Distributed photovoltaic power generation management and control system and method

Abstract

The invention belongs to the technical field of photovoltaic power generation, and particularly provides a distributed photovoltaic power generation management and control system and method, and the method comprises the steps that a data collection module transmits collected data information to a management and control terminal; the management and control terminal processes the received data information, controls the work of the photovoltaic power station based on a processing result, and defines an identification result as a first identification result and transmits the first identification result to the management and control platform when fault information is identified; the infrared image acquisition module acquires image information of photovoltaic arrays in all photovoltaic power stations in the set area to generate an image data set, and transmits the image data set to the management and control platform; and the management and control platform processes the received image data, identifies the fault type, outputs a second identification result, and performs fault early warning based on the first identification result and the second identification result. By combining two recognition results from the data acquisition module and the infrared image acquisition module, false alarm and missing alarm are effectively reduced, and the accuracy of fault recognition is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power generation, and in particular to a distributed photovoltaic power generation control system and method. Background Art

[0002] With the continuous growth of global energy demand and the increasing awareness of environmental protection, distributed photovoltaic power generation, as a clean and renewable energy form, has received widespread attention and promotion. However, the construction and operation of distributed photovoltaic power generation systems also face many challenges, one of which is how to achieve efficient and accurate management and control.

[0003] Early photovoltaic power generation management and control methods mostly relied on manual inspections and manual control, which was not only inefficient but also difficult to achieve real-time monitoring and precise control of the operating status of photovoltaic power stations. In addition, since photovoltaic power stations are usually distributed in different regions and in large numbers, traditional management and control methods also have many difficulties in information collection, processing and transmission.

[0004] In order to overcome these difficulties, people began to explore distributed photovoltaic power generation management and control systems based on the Internet of Things, big data and artificial intelligence technologies. However, the existing distributed photovoltaic power generation management and control systems still have some shortcomings in fault warning and diagnosis. For example, some systems can only rely on a single monitoring method (such as data acquisition module) to identify faults. This method is easily affected by factors such as data noise and transmission delay, resulting in low accuracy and timeliness of fault warning. In addition, some systems lack intelligent means to identify fault types, making it difficult to quickly and accurately identify complex fault types. Summary of the invention

[0005] In view of the above problems existing in the existing distributed photovoltaic power generation management and control, the present invention provides a distributed photovoltaic power generation management and control system and method.

[0006] In a first aspect, the technical solution of the present invention provides a distributed photovoltaic power generation control system, including a control platform and a plurality of photovoltaic power stations distributedly arranged in a set area, the photovoltaic power station is divided into regions, a control terminal is arranged in each region, each control terminal is provided with initial identity information, and the control terminal joins the network of the control platform through the initial identity information; Each photovoltaic power station is provided with a data collection module for collecting the operating status information of the photovoltaic power station. The data collection module is connected to the control terminal in the area and transmits the collected data information to the control terminal; the control terminal processes the received data information and controls the operation of the photovoltaic power station based on the processing result. At the same time, when fault information is identified, the identification result is defined as the first identification result and transmitted to the control platform; The system also includes an infrared image acquisition module, which is used to collect image information of photovoltaic arrays in all photovoltaic power stations in the set area to generate an image data set. The infrared image acquisition module communicates with the management and control platform to transmit the image data set to the management and control platform; the management and control platform processes the received image data to identify the fault type and output a second recognition result, and issues a fault warning based on the first recognition result and the second recognition result.

[0007] As a further limitation of the technical solution of the present invention, the data acquisition module acquires data information of the photovoltaic power station in real time and outputs it to the management and control terminal in the area. The management and control terminal processes and analyzes the acquired data to obtain the current output power of the photovoltaic power station, and determines whether the current output power is within the set pre-output power range. When the current output power is not within the pre-output power range, an adjustment instruction is generated to adjust the output power of the photovoltaic power station.

[0008] As a further limitation of the technical solution of the present invention, the management and control terminal performs abnormal feature extraction based on wavelet packet energy, and extracts a feature vector that can describe the abnormal type from the abnormal data; fault analysis is performed through the extracted abnormal feature vector to determine the corresponding fault state type and determine the location where the fault occurs.

[0009] As a further limitation of the technical solution of the present invention, the control terminal performs j-layer wavelet packet decomposition on the abnormal data, extracts 2 j wavelet packet decomposition coefficients; extract the reconstructed signal of each frequency band as the sub-signal; calculate the energy of each sub-signal frequency range, combine the energy of each wavelet packet sub-signal to generate a feature vector for judging abnormal conditions; use BP neural network for fault detection of photovoltaic power stations; the input of BP neural network is the feature vector generated by wavelet packet energy and the proportion of each phase current amplitude, the number of nodes in the input layer is equal to the dimension of the input feature vector, and the output is the encoding value of each type of fault set. The absolute value of the proportion of the forward current amplitude of each phase current to the peak-to-peak value is the proportion of the current amplitude of each phase.

[0010] As a further limitation of the technical solution of the present invention, the control platform trains U 2 - Net network parameters, generate a photovoltaic array extraction model, import the collected infrared image as input into the photovoltaic array extraction model, extract the photovoltaic array from the image to obtain a segmented photovoltaic component image, use data enhancement technology to expand the scarce sample data, and divide the sample data set according to the fault category; based on the divided sample data set, train the EfficientNet network to obtain a fault recognition model for fault recognition.

[0011] As a further limitation of the technical solution of the present invention, the management and control platform also identifies the fault type and pixel position of the faulty photovoltaic component; if there is a photovoltaic bracket area scattered in two consecutive images, a complete bracket picture is obtained by an image stitching method.

[0012] As a further limitation of the technical solution of the present invention, when the management and control platform performs fault warning based on the first recognition result and the second recognition result, the obtained fault component information and photovoltaic bracket area are matched with the map to obtain location information for fault display.

[0013] As a further limitation of the technical solution of the present invention, the control terminal within the radiation range of the set radius with the center of the entire set area as the circle is the master control terminal, and the remaining control terminals in the set area are slave control terminals; each control terminal is provided with initial identity information, and the initial identity information includes the ID of the control terminal and the master-slave identity tag; When the control terminal joins the network of the control platform through the initial identity information, the master control terminal interacts with the control platform, and each slave control platform interacts with the control platform through a master control terminal.

[0014] In a second aspect, the technical solution of the present invention provides a distributed photovoltaic power generation control method, which is applied to the system described in the first aspect, and the method includes: The data acquisition module collects the operating status information of the photovoltaic power station and transmits the collected data information to the control terminal; The control terminal processes the received data information and controls the operation of the photovoltaic power station based on the processing results. At the same time, when fault information is detected, the identification result is defined as the first identification result and transmitted to the control platform; The infrared image acquisition module collects image information of photovoltaic arrays in all photovoltaic power stations in the set area to generate an image data set, and transmits the image data set to the management and control platform; The management and control platform processes the received image data to identify the fault type and output a second recognition result, and issues a fault warning based on the first recognition result and the second recognition result.

[0015] As a further limitation of the technical solution of the present invention, the data acquisition module acquires data information of the photovoltaic power station in real time and outputs it to the management and control terminal in the area. The management and control terminal processes and analyzes the acquired data to obtain the current output power of the photovoltaic power station, and determines whether the current output power is within the set pre-output power range. When the current output power is not within the pre-output power range, an adjustment instruction is generated to adjust the output power of the photovoltaic power station.

[0016] As a further limitation of the technical solution of the present invention, the management and control terminal performs abnormal feature extraction based on wavelet packet energy, and extracts a feature vector that can describe the abnormal type from the abnormal data; fault analysis is performed through the extracted abnormal feature vector to determine the corresponding fault state type and determine the location where the fault occurs.

[0017] As a further limitation of the technical solution of the present invention, the control terminal performs j-layer wavelet packet decomposition on the abnormal data, extracts 2 j wavelet packet decomposition coefficients; extract the reconstructed signal of each frequency band as the sub-signal; calculate the energy of each sub-signal frequency range, combine the energy of each wavelet packet sub-signal to generate a feature vector for distinguishing abnormal states; use BP neural network for fault detection of photovoltaic power stations; the input of BP neural network is the feature vector generated by wavelet packet energy and the proportion of each phase current amplitude, the number of nodes in the input layer is equal to the dimension of the input feature vector, the output is the code value of each type of fault set, and the absolute value of the proportion of the forward current amplitude of each phase current to the peak-to-peak value is the proportion of the current amplitude of each phase.

[0018] As a further limitation of the technical solution of the present invention, the control platform processes the received image data to identify the fault type and outputs a second recognition result, and the steps of performing fault warning based on the first recognition result and the second recognition result include: The control platform trains U through sample images and annotated target area information 2 - Net network parameters, generate a photovoltaic array extraction model, import the collected infrared image as input into the photovoltaic array extraction model, extract the photovoltaic array from the image to obtain a segmented photovoltaic component image, use data enhancement technology to expand scarce sample data, and divide the sample data set according to the fault category; based on the divided sample data set, train the EfficientNet network to obtain a fault recognition model for fault recognition, and the recognition result is defined as the second recognition result; The management and control platform identifies the fault type and pixel location of the faulty PV module; if there is a PV bracket area scattered in two consecutive images, the complete bracket picture is obtained through image stitching method; when a fault warning is performed based on the first recognition result and the second recognition result, the location information of the obtained faulty component information and PV bracket area is obtained by matching with the map to display the fault.

[0019] As a further limitation of the technical solution of the present invention, the management and control platform also identifies the fault type and pixel position of the faulty photovoltaic component; if there is a photovoltaic bracket area scattered in two consecutive images, a complete bracket picture is obtained by an image stitching method.

[0020] As a further limitation of the technical solution of the present invention, the control terminal within the radiation range of the set radius with the center of the entire set area as the circle is the main control terminal, and the remaining control terminals in the set area are slave control terminals; initial identity information is set in each control terminal, and the initial identity information includes the ID of the control terminal and the master-slave identity tags; when the control terminal joins the network of the control platform through the initial identity information, the main control terminal exchanges information with the control platform, and each slave control platform exchanges information with the control platform through a main control terminal.

[0021] It can be seen from the above technical solutions that the present invention has the following advantages: by collecting the operating status information of the photovoltaic power station in real time through the data acquisition module and transmitting it to the management and control terminal for processing and analysis, the abnormal state of the photovoltaic power station can be discovered in time and early warning can be issued. At the same time, the image information of the photovoltaic array is collected by the infrared image acquisition module, which can further assist in determining the type and location of the fault, thereby improving the accuracy and timeliness of the fault early warning. The management and control terminal extracts abnormal features based on wavelet packet energy and uses BP neural network for fault detection, which can realize rapid and accurate identification of complex fault types.

[0022] This method not only improves the intelligent level of fault detection, but also provides strong support for subsequent fault handling and maintenance. By acquiring data information of the photovoltaic power station in real time through the data acquisition module and adjusting the output power of the photovoltaic power station based on this information, it can ensure that the photovoltaic power station always operates in the best state, thereby improving power generation efficiency and energy utilization. The present invention divides the photovoltaic power station into regions and sets up a management and control terminal for local management, which can reduce the impact of a single fault on the entire system and improve the reliability and stability of the management and control system. At the same time, by joining the network of the management and control platform through the initial identity information, the security and maintainability of the system can be ensured. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0024] Figure 1 is a schematic block diagram of a system according to an embodiment of the present invention.

[0025] Figure 2 is a schematic flow chart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0027] like Figure 1 As shown, an embodiment of the present invention provides a distributed photovoltaic power generation control system, including a control platform and a plurality of photovoltaic power stations distributedly arranged in a set area, the photovoltaic power station is divided into regions, a control terminal is arranged in each region, each control terminal is arranged with initial identity information, and the control terminal joins the network of the control platform through the initial identity information; Each photovoltaic power station is provided with a data collection module for collecting the operating status information of the photovoltaic power station. The data collection module is connected to the control terminal in the area and transmits the collected data information to the control terminal; the control terminal processes the received data information and controls the operation of the photovoltaic power station based on the processing result. At the same time, when fault information is identified, the identification result is defined as the first identification result and transmitted to the control platform; The system also includes an infrared image acquisition module, which is used to collect image information of photovoltaic arrays in all photovoltaic power stations in the set area to generate an image data set. The infrared image acquisition module communicates with the management and control platform to transmit the image data set to the management and control platform; the management and control platform processes the received image data to identify the fault type and output a second recognition result, and issues a fault warning based on the first recognition result and the second recognition result.

[0028] In some embodiments, the data acquisition module acquires data information of the photovoltaic power station in real time and outputs it to the management and control terminal in the area. The management and control terminal processes and analyzes the acquired data to obtain the current output power of the photovoltaic power station, and determines whether the current output power is within the set pre-output power range. When the current output power is not within the pre-output power range, an adjustment instruction is generated to adjust the output power of the photovoltaic power station.

[0029] Here, the setting of the pre-output power range needs to be calculated based on historical data; it should be noted that the data acquisition module also collects environmental information of the photovoltaic power station and sends it to the other management and control terminals. The environmental information includes temperature, light intensity and cloud coverage. The process of setting the pre-output power range of the management and control terminal is as follows: the management and control terminal obtains the current temperature and light intensity, and finds the output power of the time period closest to the current temperature, light intensity and cloud coverage from the stored historical data, which is recorded as the historical output power here. ; Then calculate the current output power based on the current temperature and light intensity collected by the data acquisition module ; Pre-output power range [P A1 , P A2 ], P A1 = -P0,P A2 = +P0, where P0 is the power adjustment constant, which is set according to the historical power changes.

[0030] The specific formula is as follows:

[0031] In the formula, P S It is the rated power of a photovoltaic panel under standard test conditions, which are usually 1000 W / m² of light intensity, 25°C of temperature and AM1.5 of spectrum. T is the current temperature, T S is the temperature under standard test conditions, is the temperature coefficient of the PV panel (usually a negative value, indicating that the output power decreases as the temperature increases). This value can be found in the technical specifications of the PV panel; I S is the light intensity under standard test conditions, I D is the current light intensity (W / m²).

[0032] The control terminal generates adjustment instructions to adjust the output power of the photovoltaic power station, including: when From 0 to P A1 When the power is within the range, a power reduction command is generated to adjust the output power of the photovoltaic power station to ( -P Q , ) range, when In P A2 When the range reaches positive infinity, a power increase instruction is generated to adjust the output power of the photovoltaic power station to ( , +P Q ) range, P Q Set based on historical control data.

[0033] In some embodiments, the management and control terminal extracts abnormal features based on wavelet packet energy, and extracts feature vectors that can describe the abnormal type from the abnormal data; fault analysis is performed through the extracted abnormal feature vectors to determine the corresponding fault state type and determine the location where the fault occurs.

[0034] Specifically, the control terminal performs j-layer wavelet packet decomposition on the abnormal data and extracts 2 j wavelet packet decomposition coefficients; extract the reconstructed signal of each frequency band as the sub-signal; calculate the energy of each sub-signal frequency range, combine the energy of each wavelet packet sub-signal to generate a feature vector for judging abnormal conditions; use BP neural network for fault detection of photovoltaic power stations; the input of BP neural network is the feature vector generated by wavelet packet energy and the proportion of each phase current amplitude, the number of nodes in the input layer is equal to the dimension of the input feature vector, and the output is the encoding value of each type of fault set. The absolute value of the proportion of the forward current amplitude of each phase current to the peak-to-peak value is the proportion of the current amplitude of each phase.

[0035] The specific implementation steps include: Obtain time series data of key parameters such as current and voltage from the monitoring system of the photovoltaic power station. Preprocess the data, including denoising and outlier processing, to ensure the accuracy and consistency of the data. Perform j-layer wavelet packet decomposition on the abnormal data. Select the appropriate wavelet basis function and decomposition layer number j to decompose the signal into 2 j Sub-signals of different frequency bands. Extract the reconstructed signal of each frequency band, i.e., the sub-signal. Calculate the energy of each sub-signal. Combine the energies of each wavelet packet sub-signal into a feature vector for subsequent fault detection. Calculate the absolute value of the proportion of the forward current amplitude of each phase current to the peak-to-peak value as another input feature. Calculate the difference between the maximum and minimum values ​​of each phase current (peak-to-peak value), and then calculate the ratio of the maximum value of each phase current to the peak-to-peak value, and take its absolute value.

[0036] Construct a BP neural network, with the number of input layer nodes equal to the dimension of the feature vector (the sum of the wavelet packet energy characteristics and the current amplitude ratio of each phase). Determine the number of hidden layers and nodes, and generally determine the optimal structure through trial and error. Set the number of output layer nodes to be equal to the number of fault types, and output the encoding value of each type of fault. Use historical fault data (including wavelet packet energy characteristics and current amplitude ratio) as a training set to train the BP neural network. Process the real-time collected data into a feature vector through the above steps. Input the feature vector into the trained BP neural network to obtain the output result. Determine the fault type based on the output result and output the corresponding fault code.

[0037] By extracting signal features of different frequency bands through wavelet packet decomposition and combining the proportion of current amplitude of each phase, the operating status of the photovoltaic power station can be more comprehensively reflected, thereby improving the accuracy and reliability of fault detection. BP neural network can learn complex nonlinear relationships, learn fault features through a large amount of training data, realize effective identification of multiple fault types, and reduce false alarms and missed alarms.

[0038] This method can process real-time collected data, realize online fault detection of photovoltaic power plants, discover and handle faults in a timely manner, and improve the operation efficiency and safety of power plants. Through automated fault detection, the fault point can be quickly located, reducing the time and cost of manual troubleshooting and improving maintenance efficiency. This method has strong adaptability to different photovoltaic power plants and fault types. By adjusting the wavelet packet decomposition parameters and BP neural network structure, it can be applied to photovoltaic power plants of different sizes and types.

[0039] Photovoltaic power station is a photovoltaic power generation system that uses solar energy, special materials (such as crystalline silicon panels) and electronic components such as inverters, and is connected to the power grid and transmits electricity to the grid. The components of a photovoltaic power station mainly include the following aspects: Photovoltaic modules: also known as photovoltaic panels, are the core components of photovoltaic power generation systems. They use the principle of the photoelectric effect to directly convert solar energy into electrical energy. The voltage and current of the photovoltaic module determine the type of inverter, and the length and width determine the number of modules that can be installed on a mounting surface (such as a roof).

[0040] Inverter: An inverter is a device used to convert DC power into stable AC power. In a photovoltaic power generation system, it converts the DC power output by the photovoltaic module into AC power supported by general electrical appliances. At the same time, the inverter also has the function of maximizing the performance of solar cells and system fault protection.

[0041] Bracket: Brackets are mainly used to connect photovoltaic modules to the installation surface (such as the roof) to ensure that photovoltaic modules can be installed stably and safely. The design of the bracket must take into account the load-bearing capacity and wind resistance of the power station.

[0042] Grid connection box: The grid connection box is mainly composed of knife switches, circuit breakers and related control components. As the end of the photovoltaic power generation system, it connects the power grid and the power generation system, protects and measures the total amount of photovoltaic power generation. It facilitates fault inspection and management, and improves the safety and economic benefits of the power generation system.

[0043] Cables: including DC cables and AC cables, used to transmit the power generated by photovoltaic modules and the AC power converted by inverters.

[0044] Battery pack: used to store the electricity obtained by the photovoltaic array under sunlight to ensure the continuity of power supply. The battery pack is an important part of the off-grid photovoltaic power generation system.

[0045] Controller: A photovoltaic controller is an automatic control device used to manage photovoltaic arrays. It can control multiple photovoltaic arrays to charge the battery pack or provide loads to the solar inverter at the same time to maximize energy utilization. In addition, the photovoltaic controller also has a serial communication data transmission function, which can centrally manage and remotely control multiple photovoltaic system substations.

[0046] Specifically, the energy of each sub-signal frequency range is calculated. Taking the three-layer decomposition of wavelet packet as an example, the sub-signal The corresponding energy ; represents the ith sub-signal of the third layer;

[0047] In the formula, Sub-signal The amplitude of the grid-connected current signal is decomposed into 7 layers of wavelet packets using the db3 wavelet basis function. In fact, 128 frequency bands with a width of 50Hz will be extracted. In order to reduce the dimension of the input BP network, the 128 frequency bands are re-integrated and divided into 6 frequency bands according to the characteristic frequency bands of the three abnormal states. In this way, only the energy values ​​on the new 6 frequency bands need to be calculated to generate feature vectors for input into the BP neural network. The parameter training of the BP neural network uses conventional existing methods to determine the structural parameters of the BP neural network. The trained BP neural network is used to identify the source of power quality faults, and effectively determine whether the inverter is in a switch tube disconnection state, a common mode interference state, or a filter resonance effect.

[0048] In some embodiments, the control platform trains the U through sample images and annotated target area information. 2 - Net network parameters, generate a photovoltaic array extraction model, import the collected infrared image as input into the photovoltaic array extraction model, extract the photovoltaic array from the image to obtain the segmented photovoltaic component image, use data enhancement technology to expand the scarce sample data, and divide the sample data set according to the fault category; based on the divided sample data set, train the EfficientNet network to obtain a fault recognition model for fault recognition.

[0049] The management and control platform also identifies the fault type and pixel location of the faulty PV module; if there are PV bracket areas scattered in two consecutive images, the complete bracket picture is obtained through image stitching method.

[0050] When the management and control platform issues a fault warning based on the first recognition result and the second recognition result, the fault component information and photovoltaic bracket area obtained are matched with the map to obtain the location information for fault display.

[0051] In an embodiment of the present invention, the infrared image acquisition module uses a drone with an infrared camera to acquire infrared images. Here, the path planning of the drone for image acquisition and the parameter setting of the image parameters are planned and set using existing methods to ensure that the drone can collect the required infrared image data. The control process of the drone is not the content to be protected by this application and will not be elaborated here.

[0052] The specific implementation steps are as follows: Collect infrared image samples of the photovoltaic array, manually annotate the target area (i.e., photovoltaic modules and fault areas), and generate annotation files. Divide the annotated image samples into training sets, validation sets, and test sets to ensure the diversity and representativeness of the data set. Preprocess the images, including resizing and normalizing them, to meet the input requirements of the U2-Net network.

[0053] When extracting the photovoltaic array, the preprocessed image samples and annotation files are used to train the U2-Net network to generate a photovoltaic array extraction model. The collected infrared image is imported into the photovoltaic array extraction model, and the model is used to segment the image and extract the photovoltaic array area. The segmented photovoltaic component images are further processed, such as noise removal and edge detection, to improve the accuracy of subsequent fault identification. For scarce sample data, data enhancement techniques (such as rotation, scaling, flipping, etc.) are used to expand the data set. The enhanced sample data set is divided according to the fault category (such as hot spots, cracks, circuit breaks, etc.) to ensure that each fault category has enough samples for training.

[0054] The EfficientNet network is trained using the divided sample data set to generate a fault recognition model. The segmented photovoltaic module image is input into the fault recognition model to identify the fault type and locate the pixel position. In addition, for the photovoltaic bracket area scattered in two consecutive images, an image stitching method (such as feature point matching, image fusion, etc.) is used to obtain a complete bracket image. Based on the fault component information and the recognition results of the photovoltaic bracket area, the fault location information is obtained in combination with map matching technology. According to the fault location and fault type, fault warning and display are carried out to provide intuitive fault information and location indication.

[0055] Based on the comprehensive analysis of the two identification results, the management and control platform determines the authenticity and severity of the fault. When a fault is confirmed, the management and control platform immediately triggers the fault warning mechanism and sends warning information to relevant personnel, including the fault type, location, severity, etc. The fault level is pre-set according to the fault type and location. According to the warning information, relevant personnel quickly take countermeasures to troubleshoot and repair the fault.

[0056] By combining the two recognition results from the data acquisition module and the infrared image acquisition module, they can verify and complement each other, effectively reduce false alarms and missed alarms, and improve the accuracy of fault identification. The comprehensive analysis of the two recognition results can more comprehensively reflect the operating status and fault conditions of the photovoltaic power station, and enhance the reliability and credibility of fault warning. Based on accurate fault identification results, operation and maintenance personnel can formulate response measures more quickly and accurately, optimize operation and maintenance decisions, and reduce operation and maintenance costs and time costs. Timely fault warning and disposal can effectively avoid the expansion and deterioration of faults, improve the safety and stability of photovoltaic power stations, and ensure the reliability and quality of power supply.

[0057] In some embodiments, the control terminal within the radiation range of the set radius with the center of the entire set area as the circle is the master control terminal, and the remaining control terminals in the set area are slave control terminals; each control terminal is provided with initial identity information, and the initial identity information includes the ID of the control terminal and the master-slave identity tag; When the control terminal joins the network of the control platform through the initial identity information, the master control terminal interacts with the control platform, and each slave control platform interacts with the control platform through a master control terminal.

[0058] It should be noted here that after the management and control terminal joins the network of the management and control platform through the initial identity information, if there is no information interaction with the management and control platform for an interval greater than the set threshold, identity authentication is required to communicate with the management and control platform again. After each communication is established, the management and control terminal will generate its own identity hash chain based on its own random seed value, and write the setting information of the hash chain that can verify its own identity into the management and control platform.

[0059] During identity authentication, the control terminal to be authenticated sends an authentication request to any control terminal N in the network, and sends it together with the setting information of the current identity hash chain of the control terminal to be authenticated. The control terminal N performs a hash operation on the received hash information and compares it with the data obtained from the control platform. If the two values ​​are the same, the authentication is passed and communication is allowed to continue.

[0060] It should be noted that the control terminal selects a random number and performs multiple hash operations on it to transform an input of any length into an output of a fixed length through hash operations. The output is the hash value, and the results of each hash operation are sequentially combined into a long chain, which is the device's own identity hash chain; Among them, the setting information of the hash chain is the hash value of the Mth hash operation.

[0061] In addition, the control terminal N obtains the setting information of the Hash chain from the control platform according to the ID of the control terminal to be authenticated; If the control terminal N is the main control terminal, the control terminal N communicates with the control platform and directly obtains the data backup query on the control platform through search to obtain the setting information of the Hash chain; If the control terminal N is a slave control terminal, the control terminal N does not interact with the control platform at this time. It initiates a query request to the network, and the request contains the ID of the control terminal to be authenticated; the master control terminal in the network directly obtains the data backup query on the control platform to obtain the setting information of the Hash chain through search, and returns the queried Hash chain setting information to the control terminal N.

[0062] like Figure 2 As shown, an embodiment of the present invention provides a distributed photovoltaic power generation control method, which is applied to the system described in the above embodiment, and the method includes: Step 1: The data acquisition module collects the operating status information of the photovoltaic power station and transmits the collected data information to the control terminal; Step 2: The control terminal processes the received data information, controls the operation of the photovoltaic power station based on the processing result, and when fault information is detected, defines the identification result as the first identification result and transmits it to the control platform; Step 3: The infrared image acquisition module collects image information of photovoltaic arrays in all photovoltaic power stations in the set area to generate an image data set, and transmits the image data set to the management and control platform; Step 4: The management and control platform processes the received image data to identify the fault type and output a second recognition result, and issues a fault warning based on the first recognition result and the second recognition result.

[0063] In some embodiments, the data acquisition module acquires data information of the photovoltaic power station in real time and outputs it to the management and control terminal in the area. The management and control terminal processes and analyzes the acquired data to obtain the current output power of the photovoltaic power station, and determines whether the current output power is within the set pre-output power range. When the current output power is not within the pre-output power range, an adjustment instruction is generated to adjust the output power of the photovoltaic power station.

[0064] In some embodiments, the management and control terminal extracts abnormal features based on wavelet packet energy, and extracts feature vectors that can describe the abnormal type from the abnormal data; fault analysis is performed through the extracted abnormal feature vectors to determine the corresponding fault state type and determine the location where the fault occurs.

[0065] Specifically, the control terminal performs j-layer wavelet packet decomposition on the abnormal data and extracts 2 jwavelet packet decomposition coefficients; extract the reconstructed signal of each frequency band as the sub-signal; calculate the energy of each sub-signal frequency range, combine the energy of each wavelet packet sub-signal to generate a feature vector for distinguishing abnormal states; use BP neural network for fault detection of photovoltaic power stations; the input of BP neural network is the feature vector generated by wavelet packet energy and the proportion of each phase current amplitude, the number of nodes in the input layer is equal to the dimension of the input feature vector, the output is the code value of each type of fault set, and the absolute value of the proportion of the forward current amplitude of each phase current to the peak-to-peak value is the proportion of the current amplitude of each phase.

[0066] In some embodiments, the control platform processes the received image data to identify the fault type and outputs a second recognition result, and the steps of performing a fault warning based on the first recognition result and the second recognition result include: The control platform trains U through sample images and annotated target area information 2 - Net network parameters, generate a photovoltaic array extraction model, import the collected infrared image as input into the photovoltaic array extraction model, extract the photovoltaic array from the image to obtain a segmented photovoltaic component image, use data enhancement technology to expand scarce sample data, and divide the sample data set according to the fault category; based on the divided sample data set, train the EfficientNet network to obtain a fault recognition model for fault recognition, and the recognition result is defined as the second recognition result; The management and control platform identifies the fault type and pixel location of the faulty PV module; if there is a PV bracket area scattered in two consecutive images, the complete bracket picture is obtained through image stitching method; when a fault warning is performed based on the first recognition result and the second recognition result, the location information of the obtained faulty component information and PV bracket area is obtained by matching with the map to display the fault.

[0067] In some embodiments, the control terminal within the radiation range of a set radius with the center of the entire set area as the circle is the master control terminal, and the remaining control terminals within the set area are slave control terminals; initial identity information is set in each control terminal, and the initial identity information includes the ID of the control terminal and the master-slave identity tags; when the control terminal joins the network of the control platform through the initial identity information, the master control terminal exchanges information with the control platform, and each slave control platform exchanges information with the control platform through a master control terminal.

[0068] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0069] An embodiment of a distributed photovoltaic power generation control method provided in an embodiment of the present invention belongs to the same inventive concept as the distributed photovoltaic power generation control system of the above-mentioned embodiments. For details not described in detail in the embodiment of the distributed photovoltaic power generation control method, reference can be made to the embodiment of the above-mentioned distributed photovoltaic power generation control system.

[0070] Although the present invention has been described in detail with reference to the accompanying drawings and in combination with the preferred embodiments, the present invention is not limited thereto. Without departing from the spirit and essence of the present invention, a person of ordinary skill in the art may make various equivalent modifications or substitutions to the embodiments of the present invention, and these modifications or substitutions shall be within the scope of the present invention. Any person of ordinary skill in the art may easily think of changes or substitutions within the technical scope disclosed by the present invention, and these shall be within the scope of protection of the present invention.

Claims

1. A distributed photovoltaic power generation control system, characterized in that: It includes a control platform and several photovoltaic power stations distributed in a set area. The photovoltaic power stations are divided into regions, and a control terminal is set in each region. Each control terminal is set with initial identity information. The control terminal joins the network of the control platform through the initial identity information; Each photovoltaic power station is provided with a data acquisition module for collecting the operating status information of the photovoltaic power station, and the data acquisition module is connected to the control terminal in the area and transmits the collected data information to the control terminal; The control terminal processes the received data information, controls the operation of the photovoltaic power station based on the processing result, and when fault information is detected, defines the identification result as a first identification result and transmits it to the control platform; The system also includes an infrared image acquisition module, which is used to collect image information of photovoltaic arrays in all photovoltaic power stations in the set area to generate an image data set. The infrared image acquisition module communicates with the management and control platform to transmit the image data set to the management and control platform; the management and control platform processes the received image data to identify the fault type and output a second recognition result, and issues a fault warning based on the first recognition result and the second recognition result.

2. The distributed photovoltaic power generation control system according to claim 1, characterized in that: The data acquisition module acquires data information of the photovoltaic power station in real time and outputs it to the management and control terminal in the area. The management and control terminal processes and analyzes the acquired data to obtain the current output power of the photovoltaic power station, and determines whether the current output power is within the set pre-output power range. When the current output power is not within the pre-output power range, an adjustment instruction is generated to adjust the output power of the photovoltaic power station.

3. The distributed photovoltaic power generation control system according to claim 2 is characterized in that: The control terminal extracts abnormal features based on wavelet packet energy and extracts feature vectors that can describe the abnormal type from the abnormal data; fault analysis is performed through the extracted abnormal feature vectors to determine the corresponding fault state type and the location of the fault.

4. The distributed photovoltaic power generation control system according to claim 3 is characterized in that: The control terminal performs j-layer wavelet packet decomposition on the abnormal data and extracts 2 j wavelet packet decomposition coefficients; extract the reconstructed signal of each frequency band as a sub-signal; calculate the energy of each sub-signal frequency range, combine the energy of each wavelet packet sub-signal to generate a feature vector for distinguishing abnormal states; use BP neural network to detect faults in photovoltaic power plants; The input of the BP neural network is the characteristic vector generated by the wavelet packet energy and the proportion of the current amplitude of each phase. The number of nodes in the input layer is equal to the dimension of the input characteristic vector, and the output is the code value of each type of fault set.

5. The distributed photovoltaic power generation control system according to claim 4, characterized in that: The control platform trains U through sample images and annotated target area information 2 - Net network parameters, generate a photovoltaic array extraction model, import the collected infrared image as input into the photovoltaic array extraction model, extract the photovoltaic array from the image to obtain a segmented photovoltaic component image, use data enhancement technology to expand the scarce sample data, and divide the sample data set according to the fault category; based on the divided sample data set, train the EfficientNet network to obtain a fault recognition model for fault recognition.

6. The distributed photovoltaic power generation control system according to claim 5, characterized in that: The management and control platform also identifies the fault type and pixel location of the faulty PV module; if there are PV bracket areas scattered in two consecutive images, the complete bracket picture is obtained through image stitching method.

7. The distributed photovoltaic power generation control system according to claim 6, characterized in that: When the management and control platform issues a fault warning based on the first recognition result and the second recognition result, the fault component information and photovoltaic bracket area obtained are matched with the map to obtain the location information for fault display.

8. The distributed photovoltaic power generation control system according to claim 1, characterized in that: The control terminal within the radiation range of the set radius with the center of the entire set area as the circle is the master control terminal, and the remaining control terminals within the set area are slave control terminals; each control terminal is set with initial identity information, which includes the ID of the control terminal and the master-slave identity tag; When the control terminal joins the network of the control platform through the initial identity information, the master control terminal interacts with the control platform, and each slave control platform interacts with the control platform through a master control terminal.

9. A distributed photovoltaic power generation control method, characterized in that: Applied to the system according to any one of claims 1 to 8, the method comprises: The data acquisition module collects the operating status information of the photovoltaic power station and transmits the collected data information to the control terminal; The control terminal processes the received data information and controls the operation of the photovoltaic power station based on the processing results. At the same time, when fault information is detected, the identification result is defined as the first identification result and transmitted to the control platform; The infrared image acquisition module collects image information of photovoltaic arrays in all photovoltaic power stations in the set area to generate an image data set, and transmits the image data set to the management and control platform; The management and control platform processes the received image data to identify the fault type and output a second recognition result, and issues a fault warning based on the first recognition result and the second recognition result.

10. The distributed photovoltaic power generation control method according to claim 9, characterized in that: The control platform processes the received image data to identify the fault type and output a second recognition result, and performs a fault warning based on the first recognition result and the second recognition result, including: The control platform trains U through sample images and annotated target area information 2 - Net network parameters, generate a photovoltaic array extraction model, import the collected infrared image as input into the photovoltaic array extraction model, extract the photovoltaic array from the image to obtain a segmented photovoltaic component image, use data enhancement technology to expand scarce sample data, and divide the sample data set according to the fault category; based on the divided sample data set, train the EfficientNet network to obtain a fault recognition model for fault recognition, and the recognition result is defined as the second recognition result; The management and control platform identifies the fault type and pixel location of the faulty PV module; if there is a PV bracket area scattered in two consecutive images, the complete bracket picture is obtained through image stitching method; when a fault warning is performed based on the first recognition result and the second recognition result, the location information of the obtained faulty component information and PV bracket area is obtained by matching with the map to display the fault.

Images