Rural photovoltaic grid-connected management method and system

Abstract

The invention discloses a rural photovoltaic grid-connected management method and system, relates to the technical field of energy management, and solves the problem that a harmonic change rate pollution value, an energy loss rate and an environmental influence coefficient are difficult to analyze according to operation data and environmental data. Secondly, potential faults of the photovoltaic power station are difficult to identify and analyze according to video images; thirdly, fault coefficients are analyzed according to the recognition result, and the inspection frequency is planned; and finally, comprehensive analysis is difficult to carry out according to various data, and the inspection frequency is difficult to dynamically adjust after threshold evaluation operation states are divided according to analysis results. According to the invention, real-time fault identification is realized by analyzing operation data and environment data and by means of a deep learning model, and the inspection frequency is planned according to a fault coefficient; the operation condition of the photovoltaic power station is analyzed in multiple dimensions, the operation level is dynamically divided, the inspection frequency is dynamically adjusted accordingly, and the inspection efficiency is improved.

Description

Technical Field

[0001] The present invention belongs to the field of energy management, and in particular relates to a rural photovoltaic grid-connected management method and system. Background Art

[0002] With the development of rural economy and the improvement of living standards, the demand for electricity in rural areas continues to grow. Photovoltaic power generation is clean, renewable and pollution-free, and is an important direction for future energy development. However, the operation and maintenance of rural photovoltaic power stations face many challenges, which seriously affect the power generation efficiency and service life of photovoltaic power stations. Manual inspection is not only time-consuming and labor-intensive, but also difficult to cover all equipment, which increases additional operating expenses. By combining advanced technologies such as big data and machine learning models, intelligent inspection management can be carried out.

[0003] The existing problems are as follows: first, it is difficult to analyze the harmonic rate pollution value, energy loss rate, and environmental impact coefficient based on operating data and environmental data; second, it is difficult to analyze the potential faults of photovoltaic power stations based on video image recognition; then, it is difficult to analyze the fault coefficient and plan the inspection frequency based on the recognition results; finally, it is difficult to conduct a comprehensive analysis based on various data, and it is difficult to divide the threshold value according to the analysis results to evaluate the operating status and dynamically adjust the inspection frequency. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes a rural photovoltaic grid-connected management method and system for solving the following technical problems:

[0005] First, it is difficult to analyze the harmonic rate pollution value, energy loss rate, and environmental impact coefficient based on operating data and environmental data; second, it is difficult to analyze the potential faults of photovoltaic power stations based on video image recognition; then, it is difficult to analyze the fault coefficient and plan the inspection frequency based on the recognition results; finally, it is difficult to conduct a comprehensive analysis based on various data, and it is difficult to divide the threshold value according to the analysis results and evaluate the operating status and dynamically adjust the inspection frequency.

[0006] To solve the above problems, the first aspect of the present invention provides a rural photovoltaic grid-connected management method, comprising the following steps:

[0007] S1: Collect the operation data and environmental data of the photovoltaic power station; monitor the operation status of the photovoltaic power station in real time and collect monitoring video images of the photovoltaic power station;

[0008] S2: preprocessing the collected operation data and environmental data including: cleaning and standardization; preprocessing the collected video images including: denoising and image enhancement;

[0009] S3: Analyze and calculate the harmonic rate pollution value and energy loss rate based on the collected and processed operating data; analyze and calculate the environmental impact coefficient based on the collected and processed environmental data; analyze and analyze the potential faults of the photovoltaic power station based on the collected and processed video images; analyze and calculate the fault coefficient based on the recognition results and plan the inspection frequency;

[0010] S4: Calculate the comprehensive evaluation index based on the harmonic rate pollution value, energy loss rate, environmental impact coefficient and fault coefficient, divide the threshold value into evaluation values ​​according to the analysis results, and dynamically adjust the inspection frequency.

[0011] As a further solution of the present invention: the operation data and environmental data of the photovoltaic power station collected in step S1 includes the following steps:

[0012] Operation data include: current, voltage, grid-connected power generation, photovoltaic power generation, total area of ​​photovoltaic panels, inverter housing area, total length of cables and wiring, and operating time;

[0013] Environmental data include: ambient temperature, ambient humidity, air pressure and wind speed.

[0014] As a further solution of the present invention: the step S3 of analyzing and calculating the harmonic rate pollution value and the energy loss rate according to the collected and processed operation data includes the following steps:

[0015] Perform frequency domain analysis on the collected and processed voltage and current signals according to the fast Fourier transform algorithm FFT; convert the time domain signal into a frequency domain signal; draw a spectrum diagram according to the conversion result, where the horizontal axis of the spectrum diagram represents the frequency and the vertical axis of the spectrum diagram represents the signal strength or power; identify and extract the fundamental wave and harmonic wave according to the spectrum diagram;

[0016] By analyzing the formula:

[0017] Get the effective value of current I e ; Where i(n) represents the nth sampling value of the current, and N represents the total number of sampling points;

[0018] By analyzing the formula:

[0019] Get the voltage effective value U e ; Wherein, U(n) represents the nth sampling value of the voltage, and N represents the total number of sampling points;

[0020] By analyzing the formula:

[0021]

[0022] The current harmonic distortion rate THD I and voltage harmonic distortion rate THDU are obtained; where I1 Indicates the effective value of fundamental current, U 1 Indicates the effective value of the fundamental voltage, I h Indicates the effective value of the hth harmonic current, U h It represents the effective value of the hth harmonic voltage, where H is the total number of harmonics;

[0023] By analyzing the formula:

[0024]

[0025] Get the harmonic rate pollution value R I,U Among them, B I Indicates the standard limit of current harmonic distortion rate, B U Indicates the standard limit of voltage harmonic distortion rate;

[0026] By analyzing the formula:

[0027] The energy loss rate λ is obtained; where Q represents photovoltaic power generation and L represents grid-connected power generation.

[0028] As a further solution of the present invention: the step S3 of analyzing and calculating the environmental impact coefficient according to the collected and processed environmental data includes the following steps:

[0029] By analyzing the formula: C = (t * ε 1 +w*ε 2 +v*ε 3 +f*ε 4 )×100%

[0030] Get the environmental impact coefficient C; where t represents the ambient temperature, w represents the ambient humidity, v represents the wind speed, and f represents the air pressure; ε 1 , ε 2 , ε 3 and ε 4 They represent the influencing factors of ambient temperature, ambient humidity, wind speed and air pressure respectively.

[0031] As a further solution of the present invention: the step S3 of analyzing the potential faults of the photovoltaic power station according to the collected and processed video images includes the following steps:

[0032] Based on the real-time collected and pre-processed video images, deep learning models are used to identify potential faults;

[0033] Collect and pre-process historical video images, cut the historical video images into image frames, and mark the potential fault types and locations in the image frames. The potential fault types include: photovoltaic panel fault, inverter fault and wire fault; photovoltaic panel fault includes: stains, cracks and damage, hot spots and shadows; inverter fault includes: inverter housing deformation and cracks; wire fault includes: cable sheath wear and cracking, loose wiring and / or falling off;

[0034] Input the labeled image frames of potential faults into the convolutional neural network model in the deep learning model for training and learning;

[0035] According to the trained convolutional neural network model, the video images collected and processed in real time are input into the trained convolutional neural network model to automatically identify the type and location of potential faults and output the results.

[0036] As a further solution of the present invention: the step S3 of analyzing and calculating the fault coefficient and planning the inspection frequency according to the identification result includes the following steps:

[0037] According to the recognition results, the pixel area of ​​the potential fault in the video image is extracted, and the total number of times the potential fault occurs during the running time is counted;

[0038] By analyzing the formula:

[0039] Obtain the fault coefficient G; where M represents the pixel area of ​​potential faults, B represents the total area of ​​the photovoltaic panel, Z represents the area of ​​the inverter housing, D represents the total length of the cables and wiring, N represents the total number of potential faults occurring during the operating time, and T represents the operating time;

[0040] Plan the inspection frequency based on the failure coefficient. When the failure coefficient does not exceed the threshold, the inspection frequency is once a year. When the failure coefficient exceeds the threshold, the inspection frequency is once a month.

[0041] As a further solution of the present invention: the step S4 comprises the following steps:

[0042] By analyzing the formula:

[0043] Get the comprehensive evaluation index β; where R I,U represents the harmonic rate pollution value, λ represents the energy loss rate, C represents the environmental impact coefficient, G represents the fault coefficient, W 1 , W 2 , W 3 and W 4 Respectively represent the weight coefficients of harmonic rate pollution value, energy loss rate, environmental impact coefficient and fault coefficient;

[0044] According to the value of the comprehensive evaluation index, set the threshold and divide the evaluation of the operating status into levels;

[0045] When β<20%, the operation status is the attention status, and inspection is performed according to the inspection frequency set by the fault coefficient;

[0046] When 20%≤β≤60%, the operation status is the warning status, and the inspection frequency set according to the fault coefficient is increased from once a month to once a week;

[0047] When β>60%, the operating status is warning status and the inspection frequency is once a day.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] The present invention collects operation data and environmental data for in-depth data analysis to accurately obtain harmonic rate pollution value, energy loss rate, and environmental impact coefficient;

[0050] The present invention trains historical video images through a deep learning model, and the model can learn the characteristics of various fault types, so as to accurately identify faults in real-time video images, including the specific type and precise location of the fault; this accurate and real-time fault identification capability enables operation and maintenance personnel to respond quickly and reduce troubleshooting time; the inspection frequency is planned based on the fault coefficient analyzed according to the identification results, thereby improving the efficiency of fault inspection;

[0051] The present invention provides a comprehensive decision-making basis for the operation and maintenance of photovoltaic power stations by introducing a comprehensive evaluation index; accurately analyzes the overall operation status of the photovoltaic power station by comprehensively analyzing data in multiple dimensions such as harmonic rate pollution value, energy loss rate, environmental impact coefficient and fault coefficient; dynamically divides the operation levels and formulates corresponding inspection plans according to different levels; this dynamic adjustment mechanism ensures the rational allocation of inspection resources and improves the pertinence and effectiveness of inspections. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] 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, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0053] Figure 1 is a flow chart of the method of the present invention;

[0054] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0055] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than 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 are within the scope of protection of the present invention.

[0056] See also Figure 1 The first embodiment of the present invention provides a rural photovoltaic grid-connected management method, comprising the following steps:

[0057] S1: Collect the operation data and environmental data of the photovoltaic power station; monitor the operation status of the photovoltaic power station in real time and collect monitoring video images of the photovoltaic power station;

[0058] S2: preprocessing the collected operation data and environmental data including: cleaning and standardization; preprocessing the collected video images including: denoising and image enhancement;

[0059] S3: Analyze and calculate the harmonic rate pollution value and energy loss rate based on the collected and processed operating data; analyze and calculate the environmental impact coefficient based on the collected and processed environmental data; analyze and analyze the potential faults of the photovoltaic power station based on the collected and processed video images; analyze and calculate the fault coefficient based on the recognition results and plan the inspection frequency;

[0060] S4: Calculate the comprehensive evaluation index based on the harmonic rate pollution value, energy loss rate, environmental impact coefficient and fault coefficient, divide the threshold value into evaluation values ​​according to the analysis results, and dynamically adjust the inspection frequency.

[0061] Specifically, the operation data of the photovoltaic power station is collected in real time through the sensor network of the photovoltaic power station. At the same time, environmental data is collected, which has a direct impact on the performance of the photovoltaic power station. Install high-definition cameras to monitor the photovoltaic array area, inverter area, grid cabinet area, connection point and other areas of the photovoltaic power station in real time and collect video image data. Ensure that the camera covers the main equipment, connection lines and surrounding environment of the power station so as to detect abnormal conditions in time. Clean the collected operation data and environmental data to remove abnormal values ​​and missing values. Standardize the data so that data from different sources and different dimensions can be compared and analyzed on the same scale. De-noise the collected video images to enhance the image contrast and edge information. Analyze and calculate the harmonic rate pollution value, energy loss rate and environmental impact coefficient based on the processed operation data and environmental data. Use image processing technology to identify potential faults in the processed video images. Analyze and calculate the fault coefficient based on the identification results. Dynamically plan the inspection frequency based on the fault coefficient. Comprehensively analyze the calculated data to obtain a comprehensive evaluation index. The comprehensive evaluation index reflects the overall operation status and performance of the power station. According to the comprehensive evaluation index, different thresholds are set to divide the operating status of the power station into different levels; according to the operating status level, the inspection frequency is dynamically adjusted.

[0062] In one embodiment of the present invention, the step S1 of collecting the operation data and environmental data of the photovoltaic power station includes the following steps:

[0063] Operation data include: current, voltage, grid-connected power generation, photovoltaic power generation, total area of ​​photovoltaic panels, inverter housing area, total length of cables and wiring, and operating time;

[0064] Environmental data include: ambient temperature, ambient humidity, air pressure and wind speed.

[0065] Specifically, the current sensor is used to monitor the current value of each branch of the photovoltaic power station in real time. The voltage sensor is used to measure the voltage value of key locations such as the photovoltaic cell array and the inverter output. The reading from the meter at the gateway between the power station and the power grid indicates the total power transmitted by the power station to the power grid. The power generation system of the photovoltaic power station is used to measure the actual power generated by the power station. The total area of ​​all photovoltaic panels is counted; the outer shell area of ​​the inverter is measured and recorded. The total length of all cables and wiring in the power station is counted; and the continuous operation time of the power station is recorded. The ambient temperature in the photovoltaic power station is measured by an environmental monitor. The humidity sensor is used to monitor the air humidity in the power station in real time. The air pressure value in the area where the power station is located is measured by a barometer. The wind speed around the power station is monitored by an anemometer.

[0066] In one embodiment of the present invention, the step S3 includes the following steps:

[0067] Perform frequency domain analysis on the collected and processed voltage and current signals according to the fast Fourier transform algorithm FFT; convert the time domain signal into a frequency domain signal; draw a spectrum diagram according to the conversion result, where the horizontal axis of the spectrum diagram represents the frequency and the vertical axis of the spectrum diagram represents the signal strength or power; identify and extract the fundamental wave and harmonic wave according to the spectrum diagram;

[0068] By analyzing the formula:

[0069] Get the effective value of current I e ; Where i(n) represents the nth sampling value of the current, and N represents the total number of sampling points;

[0070] By analyzing the formula:

[0071] Get the voltage effective value U e ; Wherein, U(n) represents the nth sampling value of the voltage, and N represents the total number of sampling points;

[0072] By analyzing the formula:

[0073]

[0074] The current harmonic distortion rate THD I and voltage harmonic distortion rate THDU are obtained; where I 1 Indicates the effective value of fundamental current, U 1 Indicates the effective value of the fundamental voltage, I h Indicates the effective value of the hth harmonic current, U h It represents the effective value of the hth harmonic voltage, where H is the total number of harmonics;

[0075] By analyzing the formula:

[0076]

[0077] Get the harmonic rate pollution value R I,U Among them, B I Indicates the standard limit of current harmonic distortion rate, B U Indicates the standard limit of voltage harmonic distortion rate;

[0078] By analyzing the formula:

[0079] The energy loss rate λ is obtained; where Q represents photovoltaic power generation and L represents grid-connected power generation.

[0080] Specifically, the voltage and current signals of the photovoltaic power station are collected in real time. The fast Fourier transform algorithm FFT is applied to the pre-processed voltage and current signals to convert the time domain signal into a frequency domain signal. The FFT algorithm can decompose the signal into sinusoidal components of different frequencies, namely the fundamental wave and the harmonic wave. According to the FFT conversion result, a spectrum diagram is drawn. The horizontal axis of the spectrum diagram represents the frequency, and the vertical axis represents the strength or power of the signal. The location and strength of the fundamental wave and the harmonic wave can be intuitively identified through the spectrum diagram. According to the spectrum diagram, the fundamental wave and the harmonic wave are identified and extracted. The fundamental wave is the sinusoidal wave component with the lowest frequency, and the harmonic wave is the sinusoidal wave component with a frequency higher than the fundamental wave. The current harmonic distortion rate and the voltage harmonic distortion rate are calculated using the formula, and the harmonic distortion rate pollution value is calculated using the formula according to the calculation results; the standard limit of the current harmonic distortion rate The standard limit of the current harmonic distortion rate is usually determined according to the grid-connected voltage level and the type of inverter. Generally speaking, for photovoltaic power generation systems connected to voltage levels of 10kV and below, the current harmonic distortion rate should not exceed 5%, and the current harmonic distortion rate should not exceed 4%. For photovoltaic power generation systems with higher voltage levels, the standard limits may be different, but usually will not exceed the stricter limits. In practical applications, these standard limits should be determined by referring to the relevant regulations of the local grid operator and the design requirements of the photovoltaic system. The specific standard limits may vary from region to region and time to time. The energy loss rate is calculated using a formula that reflects the difference between photovoltaic power generation and grid-connected power generation, that is, the energy loss caused by various factors.

[0081] In one embodiment of the present invention, the step S3 of analyzing and calculating the environmental impact coefficient based on the collected and processed environmental data includes the following steps:

[0082] By analyzing the formula: C = (t * ε 1 +w*ε 2 +v*ε 3 +f*ε 4 )×100%

[0083] Get the environmental impact coefficient C; where t represents the ambient temperature, w represents the ambient humidity, v represents the wind speed, and f represents the air pressure; ε 1 , ε 2 , ε 3 and ε 4 They represent the influencing factors of ambient temperature, ambient humidity, wind speed and air pressure respectively.

[0084] Specifically, according to the collected and pre-processed environmental data, the value of each environmental factor is multiplied by its corresponding impact factor through the formula, and then all the results are added and multiplied by the percentage to obtain the environmental impact coefficient. The value of the impact factor represents the relative contribution of each factor to the environmental impact coefficient C, ε 1 , ε 2 , ε 3and ε 4 They are 0.4, 0.25, 0.2, and 0.15 respectively. The value of the impact factor may change with time and location. Therefore, in order to ensure the accuracy of the calculation, the value of the impact factor needs to be adjusted dynamically on a regular basis. In addition, the formula can be adjusted and optimized according to actual needs to better adapt to specific application scenarios and needs.

[0085] In one embodiment of the present invention, the step S3 of analyzing potential faults of the photovoltaic power station according to the collected and processed video images includes the following steps:

[0086] Based on the real-time collected and pre-processed video images, deep learning models are used to identify potential faults;

[0087] Collect and pre-process historical video images, cut the historical video images into image frames, and mark the potential fault types and locations in the image frames. The potential fault types include: photovoltaic panel fault, inverter fault and wire fault; photovoltaic panel fault includes: stains, cracks and damage, hot spots and shadows; inverter fault includes: inverter housing deformation and cracks; wire fault includes: cable sheath wear and cracking, loose wiring and / or falling off;

[0088] Input the labeled image frames of potential faults into the convolutional neural network model in the deep learning model for training and learning;

[0089] According to the trained convolutional neural network model, the video images collected and processed in real time are input into the trained convolutional neural network model to automatically identify the type and location of potential faults and output the results.

[0090] Specifically, a high-definition camera is used to collect video images of a photovoltaic power station in real time, and the images are preprocessed. The preprocessed images will be used in the subsequent convolutional neural network of the deep learning model to analyze the images to identify potential faults in the photovoltaic power station. The model is trained to accurately detect various types of faults. Collect and process historical video image data, cut the video into image frames, and annotate each frame with the type and location of potential faults. This annotation information includes but is not limited to photovoltaic panel faults, inverter faults, and wire faults. The annotated historical image frames are input into the convolutional neural network model for training, and the model parameters are continuously optimized to improve the recognition accuracy. After the training is completed, the model has the ability to automatically identify different types of faults. Use the trained deep learning model to analyze the real-time video images, automatically identify the type and location of potential faults, and output the results. This process realizes real-time monitoring and management of the operating status of the photovoltaic power station.

[0091] In one embodiment of the present invention, the step S3 of analyzing and calculating the fault coefficient and planning the inspection frequency according to the identification result includes the following steps:

[0092] According to the recognition results, the pixel area of ​​the potential fault in the video image is extracted, and the total number of times the potential fault occurs during the running time is counted;

[0093] By analyzing the formula:

[0094] Obtain the fault coefficient G; where M represents the pixel area of ​​potential faults, B represents the total area of ​​the photovoltaic panel, Z represents the area of ​​the inverter housing, D represents the total length of the cables and wiring, N represents the total number of potential faults occurring during the operating time, and T represents the operating time;

[0095] Plan the inspection frequency based on the failure coefficient. When the failure coefficient does not exceed the threshold, the inspection frequency is once a year. When the failure coefficient exceeds the threshold, the inspection frequency is once a month.

[0096] Specifically, based on the recognition results, the fault area is located by edge detection or threshold segmentation using image processing technology. The pixel area of ​​the potential fault is extracted from the video image. During the running time, the number of times a potential fault is identified each time is recorded in order to track the frequency of fault occurrence. The fault coefficient is calculated using the analytical formula, and the corresponding inspection frequency strategy is formulated based on the calculation results. The threshold is set to 40%, and the threshold is adjusted dynamically based on actual conditions and historical data. When the fault coefficient does not exceed the threshold, the inspection frequency is set to once a year; when the fault coefficient exceeds the threshold, the inspection frequency is adjusted to once a month. This ensures that potential faults are discovered and handled in a timely manner.

[0097] In one embodiment of the present invention, the step S4 comprises the following steps:

[0098] By analyzing the formula:

[0099] Get the comprehensive evaluation index β; where R I,U represents the harmonic rate pollution value, λ represents the energy loss rate, C represents the environmental impact coefficient, G represents the fault coefficient, W 1 , W 2 , W 3 and W 4 Respectively represent the weight coefficients of harmonic rate pollution value, energy loss rate, environmental impact coefficient and fault coefficient;

[0100] According to the value of the comprehensive evaluation index, set the threshold and divide the evaluation of the operating status into levels;

[0101] When β<20%, the operation status is the attention status, and inspection is performed according to the inspection frequency set by the fault coefficient;

[0102] When 20%≤β≤60%, the operation status is the warning status, and the inspection frequency set according to the fault coefficient is increased from once a month to once a week;

[0103] When β>60%, the operating status is warning status and the inspection frequency is once a day.

[0104] Specifically, the comprehensive evaluation index is calculated according to the values ​​of harmonic rate pollution value, energy loss rate, environmental impact coefficient and failure coefficient and the corresponding weights. 1 is 0.3, and the energy loss rate weight coefficient W 2 is 0.25, and the environmental impact coefficient weight coefficient W 3 is 0.15, the weight coefficient of the failure coefficient W 4 is 0.3, and W 1 +W 2 +W 3 +W 4 =1. Through experts in related fields or by collecting a large amount of historical data, statistical methods are used to dynamically adjust the weight of each data. According to the value of the comprehensive evaluation index, different thresholds are set to divide the operating status level and dynamically adjust the inspection frequency. When β<20%, the operating status is the attention state, and the inspection is carried out according to the inspection frequency set by the fault coefficient. When 20%≤β≤60%, the operating status is the warning state, and the inspection frequency is increased to once a week based on the inspection frequency set by the fault coefficient. When β>60%, the operating status is the warning state, and the inspection frequency is once a day to ensure that potential problems are discovered and handled in a timely manner. During the implementation process, it is necessary to continuously collect a large amount of data and calculate the comprehensive evaluation index to evaluate the operating status. At the same time, according to the changes in the operating status level, the inspection frequency is adjusted in time. The weight coefficient and threshold may change with changes in the equipment or actual conditions. The weight coefficient and threshold can be dynamically adjusted to ensure the accuracy of the formula.

[0105] See also Figure 2 As shown, the present invention is a rural photovoltaic grid-connected management system, comprising the following modules:

[0106] Data acquisition module: used to collect the operation data and environmental data of the photovoltaic power station; to monitor the operation status of the photovoltaic power station in real time and to collect monitoring video images of the photovoltaic power station;

[0107] Data processing module: used to pre-process the collected operation data and environmental data, including cleaning and standardization; pre-process the collected video images, including denoising and image enhancement;

[0108] Data analysis module: Calculate the harmonic rate pollution value and energy loss rate based on the collected and processed operating data; Calculate the environmental impact coefficient based on the collected and processed environmental data; Analyze and analyze the potential faults of the photovoltaic power station based on the collected and processed video images; Calculate the fault coefficient based on the recognition results and plan the inspection frequency;

[0109] Comprehensive evaluation module: The comprehensive evaluation index is calculated based on the comprehensive analysis of the harmonic rate pollution value, energy loss rate, environmental impact coefficient and fault coefficient, and the threshold is divided according to the analysis results to evaluate the operating status and dynamically adjust the inspection frequency.

[0110] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A rural photovoltaic grid-connected management method, characterized in that: The following steps are involved: S1: Collect the operation data and environmental data of the photovoltaic power station; Monitor the operation status of the photovoltaic power station in real time and collect monitoring video images of the photovoltaic power station; S2: preprocessing the collected operation data and environmental data including: cleaning and standardization; preprocessing the collected video images including: denoising and image enhancement; S3: Analyze and calculate the harmonic rate pollution value and energy loss rate based on the collected and processed operating data; analyze and calculate the environmental impact coefficient based on the collected and processed environmental data; analyze and analyze the potential faults of the photovoltaic power station based on the collected and processed video images; analyze and calculate the fault coefficient based on the recognition results and plan the inspection frequency; S4: Calculate the comprehensive evaluation index based on the harmonic rate pollution value, energy loss rate, environmental impact coefficient and fault coefficient, divide the threshold value into evaluation values ​​according to the analysis results, and dynamically adjust the inspection frequency.

2. A rural photovoltaic grid-connected management method according to claim 1, characterized in that: The step S1 of collecting the operation data and environmental data of the photovoltaic power station includes the following steps: Operation data include: current, voltage, grid-connected power generation, photovoltaic power generation, total area of ​​photovoltaic panels, inverter housing area, total length of cables and wiring, and operating time; Environmental data include: ambient temperature, ambient humidity, air pressure and wind speed.

3. A rural photovoltaic grid-connected management method according to claim 1, characterized in that: The step S3 includes the following steps: Perform frequency domain analysis on the collected and processed voltage and current signals according to the fast Fourier transform algorithm FFT; convert the time domain signal into a frequency domain signal; draw a spectrum diagram according to the conversion result, where the horizontal axis of the spectrum diagram represents the frequency and the vertical axis of the spectrum diagram represents the signal strength or power; identify and extract the fundamental wave and harmonic wave according to the spectrum diagram; By analyzing the formula: Get the effective value of current I e ; Where i(n) represents the nth sampling value of the current, and N represents the total number of sampling points; By analyzing the formula: Get the voltage effective value U e ; Wherein, U(n) represents the nth sampling value of the voltage, and N represents the total number of sampling points; By analyzing the formula: The current harmonic distortion rate THDI and voltage harmonic distortion rate THDU are obtained; where I1 represents the effective value of the fundamental current, U1 represents the effective value of the fundamental voltage, and I h Indicates the effective value of the hth harmonic current, U h It represents the effective value of the hth harmonic voltage, where H is the total number of harmonics; By analyzing the formula: Get the harmonic rate pollution value R I,U Among them, B I Indicates the standard limit of current harmonic distortion rate, B U Indicates the standard limit of voltage harmonic distortion rate; By analyzing the formula: The energy loss rate λ is obtained; where Q represents photovoltaic power generation and L represents grid-connected power generation.

4. A rural photovoltaic grid-connected management method according to claim 1, characterized in that: The step S3 includes the following steps: By analyzing the formula: C = (t*ε1+w*ε2+v*ε3+f*ε4)×100% The environmental impact coefficient C is obtained; wherein t represents the ambient temperature, w represents the ambient humidity, v represents the wind speed, and f represents the air pressure; ε1, ε2, ε3, and ε4 represent the influencing factors of the ambient temperature, ambient humidity, wind speed, and air pressure, respectively.

5. A rural photovoltaic grid-connected management method according to claim 1, characterized in that: Analyzing potential faults of the photovoltaic power station according to the collected and processed video images in step S3 includes the following steps: Based on the real-time collected and pre-processed video images, deep learning models are used to identify potential faults; Collect and pre-process historical video images, cut the historical video images into image frames, and mark the potential fault types and locations in the image frames. The potential fault types include: photovoltaic panel fault, inverter fault and wire fault; photovoltaic panel fault includes: stains, cracks and damage, hot spots and shadows; inverter fault includes: inverter housing deformation and cracks; wire fault includes: cable sheath wear and cracking, loose wiring and / or falling off; Input the labeled image frames of potential faults into the convolutional neural network model in the deep learning model for training and learning; According to the trained convolutional neural network model, the video images collected and processed in real time are input into the trained convolutional neural network model to automatically identify the type and location of potential faults and output the results.

6. A rural photovoltaic grid-connected management method according to claim 5, characterized in that: The step S3 includes the following steps: According to the recognition results, the pixel area of ​​the potential fault in the video image is extracted, and the total number of times the potential fault occurs during the running time is counted; By analyzing the formula: Obtain the fault coefficient G; where M represents the pixel area of ​​potential faults, B represents the total area of ​​the photovoltaic panel, Z represents the area of ​​the inverter housing, D represents the total length of the cables and wiring, N represents the total number of potential faults occurring during the operating time, and T represents the operating time; Plan the inspection frequency based on the failure coefficient. When the failure coefficient does not exceed the threshold, the inspection frequency is once a year. When the failure coefficient exceeds the threshold, the inspection frequency is once a month.

7. A rural photovoltaic grid-connected management method according to claim 1, characterized in that: The step S4 comprises the following steps: By analyzing the formula: Get the comprehensive evaluation index β; where R I,U represents the harmonic rate pollution value, λ represents the energy loss rate, C represents the environmental impact coefficient, G represents the fault coefficient, W1, W2, W3 and W4 represent the weight coefficients of the harmonic rate pollution value, energy loss rate, environmental impact coefficient and fault coefficient respectively; According to the value of the comprehensive evaluation index, set the threshold and divide the evaluation of the operating status into levels; When β<20%, the operation status is the attention status, and inspection is performed according to the inspection frequency set by the fault coefficient; When 20%≤β≤60%, the operation status is the warning status, and the inspection frequency set according to the fault coefficient is increased from once a month to once a week; When β>60%, the operating status is warning status and the inspection frequency is once a day.

8. A rural photovoltaic grid-connected management system, using a rural photovoltaic grid-connected management method as claimed in any one of claims 1 to 7, characterized in that: Includes the following modules: Data acquisition module: used to collect the operation data and environmental data of the photovoltaic power station; to monitor the operation status of the photovoltaic power station in real time and to collect monitoring video images of the photovoltaic power station; Data processing module: used to pre-process the collected operation data and environmental data, including cleaning and standardization; pre-process the collected video images, including denoising and image enhancement; Data analysis module: Calculate the harmonic rate pollution value and energy loss rate based on the collected and processed operating data; Calculate the environmental impact coefficient based on the collected and processed environmental data; Analyze and analyze the potential faults of the photovoltaic power station based on the collected and processed video images; Calculate the fault coefficient based on the recognition results and plan the inspection frequency; Comprehensive evaluation module: The comprehensive evaluation index is calculated based on the comprehensive analysis of the harmonic rate pollution value, energy loss rate, environmental impact coefficient and fault coefficient, and the threshold is divided according to the analysis results to evaluate the operating status and dynamically adjust the inspection frequency.

Images