficient set is a set of importance coefficients obtained by quantifying the importance of each indicator in the initial thermal management system indicator set, which is used to measure the importance of each indicator in the evaluation of the thermal management system operation.

[0038] The initial battery management system indicator set, the initial energy storage converter indicator set and the initial thermal management system indicator set are analyzed using statistical methods to evaluate the degree of influence of each indicator in each indicator set on the performance or failure of the energy storage system, and a weight value, i.e., the importance coefficient, is assigned to each indicator, thereby obtaining the battery management system indicator importance coefficient set, the energy storage converter indicator importance coefficient set and the thermal management system indicator importance coefficient set. Taking the initial battery management system indicator set as an example, the hierarchical analysis method can be used for importance assessment. The overall operating state of the battery management system is taken as the target layer, and each indicator in the initial indicator set is taken as the criterion layer. The relative importance between the indicators is compared by constructing a judgment matrix. Exemplarily, the initial battery management system indicator set includes three indicators: energy efficiency, daily discharge capacity and total DC resistance on the DC side. By analyzing the degree of correlation between historical fault data and these indicators, the importance ratio between the indicators is determined, and the judgment matrix is ​​constructed. Then, the importance coefficient of each indicator is obtained by calculating the eigenvector of the judgment matrix to form a battery management system indicator importance coefficient set.

[0039] The initial indicator set is processed according to the importance coefficient of each indicator in each indicator importance coefficient set, and those indicators with lower importance are removed, thereby reducing the dimension of the indicator set and obtaining a more concise and effective evaluation indicator set. Taking the initial battery management system indicator set as an example, an importance coefficient threshold (such as 0.1) is set. When the importance coefficient of the indicator is lower than this threshold, the indicator is removed from the initial battery management system indicator set, and the remaining indicators constitute the battery management system evaluation indicator set. For the initial energy storage inverter indicator set and the initial thermal management system indicator set, similar processing is performed based on their respective indicator importance coefficient sets and the set thresholds to determine the energy storage inverter evaluation indicator set and the thermal management system evaluation indicator set.

[0040] The initial indicator set is reduced in dimension according to the relative importance of each indicator, and the final evaluation indicator set is more streamlined, which can effectively reflect the core operating status of the energy storage system, improve the accuracy of fault warning and performance evaluation, and reduce the consumption of computing resources.

[0041] Further, such as Figure 3 As shown, step S3 of the embodiment of the present application also includes:

[0042] Step S31: respectively analyze the value states of the battery management system evaluation index set, the energy storage converter evaluation index set and the thermal management system evaluation index set to obtain the battery management system index value state set, the energy storage converter index value state set and the thermal management system index value state set.

[0043] Step S32: parsing the calculation formulas of the battery management system evaluation index set, the energy storage converter evaluation index set and the thermal management system evaluation index set to obtain the battery management system index calculation formula set, the energy storage converter index calculation formula set and the thermal management system index calculation formula set.

[0044] Step S33: Based on the battery management system indicator value state set, the energy storage inverter indicator value state set and the thermal management system indicator value state set and the battery management system indicator calculation formula set, the energy storage inverter indicator calculation formula set and the thermal management system indicator calculation formula set, the daily changes of the energy storage system's non-uploaded signal data set are evaluated and summarized to obtain the energy storage system's daily operating status data.

[0045] Specifically, for each indicator in the battery management system evaluation indicator set, energy storage converter evaluation indicator set and thermal management system evaluation indicator set, analyze the time of data collection, value range, value type (such as discrete value or continuous value) and value restriction conditions of each indicator. Sort out the value status of each indicator to form the battery management system indicator value status set, energy storage converter indicator value status set and thermal management system indicator value status set.

[0046] At the same time, according to the equipment manual and relevant technical data and standards, the calculation method, variables involved and calculation logic of each indicator in each evaluation indicator set are determined, and the battery management system indicator calculation formula set, energy storage inverter indicator calculation formula set and thermal management system indicator calculation formula set are generated.

[0047] By comprehensively utilizing the indicator value state set and calculation formula set obtained above, the energy storage system's unuploaded signal data set is analyzed and calculated on a daily basis, and the calculation results are summarized to obtain the energy storage system's operating status data within a day.

[0048] For example, for the energy efficiency in the battery management system evaluation index set, data is collected when the battery is in an empty state, and the energy efficiency of the battery management system for the day is calculated by the calculation formula: energy efficiency = cumulative discharge amount / cumulative charge amount; for the daily discharge amount in the battery management system evaluation index set, data is collected at zero o'clock every day, and the daily discharge amount of the day is calculated by the calculation formula: daily discharge amount = cumulative discharge amount collected today - cumulative discharge amount collected last day; for the maximum temperature and average temperature of the battery in the battery management system evaluation index set, the temperature data of the whole day is collected, the maximum value is screened and the average value is calculated to determine the maximum temperature and average temperature of the battery on the day; for the total DC resistance of the DC side in the battery management system evaluation index set, data is collected at the moment of full charge and empty point state, and the total DC resistance of the DC side on the day is calculated by the calculation formula: total DC resistance of the DC side = total voltage / total current. For the maximum grid voltage deviation in the energy storage converter evaluation index set, the voltage data of the whole day is collected, and the grid voltage deviation of the whole day is calculated and the maximum value is screened by the calculation formula grid voltage deviation = |real-time output grid voltage value-standard voltage value|, and the maximum grid voltage deviation of the day is obtained. For the pressure difference in the thermal management system evaluation index concentration, the supply and return liquid pressures throughout the day are collected. The pressure difference of the whole day is calculated through the calculation formula pressure difference = supply liquid pressure - return liquid pressure, and the maximum value is selected to obtain the pressure difference of the day.

[0049] Furthermore, step S4 of the embodiment of the present application also includes:

[0050] Step S41: marking the daily operating status data of the energy storage system as fault data based on the data threshold to obtain a fault data set of the energy storage system.

[0051] Step S42: performing fault warning analysis on the energy storage system fault data set, determining a first energy storage system warning parameter, and performing energy storage system fault warning and operation and maintenance push based on the first energy storage system warning parameter.

[0052] Specifically, each data point in the daily operation status data of the energy storage system is traversed. For each data point, a judgment is made according to the previously established data threshold, and the data point exceeding the data threshold is marked as fault data, thereby obtaining an energy storage system fault data set, which contains all data points marked as fault data in the daily operation status data of the energy storage system.

[0053] Conduct an in-depth analysis of the energy storage system fault data set, and use data analysis and statistical methods, such as calculating the frequency, mean, variance and other statistics of the fault data to determine the severity of the fault, possible impact range and other information. Determine the first energy storage system warning parameter based on the analysis results. This parameter is used to describe the fault warning situation. It can be the severity level of the fault (such as mild, moderate, severe) or the frequency of the fault. For example, analyze the distribution of battery fault data in the fault data set. If the fault data of a certain battery appears multiple times in one day, the severity level of the fault may be determined as a higher level (severe). Send the fault warning information containing the first energy storage system warning parameters to the relevant operation and maintenance personnel through the warning information push channel, such as the message push platform, the internal operation and maintenance management system, etc., to help the operation and maintenance personnel quickly take countermeasures to ensure the stable operation of the energy storage system.

[0054] Furthermore, step S5 of the embodiment of the present application further includes:

[0055] Step S51: dividing the daily operating status data of the energy storage system to determine a data training set, a data verification set and a data test set.

[0056] Step S52: Analyze the daily operating status data of the energy storage system to obtain operating data characteristic information, and select a target training model based on the operating data characteristic information.

[0057] Step S53: Use the target training model to perform trend prediction training on the data training set, the data verification set and the data test set to establish the operation trend prediction model.

[0058] Specifically, the random division method is used to divide the daily status data of the energy storage system into different subsets according to a certain proportion or rule. For example, 70% is the data training set, 15% is the data verification set, and 15% is the data test set.

[0059] Perform characteristic analysis on the daily status data of the energy storage system, and determine the target training model suitable for predicting the operating trend of the energy storage system by analyzing the distribution law of the data and the correlation between the data. The target training model can be a linear regression model, a neural network model, etc. For example, by drawing a histogram of the daily status data of the energy storage system, observe the distribution of the data, and use the correlation coefficient calculation to determine the correlation between the data. If the data shows a linear relationship and the distribution is relatively regular, the linear regression model can be selected as the target training model; if the data relationship is complex and nonlinear, the neural network model can be selected as the target training model.

[0060] The target training model is trained using the data training set, validation set and test set so that it can predict the operating trend of the energy storage system. By repeatedly adjusting the model parameters, the model gradually converges to obtain an operating trend prediction model.

[0061] By training the operation trend prediction model, the future state or trend of the energy storage system can be predicted, potential faults can be identified in advance to optimize the operation strategy, achieve more intelligent and accurate state monitoring and early warning, and improve the reliability and operation and maintenance efficiency of the energy storage system.

[0062] Furthermore, step S53 further includes:

[0063] Step S531: Use the target training model to perform trend prediction training on the data training set to establish a basic trend prediction model.

[0064] Step S532: Use the data verification set and the data test set to perform verification testing on the basic trend prediction model to obtain prediction model performance parameters.

[0065] Step S533: Optimizing the basic trend prediction model based on the prediction model performance parameters to establish the operation trend prediction model.

[0066] Specifically, the prediction model performance parameters are indicators used to measure the performance of the basic trend prediction model on the data validation set and the data test set, such as mean square error (MSE), mean absolute error (MAE), determination coefficient (R 2 ) etc. These parameters can reflect the degree of deviation between the model prediction results and the actual results or the degree of fit of the model to the data.

[0067] During the model training process, the data training set is first input into the target training model, which is an initialized model. By learning the characteristics of the daily operating status data of the energy storage system in the data training set, the model parameters are gradually changed to reduce the error between the prediction results and the actual data, forming a basic trend prediction model. The data validation set and the data test set are respectively substituted into the basic trend prediction model to obtain the prediction output, and then the prediction model performance parameters are calculated to check the performance of the model. The basic trend prediction model is adjusted according to the prediction model performance parameters to improve the performance of the model and obtain the operation trend prediction model.

[0068] Taking the linear regression model as an example, the data training set is input into the initialized linear regression model, and the value of the regression coefficient is calculated through optimization algorithms such as the least squares method to establish a basic trend prediction model. The data validation set and the data test set are respectively substituted into the basic trend prediction model to obtain the prediction output, and then the prediction model performance parameters are calculated. If the performance parameters are not ideal, you can try to add more features or transform the existing features (such as logarithmic transformation, power transformation, etc.). During the adjustment process, the data validation set and the data test set are used again to evaluate the performance of the adjusted model, and this process is repeated until satisfactory prediction model performance parameters are obtained, thereby establishing a running trend prediction model.

[0069] Through the above training process, the obtained operation trend prediction model can more accurately predict the operation trend of the energy storage system, thereby helping to identify potential faults in advance.

[0070] Furthermore, step S5 of the embodiment of the present application further includes:

[0071] Step S54: predicting the daily operating status data of the energy storage system based on the operating trend prediction model to obtain operating trend parameters of the energy storage system.

[0072] Step S55: obtaining a data model threshold, marking the energy storage system operation trend parameter as a fault parameter based on the data model threshold, and obtaining the energy storage system fault trend parameter.

[0073] Step S56: performing fault warning analysis on the fault trend parameters of the energy storage system, determining a second energy storage system warning parameter, and performing energy storage system fault warning and operation and maintenance push based on the second energy storage system warning parameter.

[0074] Specifically, the data model threshold is a critical value used to determine whether there is an abnormality in the operating trend parameter of the energy storage system, which can be a data discrete value or a standard curve, etc. The data model threshold can be determined based on equipment performance, operating standard specifications, or normal operating modes in historical data.

[0075] The daily operation status data of the energy storage system is input into the established operation trend prediction model. The model outputs the operation trend parameters of the energy storage system. These parameters reflect the possible future operation status of the energy storage system, such as the battery charging and discharging status, temperature change trend, power output change, etc. For each energy storage system operation trend parameter, it is compared with the corresponding data model threshold. When the trend parameter exceeds the corresponding data model threshold range, the trend parameter is marked as a fault trend parameter. All marked fault trend parameters are summarized to generate the energy storage system fault trend parameter.

[0076] Using similar data analysis and statistical methods as in step S42, a fault warning analysis is performed on the fault trend parameters of the energy storage system, the change rate of the fault trend parameters, the predicted time of the fault occurrence, etc. are calculated to determine the severity level of the fault, and the second energy storage system warning parameters are generated. Through the warning information push channel, the fault warning information including the second energy storage system warning parameters, etc. is sent to the relevant operation and maintenance personnel, realizing the early fault warning of the energy storage system, and effectively improving the operation stability and reliability of the energy storage system.

[0077] In summary, the energy storage system operation evaluation and fault warning method provided in the embodiment of the present application has the following technical effects:

[0078] The platform database is back-calculated for historical data, and the energy storage system's unuploaded signal data set is summarized, and the sources of the energy storage system's unuploaded signal data set include the battery management system, the energy storage inverter, and the thermal management system. Through the back-calculation of historical data, data that is ignored in conventional evaluations can be mined, providing a comprehensive data basis for subsequent evaluations and early warnings. According to the battery management system, the energy storage inverter, and the thermal management system, the battery management system evaluation index set, the energy storage inverter evaluation index set, and the thermal management system evaluation index set are determined. Compared with the traditional overall evaluation index, this more targeted evaluation index set can more accurately reflect the operating status of each subsystem and improve the accuracy of the analysis of the operating status of each subsystem. According to the battery management system evaluation index set, the energy storage inverter evaluation index set, and the thermal management system evaluation index set, the daily changes of the energy storage system's unuploaded signal data set are evaluated and summarized to obtain the energy storage system's daily operating status data. Through the comprehensive evaluation of the daily change data, the operating status of the energy storage system is dynamically reflected, which is convenient for capturing the anomalies and trends that may exist in the short term, and providing data support for fault marking and trend prediction. According to the daily operation status data of the energy storage system, a data threshold is established, and based on the data threshold, a fault data marking warning is performed on the daily operation status data of the energy storage system, so as to realize real-time warning of potential problems of the energy storage system and solve the problems of false alarms and omissions in the static rule base. According to the daily operation status data of the energy storage system, an operation trend prediction model is established, and risk data prediction and warning are performed based on the operation trend prediction model. The operation trend prediction model can be used to analyze future trends, identify potential risks in advance, and realize the transformation from passive response to active prevention.

[0079] In general, the embodiments of the present application achieve a comprehensive assessment and accurate early warning of the operating status of the energy storage system through the back-calculation and completion of historical data and the modular design of multi-subsystem evaluation indicators, combined with daily change evaluation, dynamic data thresholds and operating trend prediction models. It not only improves the coverage and accuracy of fault detection, but also significantly enhances the ability to predict potential risks in advance, thereby greatly improving the operating safety and stability of the energy storage system.

[0080] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for energy storage system operation evaluation and fault warning, characterized in that: The method comprises: Perform historical data back calculation on the platform database to obtain a data set of signals not uploaded by the energy storage system, where the data set of signals not uploaded by the energy storage system comes from a battery management system, an energy storage converter, and a thermal management system; Determine a battery management system evaluation index set, an energy storage converter evaluation index set, and a thermal management system evaluation index set according to the battery management system, the energy storage converter, and the thermal management system; According to the battery management system evaluation index set, the energy storage converter evaluation index set and the thermal management system evaluation index set, the daily change of the energy storage system non-uploaded signal data set is evaluated and summarized to obtain the daily operation status data of the energy storage system; According to the daily operation status data of the energy storage system, a data threshold is established, and based on the data threshold, a fault data marking warning is performed on the daily operation status data of the energy storage system; An operation trend prediction model is established according to the daily operation status data of the energy storage system, and risk data prediction and early warning are performed based on the operation trend prediction model.

2. The energy storage system operation evaluation and fault warning method according to claim 1, characterized in that: The step of determining a battery management system evaluation index set, an energy storage converter evaluation index set, and a thermal management system evaluation index set includes: Based on the energy storage system testing theory, establish the operation data analysis logic; According to the operation data analysis logic, important indicators of the battery management system, energy storage converter and thermal management system are extracted in sequence to obtain an initial battery management system indicator set, an initial energy storage converter indicator set and an initial thermal management system indicator set; Dimensionality reduction analysis is performed on the initial battery management system indicator set, the initial energy storage inverter indicator set, and the initial thermal management system indicator set, respectively, to determine the battery management system evaluation indicator set, the energy storage inverter evaluation indicator set, and the thermal management system evaluation indicator set.

3. The energy storage system operation evaluation and fault warning method according to claim 2, characterized in that: The determining of the battery management system evaluation index set, the energy storage converter evaluation index set and the thermal management system evaluation index set includes: Performing importance evaluation on the initial battery management system indicator set, the initial energy storage converter indicator set, and the initial thermal management system indicator set, respectively, to obtain a battery management system indicator importance coefficient set, an energy storage converter indicator importance coefficient set, and a thermal management system indicator importance coefficient set; Based on the battery management system indicator importance coefficient set, energy storage inverter indicator importance coefficient set and thermal management system indicator importance coefficient set, the initial battery management system indicator set, initial energy storage inverter indicator set and initial thermal management system indicator set are respectively subjected to dimensionality reduction analysis to determine the battery management system evaluation indicator set, energy storage inverter evaluation indicator set and thermal management system evaluation indicator set.

4. The energy storage system operation evaluation and fault warning method according to claim 1, characterized in that: The obtaining of daily operation status data of the energy storage system includes: Respectively perform value state analysis on the battery management system evaluation index set, the energy storage converter evaluation index set, and the thermal management system evaluation index set to obtain a battery management system index value state set, an energy storage converter index value state set, and a thermal management system index value state set; Parsing the calculation formulas of the battery management system evaluation index set, the energy storage converter evaluation index set, and the thermal management system evaluation index set to obtain the battery management system index calculation formula set, the energy storage converter index calculation formula set, and the thermal management system index calculation formula set; Based on the battery management system indicator value state set, the energy storage inverter indicator value state set and the thermal management system indicator value state set and the battery management system indicator calculation formula set, the energy storage inverter indicator calculation formula set and the thermal management system indicator calculation formula set, the daily changes in the signal data set not uploaded by the energy storage system are evaluated and summarized to obtain the daily operating status data of the energy storage system.

5. The energy storage system operation evaluation and fault warning method according to claim 1, characterized in that: The performing fault data marking warning on the daily operation status data of the energy storage system based on the data threshold comprises: Performing fault data marking on the daily operating status data of the energy storage system based on the data threshold to obtain an energy storage system fault data set; A fault warning analysis is performed on the energy storage system fault data set to determine a first energy storage system warning parameter, and energy storage system fault warning and operation and maintenance push are performed based on the first energy storage system warning parameter.

6. The energy storage system operation evaluation and fault warning method according to claim 1, characterized in that: The step of establishing the operation trend prediction model comprises: Dividing the daily operating status data of the energy storage system to determine a data training set, a data verification set, and a data test set; Performing characteristic analysis on the daily operating status data of the energy storage system to obtain characteristic information of the operating data, and selecting a target training model according to the characteristic information of the operating data; The target training model is used to perform trend prediction training on the data training set, data verification set and data test set to establish the operation trend prediction model.

7. The energy storage system operation evaluation and fault warning method according to claim 6, characterized in that: The step of establishing the operation trend prediction model comprises: Perform trend prediction training on the data training set using the target training model to establish a basic trend prediction model; Using the data validation set and the data test set to perform validation testing on the basic trend prediction model to obtain prediction model performance parameters; The basic trend prediction model is optimized based on the prediction model performance parameters to establish the operation trend prediction model.

8. The energy storage system operation evaluation and fault warning method according to claim 1, characterized in that: The risk data prediction and early warning based on the operation trend prediction model includes: Predicting the daily operating status data of the energy storage system based on the operating trend prediction model to obtain operating trend parameters of the energy storage system; Acquire a data model threshold, and mark the energy storage system operation trend parameter as a fault parameter based on the data model threshold to obtain the energy storage system fault trend parameter; Perform a fault warning analysis on the fault trend parameters of the energy storage system, determine a second energy storage system warning parameter, and perform energy storage system fault warning and operation and maintenance push based on the second energy storage system warning parameter.

Images