A multi-modal state perception and intelligent diagnosis system for power equipment

Abstract

This invention discloses a multimodal state perception and intelligent diagnosis system for power equipment, belonging to the field of intelligent operation and maintenance technology for power equipment. The multimodal perception unit collects visible light imaging information, infrared thermal imaging information, ultraviolet imaging information, and acoustic imaging information from the power equipment; the data preprocessing unit performs time synchronization, spatial alignment, and data normalization on the multimodal perception data; the equipment identification unit identifies the type and key structures of the power equipment based on visible light imaging information; the multimodal fusion diagnosis unit selects appropriate multimodal fusion and diagnosis strategies according to equipment type information, performs joint analysis on the multimodal perception data, and outputs equipment state assessment results and / or fault diagnosis results; the result output unit displays, stores, or transmits the diagnostic results. This invention can improve the accuracy and reliability of power equipment state perception and has good engineering application value.

Description

Technical Field

[0001] This invention relates to the field of power equipment condition detection and intelligent operation and maintenance technology, specifically to a multimodal condition perception and intelligent diagnosis system that integrates visible light, infrared, ultraviolet and acoustic imaging, applicable to intelligent detection and fault diagnosis in power equipment operation scenarios such as substations. Background Technology

[0002] In intelligent operation and maintenance scenarios for power equipment, the operating environment is complex and the equipment types are diverse, making it difficult for a single sensing method to comprehensively and accurately reflect the equipment's operating status. Existing detection methods mostly rely on visible light or infrared imaging, which is insufficient for effectively identifying early faults such as partial discharge and hidden defects. Furthermore, different equipment types have significant differences in detection requirements and diagnostic rules, easily leading to misjudgments or missed diagnoses. Therefore, there is an urgent need for an intelligent detection technology solution that can integrate multiple sensing information and possess differentiated diagnostic capabilities for different equipment. Summary of the Invention

[0003] To address the problems of existing power equipment condition detection mainly relying on single sensing methods, insufficient information utilization, and inadequate adaptability to complex operating conditions, especially in typical application scenarios such as substations, where single visible light, infrared, or acoustic detection methods are difficult to comprehensively and accurately reflect the equipment operating status, easily leading to misjudgments or omissions, this invention proposes a multimodal condition perception and intelligent diagnosis system for power equipment.

[0004] The purpose of this invention is to construct a unified multimodal perception and fusion diagnostic system by introducing multiple sensing modalities such as visible light, infrared, ultraviolet and acoustic imaging, and to combine it with an equipment identification and guidance mechanism to achieve differentiated status assessment and fault diagnosis for different types of power equipment, thereby improving the accuracy, reliability and engineering applicability of power equipment status perception.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: This invention provides a multimodal state perception and intelligent diagnosis system for power equipment. The system includes a multimodal perception unit, a data preprocessing unit, an equipment identification unit, a multimodal fusion diagnosis unit, and a result output unit.

[0006] The multimodal sensing unit is used to collect multi-source information from the power equipment under test. The collected multi-source information includes at least visible light imaging information for acquiring the appearance and structure information of the power equipment, infrared thermal imaging information for acquiring the temperature distribution characteristics of the power equipment, ultraviolet imaging information for acquiring the partial discharge radiation characteristics of the power equipment, and acoustic imaging information for acquiring the discharge acoustic characteristics of the power equipment.

[0007] The data preprocessing unit is connected to the multimodal sensing unit and is used to perform time synchronization, spatial alignment and data normalization processing on the acquired visible light imaging information, infrared thermal imaging information, ultraviolet imaging information and acoustic imaging information to form multimodal sensing data that can be used for joint analysis.

[0008] The equipment identification unit is connected to the data preprocessing unit and is used to automatically identify power equipment based on visible light imaging information, and obtain equipment type information and / or key structural location information of the power equipment.

[0009] The multimodal fusion diagnostic unit is connected to the device identification unit and the data preprocessing unit. It is used to select a multimodal fusion and diagnostic strategy that matches the corresponding power equipment based on the acquired device type information, and to jointly analyze visible light imaging information, infrared thermal imaging information, ultraviolet imaging information and acoustic imaging information to output the status assessment results and / or fault diagnosis results of the power equipment.

[0010] The result output unit is connected to the multimodal fusion diagnostic unit and is used to display, store or transmit the state assessment results and / or fault diagnosis results to realize multimodal state perception and intelligent diagnosis of power equipment.

[0011] Furthermore, the system described in this invention can be used to implement a multimodal state perception and diagnosis process based on device identification guidance. This process is an example of one application method of the system, used to illustrate how the system is used, and does not constitute a limitation of this invention.

[0012] The present invention aims to provide a multimodal state perception and intelligent diagnosis system for power equipment. By constructing a collaborative architecture of multimodal perception unit and intelligent diagnosis unit, it realizes comprehensive perception and diagnosis of the operating status of power equipment.

[0013] The technical solution proposed in this invention mainly includes the following key points: 1. Construct a multimodal state perception system that integrates visible light, infrared, ultraviolet and acoustic imaging to achieve multi-dimensional information collection on the operating status of power equipment; 2. The equipment identification unit automatically identifies the type and key structure of power equipment, and uses this identification as a priori basis for multimodal fusion and diagnosis; 3. Based on different types of power equipment, dynamically select the corresponding multimodal information fusion strategy and diagnostic criteria to achieve differentiated intelligent diagnosis; 4. Perform multimodal data preprocessing, fusion analysis, and preliminary diagnosis at the edge to reduce communication load and improve system response real-time performance; 5. Use the diagnostic results for status display, fault warning and operation and maintenance decision support, and support data interaction with cloud systems.

[0014] Compared with existing detection methods that rely primarily on single sensing methods or fixed fusion rules, the advantages of this invention are: 1. This invention introduces multiple sensing modes such as visible light, infrared, ultraviolet and acoustic imaging to achieve multi-dimensional comprehensive sensing of the operating status of power equipment, which can effectively make up for the limitations of a single sensing method and improve the accuracy of status assessment and fault diagnosis.

[0015] 2. This invention adopts a multimodal fusion and diagnostic strategy guided by device identification. It dynamically selects fusion rules and diagnostic criteria according to different types of power equipment to achieve differentiated and targeted state perception and diagnosis, thereby enhancing the system's adaptability to multiple devices and multiple operating conditions.

[0016] 3. This invention introduces data preprocessing and edge computing mechanisms into the system architecture. By completing the preliminary processing and analysis of multimodal data on-site, it reduces communication load and improves the real-time response and engineering application feasibility of the system.

[0017] 4. The invention has a clear overall structure, strong engineering feasibility, and is suitable for scenarios with concentrated power equipment such as substations, and has good prospects for promotion and application. Attached Figure Description

[0018] To more clearly illustrate the technical solution of the present invention, the present invention will be further described below with reference to the accompanying drawings. The accompanying drawings are only used to schematically illustrate the embodiments of the present invention and do not constitute a limitation on the scope of protection of the present invention.

[0019] Figure 1 is an overall schematic diagram of the multimodal state perception and intelligent diagnosis system for power equipment in an embodiment of the present invention.

[0020] As shown in the figure, the system includes a multimodal terminal access layer, an edge-side intelligent analysis and lightweight deployment module, a cloud service layer, and an application display layer. The multimodal terminal access layer is used to collect multi-source sensing data such as visible light, infrared, ultraviolet and acoustic imaging. The edge-side module is used to complete device identification, multimodal fusion and diagnostic analysis. The cloud service layer is used for data management, model services and device management. The application display layer is used for status display, report generation and alarm output.

[0021] Figure 2 is a schematic diagram of the operation flow of the multimodal state perception and intelligent diagnosis system in an embodiment of the present invention.

[0022] As shown in the figure, the process includes a perception preparation layer, a perception layer, a cognition layer, and a decision layer. After completing the self-test and calibration of the multimodal sensors, the system synchronously collects visible light, infrared, ultraviolet and acoustic imaging information, and completes intelligent analysis, multimodal fusion and equipment status cognition at the edge. Finally, it outputs defect diagnosis results and auxiliary decision information, and realizes data synchronization and model update with the cloud.

[0023] Figure 3 is a schematic diagram of the hardware structure of the multimodal intelligent detection terminal in an embodiment of the present invention.

[0024] As shown in the figure, the multimodal detection terminal includes a central control unit, an edge computing unit, a multimodal sensor assembly, a communication module, and a power supply system. The multimodal sensor assembly includes visible light, infrared, ultraviolet, and acoustic imaging modules. The edge computing unit is used to execute multimodal perception data processing and intelligent diagnostic algorithms. The communication module is used to realize data interaction with the cloud platform. The power supply system is used to provide stable power support for each module of the terminal. Detailed Implementation

[0025] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings. It should be noted that the following embodiments are only used to explain the technical solutions of the present invention, and are not intended to limit the scope of protection of the present invention.

[0026] Example 1: Structural Embodiment of a Multimodal State Perception and Intelligent Diagnostic System As shown in Figure 1, this embodiment provides a multimodal state perception and intelligent diagnosis system for power equipment. The system includes a multimodal perception unit 1, a data preprocessing unit 2, an equipment identification unit 3, a multimodal fusion diagnosis unit 4, and a result output unit 5.

[0027] The multimodal sensing unit 1 is used to collect multi-source information from the power equipment to be inspected, specifically including a visible light imaging module 11, an infrared thermal imaging module 12, an ultraviolet imaging module 13, and an acoustic imaging module 14. These imaging modules can be integrated on the same detection terminal or formed into a stable relative positional relationship through a fixed installation structure to achieve collaborative sensing of the same power equipment target.

[0028] Among them, the visible light imaging module 11 is used to acquire information such as the appearance, structural outline and nameplate of the power equipment; the infrared thermal imaging module 12 is used to acquire the surface temperature distribution characteristics of the power equipment to identify potential heating anomalies; the ultraviolet imaging module 13 is used to acquire the partial discharge radiation characteristics generated by the power equipment during operation; and the acoustic imaging module 14 is used to acquire the acoustic feature information generated by the power equipment during discharge or abnormal operation.

[0029] Data preprocessing unit 2 is connected to multimodal sensing unit 1 and is used to perform unified preprocessing on the acquired multimodal sensing data. Specifically, data preprocessing unit 2 includes a time synchronization submodule and a spatial registration submodule, which are used to perform timestamp alignment, spatial mapping and data normalization processing on different modal imaging information, thereby forming multimodal sensing data that can be used for joint analysis.

[0030] The equipment identification unit 3 is connected to the data preprocessing unit 2 and is used to automatically identify power equipment based on visible light imaging information. The equipment identification unit 3 can identify the equipment type, structural category, and key parts of the power equipment, providing prior information for subsequent multimodal fusion diagnosis.

[0031] The multimodal fusion diagnostic unit 4 is connected to the equipment identification unit 3 and the data preprocessing unit 2. It is used to select a multimodal fusion and diagnostic strategy that matches the corresponding power equipment based on the identified equipment type information, and to jointly analyze visible light imaging information, infrared thermal imaging information, ultraviolet imaging information and acoustic imaging information to output the status assessment results and / or fault diagnosis results of the power equipment.

[0032] The result output unit 5 is connected to the multimodal fusion diagnostic unit 4 and is used to display, store or transmit the status assessment results and / or fault diagnosis results to realize the visualization and management of the operating status of power equipment.

[0033] Example 2: Multimodal State Perception and Intelligent Diagnosis Process Example As shown in Figure 2, an embodiment of the system of the present invention provides a multimodal state perception and intelligent diagnosis application method for power equipment, which can be operated according to the following process: First, multimodal information is collected from the power equipment to be tested, acquiring multi-source sensing data including at least visible light imaging information, infrared thermal imaging information, ultraviolet imaging information, and acoustic imaging information.

[0034] Subsequently, the collected multimodal sensing data is preprocessed, including time synchronization, spatial alignment and data normalization of different modal sensing data to form a unified multimodal sensing dataset.

[0035] Next, the power equipment is automatically identified based on the visible light imaging information to obtain the equipment type information and / or key structural location information of the power equipment.

[0036] Based on the acquired equipment type information, a multimodal fusion and diagnostic strategy matching the corresponding power equipment is selected to jointly analyze visible light imaging information, infrared thermal imaging information, ultraviolet imaging information, and acoustic imaging information.

[0037] Finally, based on the joint analysis results, the condition assessment results and / or fault diagnosis results of the power equipment are output.

[0038] Example 3: Engineering Application Example under Edge Computing Conditions In a specific engineering application scenario, as shown in Figure 3, the multimodal state perception and intelligent diagnostic system can be deployed on the edge intelligent detection terminal at the substation site. The multimodal perception data undergoes preprocessing, equipment identification, and preliminary state assessment on the detection terminal side, and only the state assessment results, anomaly identification information, or key feature data are uploaded to the remote platform.

[0039] In this implementation, an event-triggered multimodal acquisition strategy can be adopted. When abnormal features are detected in the visible light, infrared or acoustic modes, ultraviolet imaging and multimodal joint analysis are triggered, thereby reducing the overall energy consumption of the system and improving the system operating efficiency while ensuring detection accuracy.

[0040] The above embodiments illustrate specific implementations of the present invention under different system structures and application scenarios. Those skilled in the art can make various changes or equivalent substitutions to the above embodiments without departing from the spirit and substance of the present invention, and all such changes should fall within the protection scope of the present invention.

[0041] This invention is applicable to intelligent inspection and condition monitoring scenarios of power facilities such as substations, power lines, and power distribution equipment. It can be used for condition assessment, defect diagnosis, and fault early warning of various types of power equipment such as transformers, switchgear, and insulators.

[0042] This invention can effectively improve the comprehensiveness and accuracy of power equipment status perception, reduce the risk of misjudgment and omission, and improve the applicability and reliability of intelligent detection systems in complex engineering environments. It has good engineering promotion value and application prospects.

Claims

1. A multimodal state perception and intelligent diagnosis system for power equipment, characterized in that, The system includes: A multimodal sensing unit is used to collect multi-source information from the power equipment under test, wherein the multi-source information includes at least: Visible light imaging information used to acquire information about the appearance and structure of power equipment. Infrared thermal imaging information used to acquire temperature distribution characteristics of power equipment. Ultraviolet imaging information used to acquire partial discharge radiation characteristics of power equipment. And acoustic imaging information used to acquire the acoustic characteristics of electrical equipment discharge; The data preprocessing unit, connected to the multimodal sensing unit, is used to perform time synchronization, spatial alignment, and data normalization processing on the collected visible light imaging information, infrared thermal imaging information, ultraviolet imaging information, and acoustic imaging information to form multimodal sensing data that can be used for joint analysis. The equipment identification unit is connected to the data preprocessing unit and is used to automatically identify the power equipment based on the visible light imaging information, and to obtain the equipment type information and / or key structural location information of the power equipment. The multimodal fusion diagnostic unit is connected to the device identification unit and the data preprocessing unit. It is used to call the multimodal fusion and diagnostic strategy that matches the corresponding power equipment according to the device type information, and to jointly analyze the visible light imaging information, infrared thermal imaging information, ultraviolet imaging information and acoustic imaging information to output the status assessment result and / or fault diagnosis result of the power equipment. The result output unit is connected to the multimodal fusion diagnostic unit and is used to display, store or transmit the state assessment results and / or fault diagnosis results to realize multimodal state perception and intelligent diagnosis of power equipment.

2. The multimodal state perception and intelligent diagnosis system for power equipment according to claim 1, characterized in that, The multimodal sensing unit includes: The system includes a visible light imaging module, an infrared thermal imaging module, an ultraviolet imaging module, and an acoustic imaging module. Each imaging module is either located on the same detection terminal or formed in a relative positional relationship through a fixed installation structure, in order to achieve collaborative perception of the same power equipment target.

3. The multimodal state perception and intelligent diagnosis system for power equipment according to claim 1, characterized in that, The data preprocessing unit includes a time synchronization submodule and a spatial registration submodule. The time synchronization submodule is used to align the timestamps of different modal sensing data, and the spatial registration submodule is used to establish the spatial mapping relationship between the imaging information of each modality, so that the multimodal sensing data can be jointly analyzed in a unified spatial coordinate system.

4. The multimodal state perception and intelligent diagnosis system for power equipment according to claim 1, characterized in that, The device identification unit is configured to perform target detection and classification identification of power equipment based on visible light imaging information, and determine the corresponding device type, device number and / or key structural area according to the identification results, so as to guide the subsequent multimodal fusion and diagnostic process.

5. The multimodal state perception and intelligent diagnosis system for power equipment according to claim 1, characterized in that, The multimodal fusion diagnostic unit is configured to select multimodal data fusion rules and diagnostic criteria that match different power equipment types based on the equipment type information, so as to achieve differentiated status assessment and fault diagnosis for different power equipment.

6. The multimodal state perception and intelligent diagnosis system for power equipment according to claim 1, characterized in that, During the fusion analysis process, the multimodal fusion diagnostic unit comprehensively utilizes the structural features provided by visible light imaging information, the temperature anomaly features provided by infrared thermal imaging information, the partial discharge features provided by ultraviolet imaging information, and the acoustic anomaly features provided by acoustic imaging information to improve the accuracy and reliability of power equipment condition assessment and fault diagnosis.

7. The multimodal state perception and intelligent diagnosis system for power equipment according to claim 1, characterized in that, The system further includes an edge computing unit, which is used to perform preliminary processing and intelligent analysis of multimodal sensing data at the detection site, and send the analysis results or abnormal information to a remote platform to reduce communication load and improve system response efficiency.

Images