Industrial and commercial photovoltaic power station intelligent management system based on Internet of Things

Abstract

The invention provides an industrial and commercial photovoltaic power station intelligent management system based on the Internet of Things, and belongs to the technical field of electric power. The system comprises a data acquisition unit used for acquiring real-time data of a photovoltaic power station; the edge computing node is internally provided with a fault prediction model and an anomaly detection model, and is used for performing fault prediction and anomaly detection on the photovoltaic power station according to the collected real-time data of the photovoltaic power station, and selecting an optimal maintenance scheme according to the results of fault prediction and anomaly detection; and the cloud server is used for constructing and training a fault prediction model based on a deep belief network and constructing an anomaly detection model based on a density spatial clustering algorithm. According to the method, the fault is accurately pre-judged through the fault prediction model, the operation state of the equipment is comprehensively monitored by using the anomaly detection model, and the optimal maintenance scheme can be selected according to the detection result, so that the power generation loss and high maintenance cost caused by sudden faults of the equipment are effectively reduced, and the maintenance benefit of the power station is maximized.

Description

Technical Field

[0001] The present invention relates to the field of power technology, and particularly to an intelligent management system for industrial and commercial photovoltaic power stations based on the Internet of Things. Background Art

[0002] With the continuous growth of the global demand for clean energy, industrial and commercial photovoltaic power stations, as an important distributed energy solution, have been widely applied and developed. In the past few decades, photovoltaic power generation technology has been continuously matured and the cost has been gradually reduced, enabling more and more industrial and commercial enterprises to choose to install photovoltaic power stations to reduce energy costs and achieve the goals of energy conservation and emission reduction. However, with the continuous expansion of the scale and the increasing number of photovoltaic power stations, traditional management methods face many challenges.

[0003] Traditional management methods of photovoltaic power stations have many limitations. Data collection relies on manual inspections or simple sensor networks, with low frequency and poor accuracy. Facing a large amount of data, the processing efficiency is low, making it difficult to meet the requirements of real-time monitoring and management; fault prediction and anomaly detection mainly rely on experience and regular maintenance, lacking scientific models, unable to accurately predict in advance, and difficult to detect subtle abnormal changes; maintenance plans are formulated based on fixed cycles and experience, which are prone to over-maintenance and may also be under-maintained, and the decision-making lacks comprehensive evaluation; monitoring systems are scattered, making it difficult to conduct centralized real-time monitoring, and lacking intelligent interaction means, which affects management efficiency and the normal operation of power stations. Summary of the Invention

[0004] The present invention provides an intelligent management system for industrial and commercial photovoltaic power stations based on the Internet of Things to solve the problems of insufficient fault prediction and anomaly detection capabilities and lack of efficient monitoring and management means in the prior art.

[0005] To achieve the above object, an embodiment of the present invention provides an intelligent management system for industrial and commercial photovoltaic power stations based on the Internet of Things. The intelligent management system for industrial and commercial photovoltaic power stations includes: a data collection unit for collecting real-time data of the photovoltaic power station; an edge computing node with a built-in fault prediction model and an anomaly detection model for performing fault prediction and anomaly detection on the photovoltaic power station according to the collected real-time data of the photovoltaic power station, and selecting an optimal maintenance plan according to the results of the fault prediction and the anomaly detection; and a cloud server for constructing and training the fault prediction model based on a deep belief network and constructing the anomaly detection model based on a density-based spatial clustering algorithm.

[0006] Optionally, the edge computing node is further configured to preprocess the data collected by the data collection unit.

[0007] Optionally, the fault prediction for the photovoltaic power station includes: inputting the data collected by the data acquisition unit into a trained fault prediction model; the fault prediction model calculates the probability of each device in the photovoltaic power station being in a potential fault state through a Softmax classifier; when the probability of a device being in a potential fault state is greater than a preset threshold, the device has a fault risk.

[0008] Optionally, the anomaly detection for the photovoltaic power station includes: constructing a device spatial distribution matrix based on the physical coordinates of the photovoltaic power station devices and the collected data; calculating a dynamic neighborhood radius based on the constructed spatial distribution matrix through the distance and electrical correlation degree between devices; performing spatial clustering on each data point of the devices through the density-based spatial clustering algorithm according to the calculated dynamic neighborhood radius to obtain the density of each data point within the dynamic neighborhood radius; determining the type of each data point according to the obtained density of each data point; judging whether there is an anomaly in the photovoltaic power station devices according to the determined type of each data point, and determining the abnormal devices.

[0009] Optionally, the dynamic calculation of the dynamic neighborhood radius ε i is expressed as: ε i =(1 / N)Σd ij ×(1+C e ·E ij ) where d ij represents the physical distance between devices i and j, E ij represents the electrical correlation coefficient, and Ce represents the electrical weight factor.

[0010] Optionally, the determining the type of each data point according to the obtained density of each data point includes: if there is a data point with a density greater than or equal to a preset minimum number of points, the data point is a core point; if there is a data point that is not a core point and is within the dynamic neighborhood radius of a core point, the data point is a boundary point; if there is a data point that does not belong to a core point or a boundary point, the data point is determined to be a noise point.

[0011] Optionally, the judging whether there is an anomaly in the photovoltaic power station devices according to the determined type of each data point, and determining the abnormal devices includes: if there is a noise point among the data points of the photovoltaic power station devices, the photovoltaic power station devices have an anomaly, and the abnormal devices are determined through the spatial distribution matrix and the location of the noise point; if there is no noise point, the photovoltaic power station devices are operating normally.

[0012] Optionally, the step of selecting an optimal maintenance plan according to the results of the fault prediction and the anomaly detection includes: if a fault is detected in the device, calculating an evaluation value of the impact on the photovoltaic power station for different maintenance plans; calculating a comprehensive score for each maintenance plan according to the evaluation value; selecting the plan with the highest score as the optimal maintenance plan; wherein, the comprehensive score is expressed as:

[0013] wherein, represents different maintenance plans, represents the evaluation value, represents the weight.

[0014] Optionally, the edge computing node connects to the devices in the photovoltaic power station through the Internet of Things communication protocol and uses a preset camera to monitor the environment of the photovoltaic power station in real time.

[0015] Optionally, the intelligent management system further includes a control terminal configured with a display module for displaying the real-time data of the photovoltaic power station, the information of the fault prediction and the anomaly detection. The control terminal is equipped with an intelligent customer service system based on natural language processing technology for providing answers and querying information related to the photovoltaic power station.

[0016] An intelligent management system for industrial and commercial photovoltaic power stations based on the Internet of Things provided by the present invention realizes real-time high-frequency data collection through the Internet of Things technology, gets rid of traditional dependence, preprocesses through edge computing nodes and deeply processes through a cloud server, and analyzes data efficiently and accurately; uses a fault prediction model constructed by a deep belief network and an anomaly detection model constructed by a density-based spatial clustering algorithm to scientifically and accurately predict faults and detect anomalies; formulates personalized maintenance plans according to the detection results, comprehensively evaluates and optimizes maintenance decisions; centrally monitors the power station in real time through the Internet of Things communication protocol, and the control terminal is configured with a display module and an intelligent customer service system, bringing a convenient and intelligent monitoring and management experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained according to these drawings. In the drawings: Figure 1 is a schematic structural diagram of an intelligent management system for industrial and commercial photovoltaic power stations provided by an embodiment of the present invention; Figure 2 is the anomaly detection process provided by an embodiment of the present invention Figure 1 ; Figure 3 is the anomaly detection process provided by the embodiments of the present invention Figure 2 ; Figure 4 is the anomaly detection process provided by the embodiments of the present invention Figure 3 ; Figure 5 is the overall flowchart of the anomaly detection provided by the embodiments of the present invention. Detailed implementation manners

[0018] The following will describe in detail the detailed implementation manners of the embodiments of the present invention with reference to the accompanying drawings. It should be understood that the detailed implementation manners described herein are only used to illustrate and explain the embodiments of the present invention, and are not used to limit the embodiments of the present invention.

[0019] It should be noted that the acquisition, transmission, storage, use, processing, etc. of data in the technical solution of this application all comply with the relevant regulations of national laws and regulations. In the embodiments of this application, some industry-existing solutions such as certain software, components, models, etc. may be mentioned, and they should be regarded as exemplary. The purpose is only to illustrate the feasibility in the implementation of the technical solution of this application, but it does not mean that the applicant has already or necessarily used this solution.

[0020] In the wave of vigorously promoting the clean energy transformation globally, industrial and commercial photovoltaic power stations, as an important pillar in the field of distributed energy, have made great progress in recent years. On the one hand, after years of precipitation, photovoltaic power generation technology has continuously made breakthroughs in aspects such as improving conversion efficiency and enhancing equipment stability, and the cost has dropped significantly. This has enabled many industrial and commercial enterprises to actively participate, reduce electricity costs through photovoltaic power stations, and actively respond to the call for energy conservation and emission reduction. However, with the continuous expansion of the scale and the increasing number of photovoltaic power stations, the disadvantages of the traditional management mode have gradually emerged, making it difficult to meet the needs of modern power station operation and management.

[0021] The present invention is committed to solving many problems in the traditional management mode of industrial and commercial photovoltaic power stations. The present invention collects data in real time by means of the Internet of Things technology, which is efficiently processed by edge computing nodes and cloud servers. Fault prediction and anomaly detection models are respectively constructed using deep belief networks and density-based spatial clustering algorithms. Maintenance plans are scientifically formulated based on the detection results, and centralized monitoring is carried out through Internet of Things communication protocols. With the display module of the control terminal and the intelligent customer service system, the overall operation and management level is improved.

[0022] The following will specifically describe the present invention in combination with Figures 1-5 specifically describe the present invention.

[0023] As Figure 1As shown, an embodiment of the present invention provides an industrial and commercial photovoltaic power station intelligent management system based on the Internet of Things. The industrial and commercial photovoltaic power station intelligent management system includes: a data acquisition unit, which is used to collect real-time data of the photovoltaic power station; an edge computing node, with a built-in fault prediction model and anomaly detection model, which is used to predict faults and detect anomalies in the photovoltaic power station according to the collected real-time data of the photovoltaic power station, and select the optimal maintenance plan according to the results of the fault prediction and anomaly detection; a cloud server, which is used to build and train a fault prediction model based on a deep belief network, and to build an anomaly detection model based on a density-based spatial clustering algorithm.

[0024] The industrial and commercial photovoltaic power station intelligent management system provided by the present invention can collect real-time operating data of each device in the photovoltaic power station through a data acquisition unit, and then transmit the data to the edge computing node. Through the built-in fault prediction model and anomaly detection model of the edge computing node, fault prediction and anomaly detection are performed on the photovoltaic power station, and the optimal maintenance plan is selected according to the results, thereby realizing more intelligent and accurate fault prediction and anomaly detection.

[0025] The cloud server uses a deep belief network to build a fault prediction model, and repeatedly trains the deep belief network, which can accurately predict the probability of each device in the photovoltaic power station being in a potential fault state during future operation. At the same time, the density-based spatial clustering algorithm to build an anomaly detection model can effectively identify abnormal points in the operation data of photovoltaic power station equipment, and then determine whether the equipment has abnormal conditions. The two models built by the cloud server provide powerful intelligent analysis capabilities for the entire intelligent management system. The fault prediction model can warn of potential equipment failures in advance, allowing operation and maintenance personnel to take preventive measures and reduce power generation losses and maintenance costs caused by failures; the anomaly detection model can monitor the operating status of the equipment in real time, detect subtle anomalies in time, prevent the expansion of potential faults, and ensure the stable and efficient operation of the photovoltaic power station.

[0026] Preferably, the edge computing node is also used to pre-process the data collected by the data collection unit.

[0027] In the preferred embodiment of the present invention, the data collected by the data collection unit may have many problems, such as low data frequency and poor accuracy, and the preprocessing function of the edge computing node can correct these defects in time. It can clean the data, remove errors or invalid data caused by equipment failure, environmental interference, etc., and improve data quality. Through data aggregation, a large amount of scattered raw data is integrated according to specific rules, reducing the amount of data transmission and alleviating network pressure.

[0028] Preferably, fault prediction for a photovoltaic power station includes: inputting the data collected by the data acquisition unit into a trained fault prediction model; the fault prediction model calculates the probability of each device in the photovoltaic power station being in a potential fault state through a Softmax classifier; when the probability of a device being in a potential fault state is greater than a preset threshold, the device has a fault risk.

[0029] In a preferred embodiment of the present invention, the Softmax classifier is essentially a multi-class logistic regression model, and its core mission is to convert the raw numerical values output by the fault prediction model into the probability values of each device being in different states (normal, potential fault, etc.). In the complex operating environment of a photovoltaic power station, the device state is not a simple binary determination of "yes" or "no", but there are multiple gradual states. The Softmax classifier performs unique mathematical operations, that is, exponentiating the input data and normalizing it between 0 and 1, so that the sum of the probabilities of all possible states is 1.

[0030] For example, if the preset threshold is 0.6, for a certain inverter, the fault prediction model accurately calculates that the probability value of the inverter being in a potential fault state is 0.7. Then, at this time, the probability of the inverter being in a potential fault state is greater than the preset threshold, and it is determined that the inverter is about to fail. The system will keenly issue a warning signal, clearly informing the operation and maintenance personnel that the device has a fault risk.

[0031] Preferably, the steps of selecting the optimal maintenance plan according to the results of fault prediction and anomaly detection include: if it is detected that a device has a fault, for different maintenance plans, calculate the evaluation value of the impact on the photovoltaic power station; according to the evaluation value, calculate the comprehensive score of each maintenance plan; select the plan with the highest score as the optimal maintenance plan; where the comprehensive score is expressed as: (1) Where represents different maintenance plans, represents the evaluation value, represents the weight.

[0032] In a preferred embodiment of the present invention, if it is predicted through fault prediction and anomaly detection that there is a device that has failed or is about to fail, an optimal maintenance plan will be selected from three plans (immediate repair, deferred repair, replace the device). First, different maintenance plans are respectively evaluated for their performance under factors such as device fault risk, maintenance cost, and downtime loss, and the corresponding evaluation values are obtained . For example, for the immediate repair plan, its score for reducing the device fault risk may be , its score for maintenance cost may be ,and its score for reducing downtime loss may be Then, based on the evaluation value, the comprehensive score of each maintenance plan is calculated, such as the comprehensive score of the immediate repair plan. By comparing the comprehensive scores of different maintenance plans, the plan with the highest score is the optimal maintenance plan recommended by the system. For example, if the immediate repair plan has the highest comprehensive score, the system will recommend immediate repair to the operation and maintenance personnel, and provide detailed maintenance procedures, suggestions for the allocation of required resources, etc., to assist the operation and maintenance personnel in efficiently carrying out equipment maintenance work and ensuring the stable operation of the photovoltaic power station.

[0033] Preferably, the edge computing node connects to the equipment in the photovoltaic power station through the Internet of Things communication protocol, and uses a preset camera to monitor the photovoltaic power station environment in real time.

[0034] In a preferred embodiment of the present invention, the edge computing node uses the Internet of Things communication protocol to achieve seamless connection with various devices in the photovoltaic power station. Whether it is a photovoltaic panel, an inverter, or other auxiliary equipment, it can use this communication protocol to perform stable data interaction with the edge computing node. The edge computing node is able to obtain the operating parameters of the equipment in real time, such as the output power of the photovoltaic panel, the operating frequency of the inverter, etc., to provide first-hand data support for subsequent fault prediction, anomaly detection and other work. At the same time, the edge computing node also integrates preset camera resources to carry out all-round real-time monitoring of the photovoltaic power station environment. The camera captures the changes in light intensity and weather conditions around the power station in real time, including whether there is cloud cover, whether there is rain, strong wind and other severe weather. These environmental data are crucial for evaluating the power generation efficiency of the photovoltaic power station and the stability of equipment operation. For example, insufficient light intensity may affect the power generation of the photovoltaic panel, while extreme weather such as strong wind and heavy rain may cause physical damage to the equipment. Through real-time collection and analysis of these environmental data, edge computing nodes can, on the one hand, assist in determining whether the equipment is operating normally; on the other hand, they can also provide more comprehensive input information for fault prediction models and anomaly detection models, thereby improving the accuracy and reliability of the entire intelligent management system's judgment of the operating status of photovoltaic power stations and ensuring efficient and stable operation of power stations.

[0035] Preferably, the intelligent management system also includes a control terminal equipped with a display module for displaying real-time data of the photovoltaic power station, fault prediction and anomaly detection information. The control terminal is equipped with an intelligent customer service system based on natural language processing technology for providing answers and querying relevant information of the photovoltaic power station.

[0036] In a preferred embodiment of the present invention, the control terminal is equipped with a display module, which clearly presents the real-time data of the photovoltaic power station, covering the output power of photovoltaic panels, the operating status of inverters, equipment operation parameters, etc. At the same time, it also displays the fault prediction and anomaly detection information in real time, enabling the management personnel to quickly know the potential fault risks of the power station equipment and the abnormal operating conditions, providing a strong basis for taking timely countermeasures. In addition, the control terminal is equipped with an intelligent customer service system based on natural language processing technology. The operator can easily query the relevant information of the photovoltaic power station, such as the historical operation data of equipment, maintenance records, etc., just by asking questions in natural language. When encountering complex problems, the intelligent customer service system can quickly give professional answers by virtue of its ability to understand and analyze natural language, without the operator spending a lot of time looking up information or consulting experts. This not only significantly improves the efficiency of obtaining information, reduces the communication cost, but also further enhances the usability of the entire intelligent management system, enabling the operator to manage the photovoltaic power station more efficiently and ensure the stable operation of the power station.

[0037] In addition, the present invention also integrates an electronic invoice system to achieve an automatic settlement function, simplify the electricity bill settlement process, and improve efficiency. It provides one-stop customer service, including bill notification, electricity bill recovery management, etc., to enhance customer satisfaction.

[0038] Please refer to Figures 2-5 , when performing anomaly detection on the equipment of the photovoltaic power station through the anomaly detection model, it is roughly divided into three stages.

[0039] Preferably, as Figure 2 shown, the anomaly detection of the photovoltaic power station includes: constructing an equipment spatial distribution matrix according to the physical coordinates of the photovoltaic power station equipment and the collected data; calculating the dynamic neighborhood radius according to the constructed spatial distribution matrix through the distance and electrical correlation degree between equipment; performing spatial clustering on each data point of the equipment through the density-based spatial clustering algorithm according to the calculated dynamic neighborhood radius to obtain the density of each data point within the dynamic neighborhood radius; determining the type of each data point according to the obtained density of each data point; judging whether there is an anomaly in the photovoltaic power station equipment according to the determined type of each data point, and determining the abnormal equipment. The dynamic calculation of the dynamic neighborhood radius ε i is expressed as: ε i =(1 / N)Σd ij ×(1+C e ·E ij ) (2) where d ij represents the physical distance between equipment i and j, E ij represents the electrical correlation coefficient, and Ce represents the electrical weight factor.

[0040] In the intelligent management system of industrial and commercial photovoltaic power stations, anomaly detection is an important link to ensure the stable operation of the power station. First, based on the physical coordinates of the photovoltaic power station equipment and the collected data, an equipment spatial distribution matrix is constructed, and a position and data relationship diagram of all equipment in the power station is established. The physical coordinates of the equipment clarify the position of the equipment in the actual space. Subsequently, according to the constructed spatial distribution matrix, the dynamic neighborhood radius is calculated by Equation (1). The physical distance between equipment reflects the degree of proximity of equipment in space, while the electrical correlation coefficient reflects the degree of closeness of equipment in electrical connection and function. The calculated dynamic neighborhood radius can more accurately reflect the mutual relationship of equipment in actual operation because the physical distance and electrical correlation degree between different equipment vary dynamically under different working conditions. After obtaining the dynamic neighborhood radius, the density-based spatial clustering algorithm (DBSCAN) is used to perform spatial clustering on each data point of the equipment and determine the type of each data point. According to the type of each determined data point, it is judged whether there is an anomaly in the photovoltaic power station equipment. Through this series of rigorous and scientific anomaly detection processes, the intelligent management system can timely and accurately detect anomalies in the operation of photovoltaic power station equipment, provide strong support for maintenance personnel to take targeted measures, ensure the continuous and stable operation of the photovoltaic power station, and improve power generation efficiency and economic benefits.

[0041] As Figure 3 and Figure 4 shown, further preferably, according to the density of each obtained data point, determining the type of each data point includes: if there is a data point with a density greater than or equal to the preset minimum number of points, then this data point is a core point; if there is a data point that is not a core point and is within the dynamic neighborhood radius of the core point, then this data point is a boundary point; if there is a data point that does not belong to the core point and the boundary point, then this data point is determined to be a noise point. According to the type of each determined data point, it is judged whether there is an anomaly in the photovoltaic power station equipment, and the abnormal equipment is determined as follows: if there is a noise point among the data points of the photovoltaic power station equipment, then the photovoltaic power station equipment has an anomaly, and the abnormal equipment is determined through the spatial distribution matrix and the position where the noise point is located; if there is no noise point, then the photovoltaic power station equipment operates normally.

[0042] In the process of determining the type of data points above, calculate the number of data points within the dynamic neighborhood radius of each data point to obtain the density of each data point within the dynamic neighborhood radius. If a data point contains a sufficient number of other points within its neighborhood (i.e., the density is greater than or equal to the preset minimum number of points), then the data point is determined to be a core point; if a data point itself is not a core point but is within the dynamic neighborhood radius of a certain core point, then it is identified as a boundary point; if a data point belongs to neither a core point nor a boundary point, it is determined to be a noise point. Since the physical coordinates and operating parameter information of each device are detailed recorded when constructing the device spatial distribution matrix. After determining that a data point is a noise point, by querying the spatial distribution matrix, the location of the abnormal device in the power station can be accurately located.

[0043] For example, assume that the preset minimum number of points is 3. If there are 4 other data points within the dynamic neighborhood radius of 1.5 of device A, then the current data point of device A is determined to be a core point. If the number of data points around the data point of device D itself is less than 3 and it is not a core point, but it is within the dynamic neighborhood radius of 1.5 of device A (a core point), then the data point of device D is identified as a boundary point. If the number of data points around the data point of device E is small and it is not within the dynamic neighborhood radius of any core point, then device E is determined to be a noise point. Since the data point of device E is determined to be a noise point, this indicates that the operating state of device E has deviated from the normal range and an abnormal situation has occurred. By tracing back the previously constructed device spatial distribution matrix, it can be clearly determined that device E is an abnormal device, and the operation and maintenance personnel can check and repair the abnormal situation of device E to ensure the stable operation of the photovoltaic power station. As Figure 5 shown, the method for anomaly detection provided in the present invention constructs a spatial distribution matrix by integrating the physical coordinates and operating parameters of the devices in the photovoltaic power station, calculates the dynamic neighborhood radius by comprehensively considering the physical distance and electrical correlation degree between devices, and uses a density-based spatial clustering algorithm to accurately determine core points, boundary points, and noise points based on the density and location of data points, thereby greatly improving the detection accuracy and avoiding misjudgment and missed judgment; once a noise point is detected and an anomaly is determined, the abnormal device can be quickly and accurately located by means of the spatial distribution matrix, allowing the operation and maintenance personnel to repair before the fault deteriorates, preventing small faults from evolving into serious faults, ensuring the coordinated and stable operation of each device in the photovoltaic power station, improving the power generation efficiency, reducing power generation losses and economic losses, and ensuring the continuous provision of clean energy for industrial and commercial users.

[0044] In summary, the intelligent management system for industrial and commercial photovoltaic power stations based on the Internet of Things provided by the present invention realizes real-time, comprehensive, and high-frequency data collection by means of Internet of Things technology, covering various key operation data of photovoltaic power station equipment, and changes the disadvantages of traditional data collection; it uses a combination of edge computing and cloud computing to efficiently process data and quickly mine key information. In terms of fault prediction and anomaly detection, the model constructed by the deep belief network accurately predicts faults and issues early warnings several days or even weeks in advance, and the density-based spatial clustering algorithm model comprehensively detects subtle anomalies. When formulating maintenance plans, they are customized individually according to the detection results to avoid over-maintenance or insufficient maintenance, and different plans are evaluated from multiple dimensions to select the optimal one to maximize maintenance benefits. In terms of monitoring and interactive management, it can centrally monitor the power station environment and equipment operation in real time, and the control terminal displays key information through the display module, and cooperates with the intelligent customer service system to improve management efficiency and user experience through natural language interaction.

[0045] It should be understood that in various embodiments of the present invention, the magnitudes of the serial numbers of the above processes do not mean the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.

[0046] In addition, the terms "system" and "network" are often used interchangeably in this article. The term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this article generally represents an "or" relationship between the associated objects before and after.

[0047] It should be understood that in the embodiments of the present invention, "B corresponding to A" means that B is associated with A, and B can be determined according to A. However, it should also be understood that determining B according to A does not mean determining B only according to A, and B can also be determined according to A and / or other information.

[0048] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0049] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0050] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection can be an indirect coupling or communication connection through some interfaces, devices, or units, or can also be an electrical, mechanical, or other form of connection.

[0051] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or can also be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiments of the present invention.

[0052] In addition, the functional units in various embodiments of the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0053] Through the description of the above embodiments, those skilled in the art can clearly understand that the present invention can be implemented by hardware, or firmware, or a combination thereof. When implemented using software, the above functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. The computer-readable medium includes computer storage media and communication media, where the communication media includes any medium that facilitates the transfer of a computer program from one place to another. The storage media can be any available medium that can be accessed by a computer. By way of example but not limitation: the computer-readable medium can include RAM, ROM, EEPROM, CD-ROM or other optical disc storage, magnetic disk storage media or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer. In addition, any connection can suitably be a computer-readable medium. For example, if the software is transmitted using coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL) or wireless technologies such as infrared, radio, and microwave from a website, server, or other remote source, then the coaxial cable, fiber optic cable, twisted pair, DSL or wireless technologies such as infrared, wireless, and microwave are included in the definition of the medium. As used in the present invention, disk and disc include compact disc (CD), laser disc, optical disc, digital versatile disc (DVD), floppy disk, and Blu-ray disc, where disks generally reproduce data magnetically, while discs reproduce data optically with a laser. The above combinations should also be included within the scope of protection of the computer-readable medium.

[0054] In summary, the above description is only a preferred embodiment of the technical solution of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An intelligent management system for industrial and commercial photovoltaic power stations based on the Internet of Things, characterized in that: The industrial and commercial photovoltaic power station intelligent management system includes: A data acquisition unit, used to collect real-time data of the photovoltaic power station; The edge computing node has a built-in fault prediction model and anomaly detection model, which is used to perform fault prediction and anomaly detection on the photovoltaic power station based on the collected real-time data of the photovoltaic power station, and select the optimal maintenance plan based on the results of the fault prediction and the anomaly detection; The cloud server is used to construct and train the fault prediction model based on a deep belief network and to construct the anomaly detection model based on a density-based spatial clustering algorithm.

2. The intelligent management system for industrial and commercial photovoltaic power stations according to claim 1, characterized in that: The edge computing node is also used to pre-process the data collected by the data collection unit.

3. The intelligent management system for industrial and commercial photovoltaic power stations according to claim 1, characterized in that: The fault prediction of the photovoltaic power station includes: Inputting the data collected by the data collection unit into a trained fault prediction model; The fault prediction model calculates the probability of each device in the photovoltaic power station being in a potential fault state through a Softmax classifier; When the probability that a device is in a potential failure state is greater than a preset threshold, the device is at risk of failure.

4. The intelligent management system for industrial and commercial photovoltaic power stations according to claim 1, characterized in that: The abnormality detection of the photovoltaic power station includes: Construct the equipment spatial distribution matrix based on the physical coordinates of the photovoltaic power station equipment and the collected data; According to the constructed spatial distribution matrix, the dynamic neighborhood radius is calculated by the distance and electrical correlation between the devices; According to the calculated dynamic neighborhood radius, spatially clustering each data point of the device by using the density-based spatial clustering algorithm to obtain the density of each data point within the dynamic neighborhood radius; Determine the type of each data point according to the obtained density of each data point; According to the determined type of each data point, it is determined whether there is any abnormality in the photovoltaic power station equipment, and the abnormal equipment is determined.

5. The intelligent management system for industrial and commercial photovoltaic power stations according to claim 4 is characterized in that: The dynamic neighborhood radius ε i The dynamic calculation of is expressed as: e i =(1 / N)Σd ij ×(1+C e ·E ij ) Among them, d ij It is represented by the physical distance between devices i and j, E ij It is expressed as the electrical correlation coefficient, and Ce is expressed as the electrical weight factor.

6. The intelligent management system for industrial and commercial photovoltaic power stations according to claim 4, characterized in that: Determining the type of each data point according to the obtained density of each data point includes: If there is a data point whose density is greater than or equal to the preset minimum number of points, then the data point is a core point; If there is a data point that is not a core point and is within the dynamic neighborhood radius of the core point, then the data point is a boundary point; If there is a data point that is neither a core point nor a boundary point, the data point is determined to be a noise point.

7. The intelligent management system for industrial and commercial photovoltaic power stations according to claim 6, characterized in that: The step of judging whether there is an abnormality in the photovoltaic power station equipment according to the determined type of each data point, and determining the abnormal equipment includes: If there are noise points in the data points of the photovoltaic power station equipment, the photovoltaic power station equipment is abnormal, and the abnormal equipment is determined through the spatial distribution matrix and the location of the noise points; If there is no noise point, the photovoltaic power station equipment operates normally.

8. The intelligent management system for industrial and commercial photovoltaic power stations according to claim 1, characterized in that: The step of selecting the optimal maintenance plan according to the results of the fault prediction and the abnormality detection includes: If a device failure is detected, Calculate the impact assessment value on the photovoltaic power station for different maintenance plans; Calculate a comprehensive score for each maintenance plan based on the evaluation value; The solution with the highest score is selected as the most optimal maintenance solution; Among them, the comprehensive score It is expressed as: in, Indicates different maintenance options, represents the evaluation value, Represents weight.

9. The intelligent management system for industrial and commercial photovoltaic power stations according to claim 1, characterized in that: The edge computing node is connected to the equipment in the photovoltaic power station through the Internet of Things communication protocol, and uses a preset camera to monitor the photovoltaic power station environment in real time.

10. The intelligent management system for industrial and commercial photovoltaic power stations according to claim 1, characterized in that: The intelligent management system also includes a control terminal configured with a display module for displaying real-time data of the photovoltaic power station, the fault prediction and the abnormality detection information, The control terminal is equipped with an intelligent customer service system based on natural language processing technology, which is used to provide answers and query relevant information of the photovoltaic power station.

Images