An energy storage power station operation and maintenance management method and system integrating AI and the Internet of Things

Abstract

The present invention discloses a method for operation and maintenance management of energy storage power stations that integrates AI and the Internet of Things. The method uses edge computing technology to preprocess status monitoring data, and the data is uploaded to the operation and maintenance management center only after preprocessing, which reduces the computing resource usage of the operation and maintenance management center and is conducive to shortening the early warning response time. Moreover, when preprocessing the status monitoring data, preliminary anomaly detection is performed through a matrix profile dynamic time warping algorithm, and suspicious data segments are marked, which can further reduce the computing resource usage of the operation and maintenance management center and further shorten the early warning response time.

Description

Technical Field

[0001] The present invention relates to the field of energy storage power stations, and in particular to an energy storage power station operation and maintenance management method and system integrating AI and the Internet of Things. Background Art

[0002] With the booming development of the new energy industry and the continued advancement of smart grid construction, the importance of intelligent status monitoring for energy storage power stations, as key facilities for ensuring the stable operation of power systems, has become increasingly prominent. Traditional energy storage power station status monitoring relies primarily on manual inspections and data monitoring. This model is not only inefficient but also makes it difficult to detect potential equipment failures in advance, increasing the operation and maintenance costs of energy storage power stations.

[0003] In recent years, the deep integration of artificial intelligence and the Internet of Things (IoT) has provided a new solution for energy storage power plant condition monitoring. The widespread deployment of IoT sensors enables the real-time collection and transmission of operating data from various equipment within energy storage power plants. Artificial intelligence algorithms can perform predictive analysis of equipment status based on historical data, significantly improving the accuracy and timeliness of condition monitoring.

[0004] However, despite the significant success of these technologies, intelligent condition monitoring in energy storage power plants still faces numerous challenges. For one thing, energy storage power plants are complex structures, involving multiple components such as battery packs, converters, cooling systems, and fire protection systems. This makes data analysis and processing challenging, impacting the accuracy of condition warnings. Furthermore, existing intelligent monitoring algorithms lack the ability to provide early warnings for tasks such as equipment performance degradation. Furthermore, some algorithms require high hardware computing resources, making efficient deployment at the edge difficult and limiting the response speed of real-time condition monitoring. Summary of the Invention

[0005] The main advantage of the present invention is that it provides an energy storage power station operation and maintenance management method that integrates AI and the Internet of Things. It uses edge computing technology to preprocess status monitoring data, and the data is uploaded to the operation and maintenance management center only after preprocessing, which reduces the computing resource usage of the operation and maintenance management center and helps shorten the early warning response time.

[0006] Another advantage of the present invention is that it provides an energy storage power station operation and maintenance management method that integrates AI and the Internet of Things. When pre-processing the status monitoring data, it uses a matrix profile dynamic time warping algorithm to perform preliminary anomaly detection and mark suspicious data segments, which can further reduce the computing resource usage of the operation and maintenance management center and further shorten the early warning response time.

[0007] Another advantage of the present invention is that it provides an energy storage power station operation and maintenance management method that integrates AI and the Internet of Things. The time series prediction model of its intelligent early warning model includes a discrete cosine transform layer, a time convolution network layer, and a state space model layer. Through the discrete cosine transform frequency domain feature extraction, the time convolution network and the time domain-frequency domain matrix addition fusion mechanism, the false alarm rate of abnormal warning is reduced.

[0008] Accordingly, according to an embodiment of the present invention, a method for operation and maintenance management of an energy storage power station integrating AI and the Internet of Things, which has at least one of the aforementioned advantages, includes the following steps:

[0009] S1. Use IoT technology to connect various devices within the energy storage power station into a network, collect real-time status monitoring data from key equipment, and use edge computing technology to pre-process the status monitoring data. During the pre-processing of the status monitoring data, a matrix profile dynamic time warping algorithm is used to perform preliminary anomaly detection and mark suspicious data segments.

[0010] S2. Upload the pre-processed status monitoring data to the operation and maintenance management center;

[0011] S3, the operation and maintenance management center uses pre-processed status monitoring data to train intelligent early warning models based on artificial intelligence algorithms;

[0012] S4. Monitor the status of key equipment in the energy storage power station through an intelligent early warning model, where the intelligent early warning model includes a time series prediction model and a time series anomaly detection model. The time series prediction model obtains prediction data based on the input status monitoring data, and the time series anomaly detection model obtains the energy storage power station status warning result based on the input prediction data. The time series prediction model includes a discrete cosine transform layer, a time convolutional network layer, and a state space model layer. The discrete cosine transform layer is used to extract frequency domain features, and the state space model layer performs dynamic modeling based on the time domain-frequency domain matrix summation and fusion features;

[0013] S5. When the energy storage power station status warning result is abnormal, the operation and maintenance disposal procedure is triggered.

[0014] In some embodiments of the present invention, when preprocessing the condition monitoring data, the steps include:

[0015] Establish dynamic data verification rules to verify the validity of the original data of the collected condition monitoring data and eliminate abnormal data points;

[0016] The sliding window mechanism is used to smooth the time series data of condition monitoring data and extract statistical features;

[0017] Mark suspicious data segments by calculating the minimum dynamic time warping distance of the subsequence of time series data;

[0018] Standardize the suspicious data segments to eliminate the dimensional differences between different features.

[0019] In some embodiments of the present invention, the minimum dynamic time warping distance of a subsequence is calculated as follows:

[0020] ,in,

[0021] MP[i] is the i-th subsequence T i The minimum dynamic time warping distance, j is the traversal index of the non-adjacent subsequence, is the neighborhood exclusion radius, T i The length of the i-th Subsequence of T j is the j-th subsequence, j is not equal to i, and DTW is a dynamic alignment distance metric function.

[0022] In some embodiments of the present invention, step S2 specifically includes the steps of:

[0023] Set the data synchronization frequency in the operation and maintenance management center to ensure that the status monitoring data of the energy storage power station can be uploaded regularly;

[0024] The pre-processed condition monitoring data is encrypted through the energy storage power station's edge computing module, and the encryption keys are securely managed and regularly updated through the operation and maintenance management center.

[0025] After receiving data from the energy storage power station, the operation and maintenance management center conducts data quality checks to ensure data integrity.

[0026] In some embodiments of the present invention, when using preprocessed status monitoring data to train an intelligent early warning model, the preprocessed status monitoring data is divided into non-overlapping training sets and test sets. The training set is used to train the intelligent early warning model, and the test set is used to evaluate the accuracy of the intelligent early warning model.

[0027] In some embodiments of the present invention, the time-domain-frequency-domain matrix addition fusion feature is a hybrid feature formed by fusing the frequency-domain feature output by the discrete cosine transform layer and the time-domain data of the state monitoring data through a matrix addition operation.

[0028] In some embodiments of the present invention, the temporal convolutional network layer is a multi-layer dilated temporal convolutional network structure, which is used to extract local details and long-term trends of the time series of condition monitoring data.

[0029] Accordingly, the present invention further provides an energy storage power station operation and maintenance management system integrating AI and the Internet of Things, which is used to implement the energy storage power station operation and maintenance management method integrating AI and the Internet of Things, including:

[0030] Data acquisition module, including sensors deployed at key monitoring nodes in the energy storage power station;

[0031] The edge computing module deployed at the edge of the system is equipped with an embedded processing unit with data cleaning functions, matrix profile dynamic time warping anomaly detection functions, and statistical feature extraction functions;

[0032] A prediction and decision support module deployed in the operation and maintenance management center, which is used to build intelligent early warning models through machine learning algorithms and analytical techniques; and

[0033] The operation and maintenance execution module deployed in the operation and maintenance management center integrates the work order management system, emergency response device and maintenance decision support system. Among them, the operation and maintenance management center is the cloud side of the system.

[0034] Accordingly, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the energy storage power station operation and maintenance management method integrating AI and the Internet of Things.

[0035] Accordingly, the present invention further provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the energy storage power station operation and maintenance management method integrating AI and the Internet of Things is implemented.

[0036] The above and other advantages of the present invention will be fully reflected in conjunction with the following description and the accompanying drawings.

[0037] The above and other advantages and features of the present invention are fully reflected in the following detailed description of the present invention and the accompanying drawings.

[0038] The summary of the invention is not considered to identify the essential technical features of the invention, nor is it considered to limit the scope of protection of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 1 is a flow chart of an energy storage power station operation and maintenance management method integrating AI and the Internet of Things according to an embodiment of the present invention. DETAILED DESCRIPTION

[0040] The following description is provided to enable those skilled in the art to implement the present invention. Those skilled in the art may conceive of other obvious replacements, modifications, and variations. Therefore, the scope of protection of the present invention should not be limited by the exemplary embodiments described herein.

[0041] Those skilled in the art should understand that, unless otherwise specified herein, the term "one" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element may be one, while in another embodiment, the number of the element may be multiple.

[0042] Those skilled in the art should understand that, unless otherwise specified herein, the directions or positions referred to by terms such as "longitudinal," "transverse," "upper," "lower," "front," "back," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer" are based on the directions or positions shown in the accompanying drawings and are intended merely to facilitate description of the present invention. They do not indicate or imply that the devices or components involved must have a specific direction or position. Therefore, the above terms should not be construed as limiting the present invention.

[0043] Refer to the appendix of the present invention Figure 1 , the energy storage power station operation and maintenance management method integrating AI and the Internet of Things according to an embodiment of the present invention is explained. The energy storage power station operation and maintenance management method integrating AI and the Internet of Things includes the following steps:

[0044] S1. Use IoT technology to connect various devices within the energy storage power station into a network, collect real-time status monitoring data from key equipment, and use edge computing technology to pre-process the status monitoring data. During the pre-processing of the status monitoring data, a matrix profile dynamic time warping algorithm is used to perform preliminary anomaly detection and mark suspicious data segments.

[0045] S2. Upload the pre-processed status monitoring data to the operation and maintenance management center;

[0046] S3, the operation and maintenance management center uses pre-processed status monitoring data to train intelligent early warning models based on artificial intelligence algorithms;

[0047] S4. Monitor the status of key equipment in the energy storage power station through an intelligent early warning model, where the intelligent early warning model includes a time series prediction model and a time series anomaly detection model. The time series prediction model obtains prediction data based on the input status monitoring data, and the time series anomaly detection model obtains the energy storage power station status warning result based on the input prediction data. The time series prediction model includes a discrete cosine transform layer, a time convolutional network layer, and a state space model layer. The discrete cosine transform layer is used to extract frequency domain features, and the state space model layer performs dynamic modeling based on the time domain-frequency domain matrix summation and fusion features;

[0048] S5. When the energy storage power station status warning result is abnormal, the operation and maintenance disposal procedure is triggered.

[0049] It can be understood that edge computing technology can be used to push data processing from the traditional cloud to the edge side of the system, that is, a location closer to the data source or terminal device. In the present invention, the status monitoring data can be pre-processed by an edge computing module deployed on the edge side of the system, and the data is uploaded to the operation and maintenance management center, that is, the cloud side of the system after pre-processing, which reduces the computing resource usage of the operation and maintenance management center and helps to shorten the early warning response time. When pre-processing the status monitoring data, the edge computing module can also perform preliminary anomaly detection through the matrix profile dynamic time warping algorithm and mark suspicious data segments, which can further reduce the computing resource usage of the operation and maintenance management center and further shorten the early warning response time.

[0050] It is worth mentioning that condition monitoring data is collected through sensors deployed at key monitoring nodes in the energy storage power station, such as the energy storage system, converter system, and environmental monitoring nodes. This includes SCADA (Supervisory Control and Data Acquisition) data on battery pack voltage / current, energy storage converter input / output power, battery pack charge and discharge power, ambient temperature, humidity, hazardous gas concentration, and switch status of circuit breakers, contactors, etc.; and BMS (Battery Management System) data on single cell voltage / temperature / SOC dispersion, battery cluster total voltage / total current / insulation resistance / SOH, overvoltage, undervoltage, and short-circuit protection records, and battery internal resistance change trends.

[0051] Specifically, when preprocessing the condition monitoring data, the following steps are included:

[0052] Establish dynamic data verification rules to verify the validity of the original data of the collected condition monitoring data and eliminate abnormal data points;

[0053] The sliding window mechanism is used to smooth the time series data of condition monitoring data and extract statistical features;

[0054] Mark suspicious data segments by calculating the minimum dynamic time warping distance of the subsequence of time series data;

[0055] Standardize the suspicious data segments to eliminate the dimensional differences between different features.

[0056] More specifically, the dynamic data verification rule is based on sliding window statistical analysis, and uses adaptive threshold detection to determine whether the current data exceeds the dynamic upper and lower limits constructed by the historical mean and standard deviation; if it exceeds, it is regarded as an abnormal data point; otherwise, the data change rate is further calculated. If the change rate exceeds the preset threshold, it is also judged as abnormal.

[0057] Preferably, in the dynamic data verification rule, if the amplitude of a data point is greater than or less than a set percentage of the average amplitude, it is regarded as an abnormal data point.

[0058] It can be understood that statistical feature extraction includes but is not limited to statistical feature extraction of battery pack voltage / current data, energy storage converter input / output power, battery pack charge and discharge power, ambient temperature, etc.

[0059] Preferably, the calculation formula for the minimum dynamic time warping distance of a subsequence is:

[0060] ,in,

[0061] MP[i] is the i-th subsequence T i The minimum dynamic time warping distance, j is the traversal index of the non-adjacent subsequence, is the neighborhood exclusion radius, T i The length of the i-th Subsequence of T j is the j-th subsequence, j is not equal to i, and DTW is a dynamic alignment distance metric function.

[0062] Specifically, the present invention deploys sensors to collect real-time status monitoring data from key equipment in energy storage power plants. First, based on dynamic data verification rules (for example, data point amplitudes exceeding ±A% of the historical mean are identified as abnormal and eliminated), the sliding window mechanism is used to smooth the time series data and extract statistical features. The MP-DTW (Matrix Profile with Dynamic Time Warping) algorithm is then used to mark suspicious data segments (such as sudden changes in battery voltage or temperature) by calculating the minimum DTW distance of subsequences. The cleaned data is then normalized to eliminate dimensional differences.

[0063] Furthermore, step S2 of the energy storage power station operation and maintenance management method integrating AI and the Internet of Things described in the present invention specifically includes the following steps:

[0064] Set the data synchronization frequency in the operation and maintenance management center to ensure that the status monitoring data of the energy storage power station can be uploaded regularly;

[0065] The pre-processed condition monitoring data is encrypted through the energy storage power station's edge computing module, and the encryption keys are securely managed and regularly updated through the operation and maintenance management center.

[0066] After receiving data from the energy storage power station, the operation and maintenance management center conducts data quality checks to ensure data integrity.

[0067] Preferably, encryption can be performed using the Advanced Encryption Standard encryption algorithm. More preferably, the key length is set to 192 bits.

[0068] Preferably, after receiving data from the energy storage power station, the operation and maintenance management center uses the SHA algorithm to perform data quality checks.

[0069] Furthermore, when using the preprocessed status monitoring data to train the intelligent early warning model, the preprocessed status monitoring data is divided into non-overlapping training sets and test sets. The training set is used to train the intelligent early warning model, and the test set is used to evaluate the accuracy of the intelligent early warning model, thereby avoiding negative impacts on model training and accuracy evaluation due to data contamination.

[0070] Furthermore, the time-domain-frequency domain matrix addition fusion feature is a hybrid feature formed by fusing the frequency domain features output by the discrete cosine transform layer and the time domain data of the status monitoring data through a matrix addition operation. The specific formula is:

[0071] ;

[0072] F fusion =X+F freq ;

[0073] Where X is the state monitoring data, T is the sequence length, c is the number of sensor channels, and F freq is the frequency domain feature output by the discrete cosine transform layer; F fusion It is the time domain-frequency domain matrix sum fusion feature.

[0074] Furthermore, the temporal convolutional network layer is a multi-layer dilated time convolutional network structure, which is used to extract local details and long-term trends of the time series of condition monitoring data, which is conducive to improving the warning accuracy of the intelligent early warning model. The formula of the multi-layer dilated time convolutional network is:

[0075] ;

[0076] Where H (l) is the output of layer l, is the convolution kernel weight of the lth layer, 2 l-1 is the expansion factor.

[0077] It can be understood that the present invention reduces the false alarm rate of abnormal warning through discrete cosine transform frequency domain feature extraction, time convolution network and time domain-frequency domain matrix addition fusion mechanism.

[0078] Specifically, the intelligent early warning model of the present invention includes a time series prediction model and a time series anomaly detection model. After the status monitoring data is input into the time series prediction model, predicted data is obtained, and the predicted data is input into the time series anomaly detection model to finally obtain the status warning result of the energy storage power station. The time series prediction model of the present invention includes a discrete cosine transform (DCT) layer, a temporal convolutional network (TCN) layer, and a state space model layer. The preprocessed multidimensional feature data is converted to the frequency domain by the DCT layer to extract periodic fluctuation characteristics, and then input into a multi-layer expanded TCN structure to further extract local details and long-term trends of the time series. The frequency domain features output by the TCN are fused with the time domain data of the status monitoring data through a matrix addition operation to form a hybrid feature. After that, it is input into the state space model layer for dynamic modeling and outputs the predicted data for the next moment. The time series anomaly detection model of the present invention divides the time series of the predicted data into non-overlapping subsequences of fixed length and converts them into D-dimensional embedding vectors through a learnable mapping process. Here, D is a hyperparameter of the model, representing the dimension of the embedding space. During the pre-training stage, these embedding vectors are randomly masked and replaced with special mask embeddings (MASK). The goal of pre-training is to allow the model to reconstruct the original input from the unmasked parts, improve its ability to understand the structure of the time series, and effectively learn the pattern of the time series by using a lightweight Transformer model architecture. The reconstruction error is used as a metric to identify anomalies. If there is a significant difference between the reconstructed time series and the original sequence, it indicates that there may be an anomaly, which in turn triggers the status warning mechanism of the energy storage power station and realizes the warning of abnormal status of the energy storage power station.

[0079] Preferably, the early warning of abnormal state of energy storage power station of the present invention adopts a multi-level early warning trigger mechanism, including early warning, mid-term warning and emergency warning, wherein,

[0080] Early warning: When the intelligent warning model outputs results indicating a slight abnormal change in the status of key equipment, but not yet reaching the level of failure, it alerts operation and maintenance personnel to pay attention to the operating status of related equipment and conduct timely inspections to identify potential hidden dangers in the early stages.

[0081] Mid-term warning: When the intelligent warning model output indicates a significant abnormal change in the status of critical equipment, indicating an impending equipment failure or a significant safety risk, a mid-term warning is activated, prompting operations and maintenance personnel to attach great importance to the situation and take appropriate measures immediately, such as strengthening equipment inspections, conducting detailed troubleshooting, preparing necessary maintenance tools and spare parts, and making preparations for emergency response.

[0082] Emergency warning: When the output results of the intelligent warning model indicate that there is a serious fault in key equipment, a series of emergency response measures will be initiated, and an emergency alarm will be issued to all relevant personnel, notifying professional departments such as fire protection and emergency rescue to come.

[0083] According to another aspect of the present invention, the present invention further provides an energy storage power station operation and maintenance management system integrating AI and the Internet of Things, which is used to implement the energy storage power station operation and maintenance management method integrating AI and the Internet of Things, including:

[0084] Data acquisition module, including sensors deployed at key monitoring nodes in the energy storage power station;

[0085] The edge computing module deployed at the edge of the system is equipped with an embedded processing unit with data cleaning functions, matrix profile dynamic time warping anomaly detection functions, and statistical feature extraction functions;

[0086] A prediction and decision support module deployed in the operation and maintenance management center, which is used to build intelligent early warning models through machine learning algorithms and analytical techniques; and

[0087] The operation and maintenance execution module deployed in the operation and maintenance management center integrates the work order management system, emergency response device and maintenance decision support system. Among them, the operation and maintenance management center is the cloud side of the system.

[0088] Specifically, the energy storage power station operation and maintenance management system that integrates AI and the Internet of Things of the present invention adopts a cloud, edge, and end collaborative architecture. The end side is a sensor for collecting status monitoring data of key equipment in the energy storage power station. An embedded processing unit is deployed on the edge side, which is responsible for the preprocessing of status monitoring data and dynamic time regularization anomaly detection of matrix profiles. The cloud is the operation and maintenance management center, which is responsible for the training and use of intelligent early warning models, and executes corresponding levels of disposal plans based on the output results of the intelligent early warning models.

[0089] According to another aspect of the present invention, the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the energy storage power station operation and maintenance management method integrating AI and the Internet of Things.

[0090] According to another aspect of the present invention, the present invention further provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the energy storage power station operation and maintenance management method integrating AI and the Internet of Things is implemented.

[0091] Those skilled in the art will appreciate that the above embodiments are merely examples, and features of different embodiments may be combined with each other to obtain implementation methods that are easily conceivable based on the disclosure of the present invention but are not explicitly indicated in the drawings.

[0092] Those skilled in the art will appreciate that the embodiments described above and shown in the accompanying drawings are intended only to illustrate the present invention and are not intended to limit the present invention. All equivalent implementations, modifications, and improvements within the spirit of the present invention are intended to fall within the scope of protection of the present invention.

Claims

1. A method for operation and maintenance management of energy storage power stations integrating AI and the Internet of Things, characterized in that: Including steps: S1. Use IoT technology to connect various devices within the energy storage power station into a network, collect real-time status monitoring data from key equipment, and use edge computing technology to pre-process the status monitoring data. During the pre-processing of the status monitoring data, a matrix profile dynamic time warping algorithm is used to perform preliminary anomaly detection and mark suspicious data segments. S2. Upload the pre-processed status monitoring data to the operation and maintenance management center; S3, the operation and maintenance management center uses pre-processed status monitoring data to train intelligent early warning models based on artificial intelligence algorithms; S4. Monitor the status of key equipment in the energy storage power station through an intelligent early warning model, where the intelligent early warning model includes a time series prediction model and a time series anomaly detection model. The time series prediction model obtains prediction data based on the input status monitoring data, and the time series anomaly detection model obtains the energy storage power station status warning result based on the input prediction data. The time series prediction model includes a discrete cosine transform layer, a time convolutional network layer, and a state space model layer. The discrete cosine transform layer is used to extract frequency domain features, and the state space model layer performs dynamic modeling based on the time domain-frequency domain matrix summation and fusion features; S5. When the energy storage power station status warning result is abnormal, the operation and maintenance disposal procedure is triggered; Among them, when preprocessing the condition monitoring data, the steps include: Mark suspicious data segments by calculating the minimum dynamic time warping distance of the subsequence of time series data; The calculation formula for the minimum dynamic time warping distance of a subsequence is: ,in, MP[i] is the i-th subsequence T i The minimum dynamic time warping distance, j is the traversal index of the non-adjacent subsequence, is the neighborhood exclusion radius, T i The length of the i-th Subsequence of T j is the j-th subsequence, j is not equal to i, and DTW is a dynamic alignment distance metric function.

2. The energy storage power station operation and maintenance management method integrating AI and the Internet of Things according to claim 1 is characterized in that: When preprocessing condition monitoring data, the following steps are also included: Establish dynamic data verification rules to verify the validity of the original data of the collected condition monitoring data and eliminate abnormal data points; The sliding window mechanism is used to smooth the time series data of condition monitoring data and extract statistical features; Standardize the suspicious data segments to eliminate the dimensional differences between different features.

3. The energy storage power station operation and maintenance management method integrating AI and the Internet of Things according to claim 2 is characterized in that: Step S2 specifically includes the following steps: Set the data synchronization frequency in the operation and maintenance management center to ensure that the status monitoring data of the energy storage power station can be uploaded regularly; The pre-processed condition monitoring data is encrypted through the energy storage power station's edge computing module, and the encryption keys are securely managed and regularly updated through the operation and maintenance management center. After receiving data from the energy storage power station, the operation and maintenance management center conducts data quality checks to ensure data integrity.

4. The energy storage power station operation and maintenance management method integrating AI and the Internet of Things according to claim 2 is characterized in that: When using the preprocessed condition monitoring data to train the intelligent early warning model, the preprocessed condition monitoring data is divided into a non-overlapping training set and a test set. The training set is used to train the intelligent early warning model, and the test set is used to evaluate the accuracy of the intelligent early warning model.

5. The energy storage power station operation and maintenance management method integrating AI and the Internet of Things according to claim 2 is characterized in that: The time-domain-frequency-domain matrix addition fusion feature is a hybrid feature formed by fusing the frequency-domain features output by the discrete cosine transform layer and the time-domain data of the state monitoring data through a matrix addition operation.

6. The energy storage power station operation and maintenance management method integrating AI and the Internet of Things according to claim 2 is characterized in that: The temporal convolutional network layer is a multi-layer dilated temporal convolutional network structure, which is used to extract local details and long-term trends of the time series of condition monitoring data.

7. An energy storage power station operation and maintenance management system integrating AI and the Internet of Things, used to implement the energy storage power station operation and maintenance management method integrating AI and the Internet of Things as described in any one of claims 1-6, characterized in that: include: Data acquisition module, including sensors deployed at key monitoring nodes in the energy storage power station; The edge computing module deployed at the edge of the system is equipped with an embedded processing unit with data cleaning functions, matrix profile dynamic time warping anomaly detection functions, and statistical feature extraction functions; A prediction and decision support module deployed in the operation and maintenance management center, which is used to build intelligent early warning models through machine learning algorithms and analytical techniques; and The operation and maintenance execution module deployed in the operation and maintenance management center integrates the work order management system, emergency response device and maintenance decision support system. Among them, the operation and maintenance management center is the cloud side of the system.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the energy storage power station operation and maintenance management method integrating AI and the Internet of Things as described in any one of claims 1 to 6 is implemented.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the energy storage power station operation and maintenance management method integrating AI and the Internet of Things as described in any one of claims 1 to 6 is implemented.

Images