An intelligent operation and maintenance management platform and method for smart photovoltaic power stations

Abstract

The present invention relates to the field of photovoltaic power station management technology, and provides an intelligent operation and maintenance management platform and method for smart photovoltaic power stations, wherein the management platform comprises: controlling a designated drone to collect real-time video streams of a target photovoltaic power station area, collecting real-time working data corresponding to each power station equipment in the target photovoltaic power station, analyzing the photovoltaic power generation power fluctuation trend of the target photovoltaic power station based on the real-time working data corresponding to each power station equipment, and displaying it, and diagnosing the fault information corresponding to each power station equipment in the target photovoltaic power station based on the real-time video stream and the real-time working data, and displaying it. The present invention is used to reduce the demand for operation and maintenance personnel and improve the quality and efficiency of inspections in response to the problems of high labor intensity and many subjective factors in the traditional operation and maintenance mode.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power station management, and in particular to an intelligent operation and maintenance management platform and method for a smart photovoltaic power station. Background Art

[0002] Photovoltaic power station is a photovoltaic power generation system that uses solar energy and special materials such as crystalline silicon panels, inverters and other electronic components to connect to the power grid and transmit electricity to the grid. Because its power generation process has no mechanical rotating parts, does not consume fuel, and does not emit any substances including greenhouse gases, it has the characteristics of no noise, no pollution, and no geographical restrictions. Compared with other new power generation technologies, it is a renewable energy power generation technology with ideal characteristics of sustainable development. According to the life cycle predetermined at the time of design, photovoltaic power stations have a profit cycle of 17-20 years. Once a photovoltaic component fails, it means a shortened profit cycle, and if it is not handled for a long time, it may cause failures of surrounding components and even cause safety accidents such as fires. The concentration of existing power stations and the surge in incremental power stations, the efficient operation of power station systems, and the importance of refined and intelligent operation and maintenance in the later stage are self-evident.

[0003] The main modes of photovoltaic operation and maintenance at present include power station trusteeship, contract energy management, and outsourcing of power station operation and maintenance services. The specific contents of operation and maintenance include equipment detection, fault repair, cleaning, safety inspection, etc. The main pain points of various operation and maintenance matters are: (1) complex environment and high maintenance and inspection costs; (2) high work intensity; (3) lack of industry standards and management standards for power station operation indicators; (4) due to labor costs, the cleaning frequency is low, resulting in a decrease in power generation. These problems will cause power stations to have impacts such as reduced power generation, equipment failures, and safety hazards. No effective solutions have been proposed for the problems in related technologies. Summary of the invention

[0004] The present invention provides an intelligent operation and maintenance management platform and method for smart photovoltaic power stations, which are used to address the problems of high labor intensity and many subjective factors in the traditional operation and maintenance mode, reduce the demand for operation and maintenance personnel and improve the quality and efficiency of inspections, provide intelligent monitoring operation and maintenance means and safety guarantees for photovoltaic components, substations, distribution networks, etc., and design an intelligent operation and maintenance system that combines communication, computer, artificial intelligence, big data and other technologies to achieve the goals of intelligent inspection, online monitoring, fault diagnosis and maintenance.

[0005] The present invention provides an intelligent operation and maintenance management platform for a smart photovoltaic power station, comprising:

[0006] The video acquisition module is used to control the designated drone to collect the real-time video stream of the target photovoltaic power station area;

[0007] A data acquisition module, used to collect real-time working data corresponding to each power station equipment in the target photovoltaic power station;

[0008] An operation and maintenance prediction module, used for analyzing the photovoltaic power generation power fluctuation trend of the target photovoltaic power station based on the real-time working data corresponding to each of the power station equipment, and displaying the same;

[0009] The intelligent diagnosis module is used to diagnose and display the fault information corresponding to each power station equipment in the target photovoltaic power station based on the real-time video stream and the real-time working data.

[0010] In one practicable manner,

[0011] The video acquisition module comprises:

[0012] An inspection control unit, used to control a designated UAV to inspect the target photovoltaic power station area according to a preset inspection plan;

[0013] The designated drones include:

[0014] An infrared thermal imager, used to collect biological activity information in the target photovoltaic power station area;

[0015] A high-resolution camera, used to collect real-time image data within the target photovoltaic power station area;

[0016] The video acquisition unit further includes:

[0017] The video collating unit is used to map the biological activity information into the real-time image data to establish a real-time video stream of the target photovoltaic power station area.

[0018] In one practicable manner,

[0019] The data acquisition module comprises:

[0020] The equipment network access unit is used to establish a converged network using ZigBee and NB-IoT, adjust the network structure of the converged network using a hybrid network topology structure, establish a short-distance communication network, and respectively connect each of the power station equipment to the short-distance communication network;

[0021] A data collection unit, used to respectively collect the real-time device data corresponding to each of the power station devices, and divide each of the real-time device data into a plurality of small packet data using a preset GPRS, so as to obtain a small packet data set corresponding to each of the power station devices;

[0022] A data processing unit, used for performing data filtering, data aggregation and data conversion on each of the small packet data sets to obtain real-time working data corresponding to each of the power station equipment;

[0023] A data display unit is used to establish and display a real-time data report of the target photovoltaic power station according to the real-time working data.

[0024] In one practicable manner,

[0025] The operation and maintenance prediction module includes:

[0026] A vibration analysis unit, used to perform variational modal decomposition on each of the real-time working data using a particle optimization algorithm to obtain vibration information and amplitude information corresponding to each of the power station equipment;

[0027] A first model building unit is used to use a preset convolutional neural network to perform convolution analysis on each of the vibration information and each of the amplitude information, obtain the working logic relationship of different power station equipment, and establish an ultra-short-term prediction model of the target photovoltaic power station in combination with a preset gated cycle model;

[0028] The feature capture unit is used to input the vibration information and amplitude information corresponding to each of the power station equipment into the ultra-short-term prediction model, capture the instantaneous change vibration of each of the power station equipment, and establish the instantaneous power generation power of the target photovoltaic power station.

[0029] In one practicable manner,

[0030] The operation and maintenance prediction module further includes:

[0031] A parameter retrieving unit, configured to extract a vibration threshold from the vibration information, extract an amplitude threshold from the amplitude information, and retrieve corresponding model parameters based on the vibration threshold and the amplitude threshold;

[0032] A second model building unit is used to optimize the vibration information and amplitude information by using a particle swarm optimization algorithm, establish a short-term prediction model corresponding to the target photovoltaic power station in combination with a preset LSTM model, and establish a short-term power generation power of the target photovoltaic power station based on the short-term prediction model;

[0033] A trend analysis unit is used to construct a power generation fluctuation direction framework based on the short-term power generation of the target photovoltaic power station, map each instantaneous power generation into the power generation fluctuation direction framework, generate the photovoltaic power generation power fluctuation trend of the target photovoltaic power station, and display it.

[0034] In one practicable manner,

[0035] The intelligent diagnosis module comprises:

[0036] A data preprocessing unit, used to clean, denoise and normalize each of the real-time working data to obtain activation data of each of the power station equipment;

[0037] A feature classification unit, used for classifying the features contained in each of the activation functions to obtain a positive feature sample set and a negative feature sample set, and establishing a total feature set of each of the activation functions;

[0038] A data dimension reduction unit, used to reduce the dimension of each of the activation data using a preset F-Score feature selection algorithm, and obtain a plurality of feature values ​​contained in each of the activation data according to formula (1);

[0039] (1)

[0040] in, represents the i-th eigenvalue in the activation data, represents the average value of the i-th feature sample in the total feature set, represents the average value of the i-th feature sample in the positive feature sample set, represents the average value of the i-th feature sample in the negative feature sample set, represents the number of positive feature samples in the positive feature sample set, Represents the number of negative feature samples in the positive feature sample set, represents the eigenvalue corresponding to the kth positive sample feature on the i-th feature, Indicates the eigenvalue corresponding to the k-th negative sample feature on the i-th feature;

[0041] The abnormality analysis unit is used to optimize each of the characteristic values ​​of the random forest by using the sparrow search algorithm to construct a forest decision tree, retrieve corresponding key parameters when constructing the forest decision tree, adjust the preset SSA-RF classification model according to the key parameters, and use the adjusted preset SSA-RF classification model to perform abnormality detection on each of the characteristic values, obtain the fault information corresponding to each power station equipment in the target photovoltaic power station, and display it.

[0042] In one practicable manner,

[0043] The intelligent diagnosis module further includes:

[0044] A fault monitoring unit is used to construct a hierarchical SoftMax model using deep learning technology, and to input each of the fault information into the hierarchical SoftMax model to locate the fault, and determine the fault location and fault type;

[0045] The fault processing unit is used to establish and display warning suggestions and processing suggestions based on the fault location and fault type.

[0046] The present invention also provides an intelligent operation and maintenance management method for a smart photovoltaic power station, using the above-mentioned intelligent operation and maintenance management platform for a smart photovoltaic power station, comprising the following steps:

[0047] Control the designated drone to collect real-time video streams of the target photovoltaic power station area;

[0048] Collecting real-time working data corresponding to each power station equipment in the target photovoltaic power station;

[0049] Analyzing the photovoltaic power generation power fluctuation trend of the target photovoltaic power station based on the real-time working data corresponding to each of the power station equipment, and displaying the result;

[0050] The fault information corresponding to each power station equipment in the target photovoltaic power station is diagnosed based on the real-time video stream and the real-time working data, and displayed.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] Using drone inspection technology, through aerial perspective and high-precision sensors, we can comprehensively monitor the equipment and environment of photovoltaic power stations, and collect data generated by each power station equipment during operation. In this way, we can continuously monitor the equipment operation data, power station real-time power data, operation and maintenance system operation and status during the operation of the photovoltaic power station. By establishing the power fluctuation trend of the photovoltaic power station, we can preliminarily analyze whether there are any abnormal signs in the power station equipment. Then, we can analyze the faults of the power station equipment based on real-time video streams and real-time working data, and provide operation and maintenance strategies for the intelligent operation and maintenance robot cluster.

[0053] The present invention establishes a data monitoring and intelligent diagnosis system to realize all-weather uninterrupted monitoring of equipment operation data, power station real-time power data, operation and maintenance system operation and maintenance status, and provide operation and maintenance strategies for intelligent operation and maintenance robot clusters; develops a fault monitoring system to locate and accurately identify all equipment faults in photovoltaic power stations in real time, avoid "aimless routine inspections", and guide manual on-site inspections and verifications to eliminate faults; constructs a photovoltaic power prediction system, and realizes full-time period prediction of photovoltaic power generation by designing two prediction models, providing data support for the smoothing of photovoltaic power fluctuations, and solving the problem of large-scale grid connection of photovoltaic power generation.

[0054] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0056] Figure 1 The figure is a schematic diagram of the composition of an intelligent operation and maintenance management platform for a smart photovoltaic power station in an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0058] Example 1

[0059] An intelligent operation and maintenance management platform for smart photovoltaic power stations, such as Figure 1 As shown, including:

[0060] The video acquisition module is used to control the designated drone to collect real-time video streams of the target photovoltaic power station area;

[0061] A data acquisition module, used to collect real-time working data corresponding to each power station equipment in the target photovoltaic power station;

[0062] An operation and maintenance prediction module, used for analyzing the photovoltaic power generation power fluctuation trend of the target photovoltaic power station based on the real-time working data corresponding to each of the power station equipment, and displaying the same;

[0063] The intelligent diagnosis module is used to diagnose and display the fault information corresponding to each power station equipment in the target photovoltaic power station based on the real-time video stream and the real-time working data.

[0064] In this example, the power station equipment includes photovoltaic modules, support structures, cable lines, inverters and other equipment;

[0065] In this example, the fault information includes: the status and operation of equipment such as photovoltaic modules, support structures, cable lines, inverters, and the impact of environmental factors such as vegetation and animal activities on the photovoltaic power station.

[0066] The working principle and beneficial effects of the above technical solution are as follows: using drone inspection technology, through aerial perspective and high-precision sensors, to comprehensively monitor the photovoltaic power station equipment and environment, and to collect data generated by each power station equipment during operation, so as to continuously monitor the equipment operation data, power station real-time power data, operation and maintenance system operation and operation and maintenance status during the operation of the photovoltaic power station. By establishing the power fluctuation trend of the photovoltaic power station, a preliminary analysis is made on whether there are abnormal precursors in the power station equipment, and then the power station equipment faults are analyzed based on real-time video streams and real-time working data, providing operation and maintenance strategies for the intelligent operation and maintenance robot cluster.

[0067] Example 2

[0068] On the basis of Example 1, the intelligent operation and maintenance management platform for a smart photovoltaic power station, the video acquisition module includes:

[0069] An inspection control unit, used to control a designated UAV to inspect the target photovoltaic power station area according to a preset inspection plan;

[0070] The designated drones include:

[0071] An infrared thermal imager, used to collect biological activity information in the target photovoltaic power station area;

[0072] A high-resolution camera, used to collect real-time image data within the target photovoltaic power station area;

[0073] The video acquisition unit further includes:

[0074] The video collating unit is used to map the biological activity information into the real-time image data to establish a real-time video stream of the target photovoltaic power station area.

[0075] In this example, the preset inspection plan refers to the inspection route and frequency set in advance, as well as the inspection contents set by the management personnel before the inspection.

[0076] The working principle and beneficial effects of the above technical solution are as follows: Select a multi-rotor drone with high-altitude flight capability and long flight time, and equip the drone with high-precision sensors such as infrared thermal imagers and high-resolution cameras to collect data in real time, which can not only ensure the timeliness but also the accuracy of the data. In addition, focus on analyzing biological activities and monitor people and animals around the photovoltaic development station.

[0077] Example 3

[0078] On the basis of Example 1, the intelligent operation and maintenance management platform for a smart photovoltaic power station, the data acquisition module includes:

[0079] The equipment network access unit is used to establish a converged network using ZigBee and NB-IoT, adjust the network structure of the converged network using a hybrid network topology structure, establish a short-distance communication network, and respectively connect each of the power station equipment to the short-distance communication network;

[0080] A data collection unit, used to respectively collect the real-time device data corresponding to each of the power station devices, and divide each of the real-time device data into a plurality of small packet data using a preset GPRS, so as to obtain a small packet data set corresponding to each of the power station devices;

[0081] A data processing unit, used for performing data filtering, data aggregation and data conversion on each of the small packet data sets to obtain real-time working data corresponding to each of the power station equipment;

[0082] A data display unit is used to establish and display a real-time data report of the target photovoltaic power station according to the real-time working data.

[0083] In this example, the ZigBee module and the NB-IoT module are connected to the microcontroller through the serial port to achieve information exchange between the two modules and send operation commands to the relay.

[0084] In this example, data transmission between power station equipment and the remote monitoring center can be achieved by sending and receiving small packets of data one by one through the wireless network;

[0085] In this example, the short-distance communication network represents the network that connects the power station equipment to the management platform;

[0086] In this example, the real-time data report includes: real-time monitoring of the operating status, power, voltage, current and other data of equipment such as inverters, combiner boxes, photovoltaic modules, and booster stations, as well as environmental data such as temperature, humidity, wind speed, and light intensity in the photovoltaic power station area, and monitoring of the instant power generation and power consumption of the power station;

[0087] In this example, GPRS representation is preset.

[0088] The working principle and beneficial effects of the above technical solution are as follows: by utilizing ZigBee and NB-IoT in combination with a hybrid network topology to establish a short-distance communication network, and then connecting all power station equipment to the short-distance communication network, the real-time device data is then packet exchanged and transmitted, and the small packet data corresponding to each real-time device data is further filtered, aggregated and converted, and the implementation work data of the power station equipment is obtained, thereby establishing a real-time data report for management personnel to view.

[0089] Example 4

[0090] On the basis of Example 1, the intelligent operation and maintenance management platform for a smart photovoltaic power station, the operation and maintenance prediction module includes:

[0091] A vibration analysis unit, used to perform variational modal decomposition on each of the real-time working data using a particle optimization algorithm to obtain vibration information and amplitude information corresponding to each of the power station equipment;

[0092] A first model building unit is used to use a preset convolutional neural network to perform convolution analysis on each of the vibration information and each of the amplitude information, obtain the working logic relationship of different power station equipment, and establish an ultra-short-term prediction model of the target photovoltaic power station in combination with a preset gated cycle model;

[0093] The feature capture unit is used to input the vibration information and amplitude information corresponding to each of the power station equipment into the ultra-short-term prediction model, capture the instantaneous change vibration of each of the power station equipment, and establish the instantaneous power generation power of the target photovoltaic power station.

[0094] In this example, the prediction duration of the short-term prediction model is 1 minute.

[0095] The working principle and beneficial effects of the above technical solution are as follows: by using the particle optimization algorithm to perform variational modal decomposition on the real-time working data, the vibration information and amplitude information of the power station equipment are determined, and then a convolutional neural network combined with a gated loop model is used to develop an ultra-short-term prediction model, thereby obtaining the instantaneous power generation of the photovoltaic development station and providing an operation and maintenance strategy for the intelligent operation and maintenance robot cluster.

[0096] Example 5

[0097] On the basis of Embodiment 4, the intelligent operation and maintenance management platform for a smart photovoltaic power station, the operation and maintenance prediction module further includes:

[0098] A parameter retrieving unit, configured to extract a vibration threshold from the vibration information, extract an amplitude threshold from the amplitude information, and retrieve corresponding model parameters based on the vibration threshold and the amplitude threshold;

[0099] A second model building unit is used to optimize the vibration information and amplitude information by using a particle swarm optimization algorithm, establish a short-term prediction model corresponding to the target photovoltaic power station in combination with a preset LSTM model, and establish a short-term power generation power of the target photovoltaic power station based on the short-term prediction model;

[0100] A trend analysis unit is used to construct a power generation fluctuation direction framework based on the short-term power generation of the target photovoltaic power station, map each instantaneous power generation into the power generation fluctuation direction framework, generate the photovoltaic power generation power fluctuation trend of the target photovoltaic power station, and display it.

[0101] In this example, the prediction period of the short-term prediction model is 1 hour.

[0102] The working principle and beneficial effects of the above technical solution: By designing two prediction models, the full-time period prediction of photovoltaic power generation is achieved, providing data support for the stabilization of photovoltaic power fluctuations, and solving the problem of large-scale photovoltaic power generation grid connection. The present invention takes the intelligent and efficient operation and maintenance of photovoltaic power stations as the demand, conducts in-depth research on the intelligent operation and maintenance management of photovoltaic power stations, breaks through a number of key technologies to provide operation and maintenance strategies for intelligent operation and maintenance robot clusters, provides fault location and type information for fault elimination, and provides data support for the stabilization of photovoltaic power fluctuations.

[0103] Example 6

[0104] On the basis of Example 1, the intelligent operation and maintenance management platform for a smart photovoltaic power station, the intelligent diagnosis module includes:

[0105] A data preprocessing unit, used to clean, denoise and normalize each of the real-time working data to obtain activation data of each of the power station equipment;

[0106] A feature classification unit, used for classifying the features contained in each of the activation functions to obtain a positive feature sample set and a negative feature sample set, and establishing a total feature set of each of the activation functions;

[0107] A data dimension reduction unit, used to reduce the dimension of each of the activation data using a preset F-Score feature selection algorithm, and obtain a plurality of feature values ​​contained in each of the activation data according to formula (1);

[0108] (1)

[0109] in, represents the i-th eigenvalue in the activation data, represents the average value of the i-th feature sample in the total feature set, represents the average value of the i-th feature sample in the positive feature sample set, represents the average value of the i-th feature sample in the negative feature sample set, represents the number of positive feature samples in the positive feature sample set, Represents the number of negative feature samples in the positive feature sample set, represents the eigenvalue corresponding to the kth positive sample feature on the i-th feature, Indicates the eigenvalue corresponding to the k-th negative sample feature on the i-th feature;

[0110] The abnormality analysis unit is used to optimize each of the characteristic values ​​of the random forest by using the sparrow search algorithm to construct a forest decision tree, retrieve corresponding key parameters when constructing the forest decision tree, adjust the preset SSA-RF classification model according to the key parameters, and use the adjusted preset SSA-RF classification model to perform abnormality detection on each of the characteristic values, obtain the fault information corresponding to each power station equipment in the target photovoltaic power station, and display it.

[0111] The working principle and beneficial effects of the above technical solution are as follows: determine the various subsystems of a photovoltaic power station, such as photovoltaic modules, inverters, grid connections, etc., assign corresponding weights to each subsystem, design a weight tree for each subsystem of the photovoltaic power station, clarify the weights and correlations of each component in each subsystem, combine the weight tree with expert experience, and use the expert system to integrate expert knowledge and experience, converting it into computable rules and reasoning processes to help identify potential failure modes in each subsystem of the photovoltaic power station, and provide targeted suggestions and diagnoses, thereby establishing a photovoltaic power station fault detection algorithm, which not only improves the accuracy of the calculation, but also allows real-time calculations and reduces latency.

[0112] Example 7

[0113] On the basis of Example 6, the intelligent operation and maintenance management platform for a smart photovoltaic power station, the intelligent diagnosis module further includes:

[0114] A fault monitoring unit is used to construct a hierarchical SoftMax model using deep learning technology, and to input each of the fault information into the hierarchical SoftMax model to locate the fault, and determine the fault location and fault type;

[0115] The fault processing unit is used to establish and display warning suggestions and processing suggestions based on the fault location and fault type.

[0116] The working principle and beneficial effects of the above technical solution are: real-time positioning and accurate diagnosis of the cause and location of failure of each subsystem and component in the photovoltaic power station, and formulation of corresponding early warning suggestions and processing suggestions to facilitate management personnel to take timely remedial measures.

[0117] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An intelligent operation and maintenance management platform for smart photovoltaic power stations, characterized in that: include: The video acquisition module is used to control the designated drone to collect real-time video streams of the target photovoltaic power station area; A data acquisition module, used to collect real-time working data corresponding to each power station equipment in the target photovoltaic power station; An operation and maintenance prediction module, used for analyzing the photovoltaic power generation power fluctuation trend of the target photovoltaic power station based on the real-time working data corresponding to each of the power station equipment, and displaying the same; An intelligent diagnosis module, used for diagnosing and displaying the fault information corresponding to each power station equipment in the target photovoltaic power station based on the real-time video stream and the real-time working data; The video acquisition module comprises: An inspection control unit, used to control a designated UAV to inspect the target photovoltaic power station area according to a preset inspection plan; The designated drones include: An infrared thermal imager, used to collect biological activity information in the target photovoltaic power station area; A high-resolution camera, used to collect real-time image data within the target photovoltaic power station area; The video acquisition module further includes: A video collating unit, used for mapping the biological activity information into the real-time image data to establish a real-time video stream of the target photovoltaic power station area; Characterized in that the data acquisition module comprises: The equipment network access unit is used to establish a converged network using ZigBee and NB-IoT, adjust the network structure of the converged network using a hybrid network topology structure, establish a short-distance communication network, and respectively connect each of the power station equipment to the short-distance communication network; A data collection unit, used to respectively collect the real-time device data corresponding to each of the power station devices, and divide each of the real-time device data into a plurality of small packet data using a preset GPRS, so as to obtain a small packet data set corresponding to each of the power station devices; A data processing unit, used for performing data filtering, data aggregation and data conversion on each of the small packet data sets to obtain real-time working data corresponding to each of the power station equipment; A data display unit, used to establish and display a real-time data report of the target photovoltaic power station according to the real-time working data; The operation and maintenance prediction module includes: A vibration analysis unit, used to perform variational modal decomposition on each of the real-time working data using a particle optimization algorithm to obtain vibration information and amplitude information corresponding to each of the power station equipment; A first model building unit is used to use a preset convolutional neural network to perform convolution analysis on each of the vibration information and each of the amplitude information, obtain the working logic relationship of different power station equipment, and establish an ultra-short-term prediction model of the target photovoltaic power station in combination with a preset gated cycle model; A feature capturing unit, used to input the vibration information and amplitude information corresponding to each of the power station equipment into the ultra-short-term prediction model, capture the instantaneous change vibration of each of the power station equipment, and establish the instantaneous power generation power of the target photovoltaic power station; The operation and maintenance prediction module further includes: A parameter retrieving unit, configured to extract a vibration threshold from the vibration information, extract an amplitude threshold from the amplitude information, and retrieve corresponding model parameters based on the vibration threshold and the amplitude threshold; A second model building unit is used to optimize the vibration information and amplitude information by using a particle swarm optimization algorithm, establish a short-term prediction model corresponding to the target photovoltaic power station in combination with a preset LSTM model, and establish a short-term power generation power of the target photovoltaic power station based on the short-term prediction model; A trend analysis unit, used to construct a power generation fluctuation pointing framework based on the short-term power generation of the target photovoltaic power station, map each instantaneous power generation into the power generation fluctuation pointing framework, generate the photovoltaic power generation power fluctuation trend of the target photovoltaic power station, and display it; The intelligent diagnosis module comprises: A data preprocessing unit, used to clean, denoise and normalize each of the real-time working data to obtain activation data of each of the power station equipment; A feature classification unit, used for classifying the features contained in each of the activation data to obtain a positive feature sample set and a negative feature sample set, and establishing a total feature set of each of the activation data; A data dimension reduction unit, used to reduce the dimension of each of the activation data using a preset F-Score feature selection algorithm, and obtain a plurality of feature values ​​contained in each of the activation data according to formula (1); （1） in, represents the i-th eigenvalue in the activation data, represents the average value of the i-th feature sample in the total feature set, represents the average value of the i-th feature sample in the positive feature sample set, represents the average value of the i-th feature sample in the negative feature sample set, represents the number of positive feature samples in the positive feature sample set, Represents the number of negative feature samples in the positive feature sample set, represents the eigenvalue corresponding to the kth positive sample feature on the i-th feature, Indicates the eigenvalue corresponding to the k-th negative sample feature on the i-th feature; An abnormality analysis unit is used to optimize each of the characteristic values ​​of the random forest by using a sparrow search algorithm to construct a forest decision tree, retrieve corresponding key parameters when constructing the forest decision tree, adjust a preset SSA-RF classification model according to the key parameters, and perform abnormality detection on each of the characteristic values ​​by using the adjusted preset SSA-RF classification model to obtain fault information corresponding to each power station equipment in the target photovoltaic power station, and display the fault information; The intelligent diagnosis module also includes: A fault monitoring unit is used to construct a hierarchical SoftMax model using deep learning technology, and to input each of the fault information into the hierarchical SoftMax model to locate the fault, and determine the fault location and fault type; The fault processing unit is used to establish and display warning suggestions and processing suggestions based on the fault location and fault type.

2. An intelligent operation and maintenance management method for a smart photovoltaic power station, using the intelligent operation and maintenance management platform for a smart photovoltaic power station according to claim 1, characterized in that: The steps include: Control the designated drone to collect real-time video streams of the target photovoltaic power station area; Collecting real-time working data corresponding to each power station equipment in the target photovoltaic power station; Analyzing the photovoltaic power generation power fluctuation trend of the target photovoltaic power station based on the real-time working data corresponding to each of the power station equipment, and displaying the result; The fault information corresponding to each power station equipment in the target photovoltaic power station is diagnosed based on the real-time video stream and the real-time working data, and displayed.

Images