High-capacity energy storage battery failure early warning system based on pattern recognition
By laying sensor arrays in energy storage batteries for real-time data acquisition and pattern recognition, the real-time and accuracy of the energy storage battery status monitoring and early warning system is solved, timely identification and early warning of abnormal situations is realized, and energy storage is improved System stability and reliability.
Patent Information
- Application Number
- CN202410866036.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-01
- Publication Date
- 2024-11-22
AI Technical Summary
The status monitoring and early warning systems of large-capacity energy storage batteries lack real-time and accuracy, resulting in poor timeliness of failure warning.
A large-capacity energy storage battery failure warning system is adopted based on pattern recognition. The battery operating parameters are collected in real time through the sensor array, stability abnormality analysis and pattern recognition are performed, and clear warning levels are generated in a timely manner.
Real-time monitoring and accurate abnormal identification of battery operating parameters is realized, troubleshooting efficiency and timely warning are improved, and the stable operation of the energy storage system is ensured.
Smart Images

Figure CN119001453A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of energy storage batteries, and in particular to a failure early warning system for large-capacity energy storage batteries based on pattern recognition. Background Art
[0002] The instability of renewable energy sources such as wind and solar energy requires energy systems to have energy storage functions. Large-capacity energy storage batteries are often used in renewable energy storage systems. Through failure warning methods, abnormal conditions of batteries can be discovered in time to ensure the stable operation of the energy storage system and improve the utilization efficiency of renewable energy. However, in past technologies, the status monitoring and warning of large-capacity energy storage batteries usually rely on manual inspections or regular inspections, which lack real-time and accuracy, resulting in technical problems such as poor timeliness of failure warnings. Summary of the invention
[0003] This application provides a large-capacity energy storage battery failure warning system based on pattern recognition, aiming to solve the technical problem that the traditional large-capacity energy storage battery status monitoring and warning system lacks real-time and accuracy, resulting in poor timeliness of failure warning.
[0004] In view of the above problems, the present application provides a large-capacity energy storage battery failure warning system based on pattern recognition.
[0005] The present application discloses a large-capacity energy storage battery failure warning system based on pattern recognition, the system comprising: a sensor array layout module, used to layout a sensor array in a target energy storage battery, wherein the sensor array comprises a plurality of sensors; an operating parameter acquisition module, used to acquire battery operating parameters through the plurality of sensors within a preset time zone, and obtain a plurality of sensor data sequences, wherein the plurality of sensor data sequences are provided with time identifiers; an abnormality analysis module, used to perform stability abnormality analysis on the plurality of sensor data sequences, generate a plurality of stability coefficients, and locate abnormal parameters based on the plurality of stability coefficients; an abnormal pattern recognition module, used to perform abnormal pattern recognition on the abnormal stability coefficients corresponding to the abnormal parameters, and generate an abnormal pattern recognition result; a matching module, used to match the abnormal pattern recognition result with a preset abnormal pattern list, wherein the preset abnormal pattern list comprises a plurality of preset abnormal patterns and a plurality of corresponding preset warning levels; a failure warning module, used to obtain a warning level according to the matching result, and perform a failure warning of the target energy storage battery according to the warning level.
[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:
[0007] By deploying a sensor array in the target energy storage battery, real-time collection and monitoring of battery operating parameters are achieved, so that abnormal conditions can be discovered in time and early warnings can be issued; by performing stability anomaly analysis on the sensor data sequence, multiple stability coefficients can be generated and abnormal parameters can be accurately located, which helps to accurately identify anomalies and improve the efficiency of troubleshooting; by identifying the abnormal level of the abnormal stability coefficient of the abnormal parameter and matching it with the preset abnormal level list, a clear early warning level can be generated, allowing the operator to take appropriate countermeasures according to the early warning level.
[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 schematic diagram of the structure of a large-capacity energy storage battery failure warning system based on pattern recognition provided in an embodiment of the present application;
[0010] Figure 2 A schematic diagram of the sensor array structure of a large-capacity energy storage battery failure warning system based on pattern recognition provided in an embodiment of the present application; Figure 3 This is a diagram of the internal structure of a computer device provided in an embodiment of the present application.
[0011] Explanation of reference numerals: sensor array layout module 10 , operation parameter acquisition module 20 , abnormality analysis module 30 , abnormal pattern recognition module 40 , matching module 50 , failure warning module 60 . DETAILED DESCRIPTION
[0012] The embodiment of the present application solves the technical problem that the state monitoring and warning system of traditional large-capacity energy storage batteries lacks real-time and accuracy, resulting in poor timeliness of failure warning, by providing a large-capacity energy storage battery failure warning system based on pattern recognition.
[0013] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically introduced below in conjunction with the accompanying drawings of the specification. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0014] like Figure 1 As shown, the embodiment of the present application provides a large-capacity energy storage battery failure warning system based on pattern recognition, and the system includes:
[0015] A sensor array deployment module 10 is used to deploy a sensor array in a target energy storage battery, wherein the sensor array includes a plurality of sensors;
[0016] First, determine the type of sensor suitable for monitoring the state of the energy storage battery, such as temperature sensors, voltage sensors, charge and discharge rate sensors, etc. According to the monitoring requirements, determine the number and layout of sensors. Usually, multiple sensors need to be deployed inside the battery to cover different areas and parameters; based on the structure and characteristics of the battery, choose a suitable location to deploy the sensor inside the battery according to the internal structure of the battery, temperature distribution, voltage changes and other factors. Sensors should be evenly distributed inside the battery to ensure that the state changes of the battery can be accurately monitored.
[0017] An operating parameter acquisition module 20, configured to acquire battery operating parameters through the multiple sensors within a preset time zone to obtain multiple sensor data sequences, wherein the multiple sensor data sequences are provided with time marks;
[0018] Determine the time range of the preset time zone. This time range can be determined according to demand, such as according to the battery usage and monitoring cycle settings. In the preset time zone, multiple sensors are used to monitor the battery at the same time. Each sensor collects corresponding battery operating parameter data according to its location and type, including temperature, voltage, current, etc., and saves the collected data as a data sequence according to the device time mark. The time mark is used to identify the time point of data collection.
[0019] An abnormality analysis module 30, used to perform stability abnormality analysis on the plurality of sensor data sequences, generate a plurality of stability coefficients, and locate abnormal parameters based on the plurality of stability coefficients;
[0020] Stability indicators are calculated for each sensor data sequence. These indicators may include mean, standard deviation, variance, coefficient of variation, etc., which are used to describe the stability of the data sequence. For example, the mean and standard deviation of each sensor data sequence are calculated as stability indicators. Based on the stability indicators, they are combined to generate a stability coefficient to reflect the overall stability of the data sequence.
[0021] By comparing the generated stability coefficient with the preset stability interval, the abnormal parameters can be located. Specifically, when the stability coefficient exceeds the preset range, it can be determined that the corresponding sensor data sequence is abnormal.
[0022] An abnormal pattern recognition module 40 is used to perform abnormal pattern recognition on the abnormal stability coefficient corresponding to the abnormal parameter and generate an abnormal pattern recognition result;
[0023] Different types of abnormal patterns are predefined. These abnormal patterns can be known and empirical, and each abnormal pattern has its own specific characteristics. For the abnormal stability coefficient corresponding to each abnormal parameter, machine learning methods such as support vector machines and neural networks are used to perform abnormal pattern recognition. The goal of abnormal pattern recognition is to distinguish abnormal data points from normal data points and identify different types of abnormal patterns for subsequent warnings. Based on the output of the abnormal pattern recognition algorithm, the abnormal pattern recognition result is generated.
[0024] A matching module 50, configured to match the abnormal pattern recognition result with a preset abnormal pattern list, wherein the preset abnormal pattern list includes a plurality of preset abnormal patterns and a corresponding plurality of preset warning levels;
[0025] Obtain a preset abnormal pattern list, which includes a variety of known abnormal patterns and corresponding preset warning levels. These abnormal patterns can be determined based on past experience, knowledge of domain experts, or results of data analysis.
[0026] The abnormal pattern recognition results are matched with the preset abnormal pattern list. For the identified abnormal pattern, check whether it matches a certain abnormal pattern in the preset abnormal pattern list. The matching can be based on the characteristics and attributes of the abnormal pattern. The preset abnormal pattern with the largest matching degree is obtained as the matching abnormal pattern. The corresponding preset warning level is obtained from the preset abnormal pattern list. This warning level can indicate the severity and urgency of the abnormality, thereby determining the subsequent response measures.
[0027] The failure warning module 60 is used to obtain a warning level according to the matching result, and to issue a failure warning for the target energy storage battery according to the warning level.
[0028] The preset warning level is used as the current warning level, and corresponding warning information is generated according to the warning level. The warning information may include a description of the abnormal mode, the time of occurrence, the location, and the corresponding warning level. The warning information is conveyed to relevant managers through system notifications, alarm sounds, graphical interfaces, etc. to deal with abnormal situations in the battery system.
[0029] Furthermore, if Figure 2 As shown, the sensor array includes a plurality of sensors, and the plurality of sensors include at least a temperature sensor, a voltage sensor, and a charge and discharge rate sensor.
[0030] The sensor array includes at least a temperature sensor, a voltage sensor, and a charge and discharge rate sensor, which are used to monitor the key parameters of the battery in the battery system. The temperature sensor is used to monitor the temperature change of the battery, the voltage sensor is used to monitor the voltage state of the battery, and the charge and discharge rate sensor is used to monitor the charge and discharge rate of the battery. Through the data collection of these sensors, the operating status of the battery can be more comprehensively understood, thereby improving the safety and reliability of the battery.
[0031] Furthermore, the system further comprises:
[0032] a deviation calculation module, used for acquiring a first sensing data sequence, performing mean calculation and standard deviation calculation on the first sensing data sequence, and generating a first data deviation central coefficient and a first data deviation dispersion coefficient;
[0033] A comprehensive calculation module, used for comprehensively calculating the first data deviation central coefficient and the first data deviation dispersion coefficient to obtain a first stability coefficient;
[0034] A coefficient adding module is used to add the first stability coefficient to the plurality of stability coefficients.
[0035] Randomly extract one from multiple sensor data sequences as the first sensor data sequence, calculate the mean of the first sensor data sequence, calculate the absolute value of the difference between each sensor data in the first sensor data sequence and the mean, then sum these differences, and finally divide them by the number of data and the mean to obtain the first data deviation center coefficient. The first data deviation center coefficient describes the change of the sensor data relative to the mean. The smaller the value, the closer the sensor data is to the mean, and vice versa, the greater the degree of dispersion;
[0036] The standard deviation is calculated for all sensor data in the first sensor data sequence. The standard deviation is the square root of the average of the squares of the differences between each sensor data and the mean. The standard deviation is divided by the mean to obtain the first data deviation dispersion coefficient. The first data deviation dispersion coefficient describes the ratio of the degree of change of the sensor data to its mean. The larger the value, the higher the dispersion of the delay time, and vice versa.
[0037] The first data deviation center coefficient and the first data deviation dispersion coefficient are normalized so that their range is between 0 and 1. After normalization, they are weighted summed according to preset weights to obtain a first stability coefficient. This stability coefficient can reflect the overall stability and degree of change of the data sequence, and is thus used for anomaly detection.
[0038] The first stability coefficient is added to the multiple stability coefficients, and the multiple sensing data sequences are traversed in sequence to calculate and obtain the multiple stability coefficients.
[0039] Furthermore, the system further comprises:
[0040] A weight index acquisition module is used to obtain a preset data deviation central coefficient weight index and a preset data deviation discrete coefficient weight index;
[0041] A weighted summation module, configured to perform weighted summation of the first data deviation central coefficient and the first data deviation discrete coefficient based on the preset data deviation central coefficient weight index and the preset data deviation discrete coefficient weight index to obtain a first data deviation factor;
[0042] A first stability coefficient acquisition module is used to obtain the first stability coefficient based on the first data deviation factor, wherein the first stability coefficient is inversely proportional to the first stability coefficient.
[0043] Determine the preset data deviation central coefficient weight index and the preset data deviation discrete coefficient weight index. These weight indicators can be determined according to system requirements and specific scenarios, for example, they can be specified through expert knowledge or data analysis.
[0044] Based on the preset data deviation center coefficient weight index, the first data deviation center coefficient is weighted, that is, the preset data deviation center coefficient weight index is multiplied by the first data deviation center coefficient; based on the preset data deviation dispersion coefficient weight index, the first data deviation dispersion coefficient is weighted, that is, the preset data deviation dispersion coefficient weight index is multiplied by the first data deviation dispersion coefficient. The weighted results are added together to obtain the first data deviation factor, which can reflect the overall deviation of the data.
[0045] The first stability coefficient is calculated according to the first data deviation factor. For example, the first data deviation factor is normalized to between 0 and 1, and the value of 1 minus the first data deviation factor is calculated to obtain the first stability coefficient.
[0046] Furthermore, the system further comprises:
[0047] A preset interval acquisition module, used to acquire a preset stability coefficient interval, wherein the preset stability coefficient interval includes a maximum preset stability coefficient and a minimum preset stability coefficient;
[0048] A stability coefficient judgment module, used for sequentially traversing the multiple stability coefficients and judging whether the multiple stability coefficients meet the preset stability coefficient interval;
[0049] A marking instruction generating module, used for generating a marking instruction when the conditions are not met;
[0050] An abnormal parameter marking module is used to mark the parameter corresponding to the current stability coefficient as an abnormal parameter based on the marking instruction.
[0051] The preset stability coefficient interval includes the maximum preset stability coefficient and the minimum preset stability coefficient. This interval can be determined according to the system requirements and the requirements of the stability analysis. For each stability coefficient, it is determined whether it is within the preset stability coefficient interval. If it is, it indicates that the stability coefficient is normal; otherwise, it indicates that there is an abnormality. When it is detected that the stability coefficient is not within the preset stability coefficient interval, a marking instruction is generated to mark the abnormal parameter. According to the marking instruction, the parameter corresponding to the current stability coefficient is marked as an abnormal parameter, so as to promptly handle the abnormal situation in the system and ensure the stability and reliability of the system.
[0052] Furthermore, the system further comprises:
[0053] A skip instruction generation module, used for generating a skip instruction when the conditions are met;
[0054] The skip operation module is used to perform a skip operation on the current stability coefficient based on the skip instruction.
[0055] When the stability coefficient is detected to be within the preset stability coefficient range, a skip instruction is generated, indicating that the current parameter is in a normal state and no further processing is required. According to the generated skip instruction, the corresponding skip operation is performed, that is, the abnormal processing step of the current stability coefficient is skipped, thereby realizing flexible management of system operation.
[0056] Furthermore, the system further comprises:
[0057] A historical record acquisition module, used to acquire historical abnormal pattern recognition records, acquire a sample abnormal stability coefficient set, and a sample abnormal pattern recognition result set;
[0058] An identifier construction module, used to construct an abnormal pattern identifier by using the sample abnormal stability coefficient set and the sample abnormal pattern recognition result set;
[0059] The recognition result acquisition module is used to use the abnormal pattern recognizer to perform abnormal pattern recognition on the abnormal stability coefficient and output the abnormal pattern recognition result.
[0060] Obtain historical abnormal pattern recognition records from the system or database, which include abnormal situations detected by the system in the past period of time, abnormal stability coefficients, abnormal pattern recognition results, and other information. Extract a set of sample abnormal stability coefficients from the historical abnormal pattern recognition records, which reflect the stability of various abnormal situations that have appeared in the system in the past period of time; similarly, extract a set of sample abnormal pattern recognition results from the historical abnormal pattern recognition records, which describe various abnormal patterns detected by the system in the past period of time.
[0061] The sample abnormal stability coefficient set and the sample abnormal pattern recognition result set are used to form a training data set for constructing an abnormal pattern recognizer, wherein the abnormal stability coefficient is used as an input feature and the abnormal pattern recognition result is used as a label.
[0062] An abnormal pattern recognizer model is constructed based on a neural network, and the model is trained on the prepared training data set. The training process aims to enable the model to accurately identify abnormal patterns from the abnormal stability coefficient and associate it with the preset abnormal level. After the training is completed, the model is evaluated using a test data set to evaluate the performance and accuracy of the model. According to the evaluation results, the model is tuned to improve the performance and generalization ability of the model. The trained and evaluated abnormal pattern recognizer can be deployed in the actual system to identify abnormal patterns.
[0063] The abnormal stability coefficient to be identified is taken as input data and passed to the constructed abnormal pattern recognizer, and the model is called for identification. According to the recognition result of the model, the abnormal pattern recognition result is generated.
[0064] Furthermore, the system further comprises:
[0065] The warning recording module is used to record the warning reaction time and warning processing effect with the failure warning time node as the starting point;
[0066] A first adjustment coefficient generating module, used to generate a first adjustment coefficient based on the warning reaction time;
[0067] A second adjustment coefficient generating module, used for generating a second adjustment coefficient based on the warning processing effect;
[0068] The warning level adjustment module is used to adjust the preset warning level of the preset abnormal mode list based on the first adjustment coefficient and the second adjustment coefficient.
[0069] For warning events, calculate the warning reaction time. The warning reaction time refers to the time interval from the issuance of the warning to the operator starting to take corresponding measures. It can be determined by calculating the time difference between the issuance of the warning and the actual start of warning processing. At the same time, record the effect after the warning processing, such as whether the potential problem is successfully solved, whether the loss is stopped or the abnormal situation is repaired in time. The warning processing effect reflects the effectiveness of the measures taken after the warning.
[0070] The first adjustment coefficient is generated based on the warning reaction time. Generally, the shorter the warning reaction time is, the stronger the system response capability is. Therefore, the first adjustment coefficient can be determined according to the length of the reaction time. The second adjustment coefficient is generated based on the warning processing effect. If the warning processing effect is good, the stability and reliability of the system will be improved. Therefore, the second adjustment coefficient can be determined according to the quality of the processing effect.
[0071] According to the values of the first adjustment coefficient and the second adjustment coefficient, the preset warning level in the preset abnormal mode list is adjusted. Specifically, if the first adjustment coefficient is small, that is, the warning reaction time is long, it may be necessary to increase the warning level to strengthen the monitoring and warning of abnormal situations, as well as to speed up the processing speed; if the second adjustment coefficient is small, that is, the warning processing effect is poor, it may be necessary to increase the warning level to ensure that the system can respond to abnormal situations and solve problems more promptly.
[0072] By adjusting the preset warning level of the preset abnormal mode list according to the first adjustment coefficient and the second adjustment coefficient, it is possible to better adapt to the current abnormal situation of the system and improve the accuracy and practicality of the warning.
[0073] 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 large-capacity energy storage battery failure warning system based on pattern recognition, characterized in that: The system comprises: A sensor array deployment module, used to deploy a sensor array in a target energy storage battery, wherein the sensor array includes a plurality of sensors; An operating parameter acquisition module, used to acquire battery operating parameters through the multiple sensors within a preset time zone to obtain multiple sensor data sequences, wherein the multiple sensor data sequences are provided with time marks; an abnormality analysis module, used for performing stability abnormality analysis on the plurality of sensor data sequences, generating a plurality of stability coefficients, and locating abnormal parameters based on the plurality of stability coefficients; An abnormal pattern recognition module is used to perform abnormal pattern recognition on the abnormal stability coefficient corresponding to the abnormal parameter and generate an abnormal pattern recognition result; A matching module, used for matching the abnormal pattern recognition result with a preset abnormal pattern list, wherein the preset abnormal pattern list includes a plurality of preset abnormal patterns and a corresponding plurality of preset warning levels; The failure warning module is used to obtain a warning level according to the matching result, and to issue a failure warning for the target energy storage battery according to the warning level.
2. The system according to claim 1, characterized in that The multiple sensors include at least a temperature sensor, a voltage sensor, and a charge and discharge rate sensor.
3. The system according to claim 1, characterized in that Performing stability anomaly analysis on the plurality of sensor data sequences to generate a plurality of stability coefficients, including: a deviation calculation module, used for acquiring a first sensing data sequence, performing mean calculation and standard deviation calculation on the first sensing data sequence, and generating a first data deviation central coefficient and a first data deviation dispersion coefficient; A comprehensive calculation module, used for comprehensively calculating the first data deviation central coefficient and the first data deviation dispersion coefficient to obtain a first stability coefficient; A coefficient adding module is used to add the first stability coefficient to the plurality of stability coefficients.
4. The system according to claim 3, characterized in that Comprehensively calculating the first data deviation center coefficient and the first data deviation dispersion coefficient to obtain a first stability coefficient includes: A weight index acquisition module is used to obtain a preset data deviation central coefficient weight index and a preset data deviation discrete coefficient weight index; A weighted summation module, configured to perform weighted summation of the first data deviation central coefficient and the first data deviation discrete coefficient based on the preset data deviation central coefficient weight index and the preset data deviation discrete coefficient weight index to obtain a first data deviation factor; A first stability coefficient acquisition module is used to obtain the first stability coefficient based on the first data deviation factor, wherein the first stability coefficient is inversely proportional to the first stability coefficient.
5. The system according to claim 1, wherein: Locating abnormal parameters based on the multiple stability coefficients includes: A preset interval acquisition module, used to acquire a preset stability coefficient interval, wherein the preset stability coefficient interval includes a maximum preset stability coefficient and a minimum preset stability coefficient; A stability coefficient judgment module, used for sequentially traversing the multiple stability coefficients and judging whether the multiple stability coefficients meet the preset stability coefficient interval; A marking instruction generating module, used for generating a marking instruction when the conditions are not met; An abnormal parameter marking module is used to mark the parameter corresponding to the current stability coefficient as an abnormal parameter based on the marking instruction.
6. The system according to claim 5, characterized in that Also includes: A skip instruction generation module, used for generating a skip instruction when the conditions are met; The skip operation module is used to perform a skip operation on the current stability coefficient based on the skip instruction.
7. The system according to claim 1, characterized in that Performing abnormal pattern recognition on the abnormal stability coefficient corresponding to the abnormal parameter to generate an abnormal pattern recognition result includes: A historical record acquisition module, used to acquire historical abnormal pattern recognition records, acquire a sample abnormal stability coefficient set, and a sample abnormal pattern recognition result set; An identifier construction module, used to construct an abnormal pattern identifier by using the sample abnormal stability coefficient set and the sample abnormal pattern recognition result set; The recognition result acquisition module is used to use the abnormal pattern recognizer to perform abnormal pattern recognition on the abnormal stability coefficient and output the abnormal pattern recognition result.
8. The system of claim 1, wherein: Also includes: The warning recording module is used to record the warning reaction time and warning processing effect with the failure warning time node as the starting point; A first adjustment coefficient generating module, used to generate a first adjustment coefficient based on the warning reaction time; A second adjustment coefficient generating module, used for generating a second adjustment coefficient based on the warning processing effect; The warning level adjustment module is used to adjust the preset warning level of the preset abnormal mode list based on the first adjustment coefficient and the second adjustment coefficient.
