An inverter health assessment system

Abstract

The invention discloses an inverter health assessment system, which relates to the technical field of photovoltaic power station equipment, and comprises a data collection module, wherein the data collection module is used to collect real-time data of the inverter; a dust accumulation assessment module, wherein the dust accumulation assessment module is used to decompose and calculate the loss amount of the photovoltaic power station caused by dust accumulation based on the inverter as a unit; and a health assessment module; by diagnosing photovoltaic components through a data analysis method, not only can the inefficient operation of the equipment be efficiently identified, but also the fault type can be accurately determined in combination with data characteristics, and the operation efficiency and health of the power station can be quantitatively evaluated; by decomposing and calculating the loss of electricity caused by dust accumulation in the photovoltaic power station based on the inverter as a unit, based on the actual power generation power of the photovoltaic inverter, and the irradiation, temperature and other data measured on site of the power station, the dust accumulation conditions in different areas of the power station can be automatically and individually evaluated.

Description

Technical Field

[0001] The present invention relates to the technical field of photovoltaic power station equipment, and more specifically, to an inverter health assessment system. Background Art

[0002] Photovoltaic power stations can only guarantee profits if they generate electricity stably, continuously and efficiently. However, various equipment problems and external environmental influences lead to frequent failures, which reduces the power generation efficiency of photovoltaic power generation systems. Among them, photovoltaic modules are the core components of photovoltaic power generation units. Their fault diagnosis and inefficiency identification methods have become key technologies. This technology is also a key research direction to help photovoltaic power stations refine their operation and maintenance control and improve the energy efficiency of photovoltaic power generation systems.

[0003] Patent application number CN117277435A discloses a method, system and device for health assessment of photovoltaic inverters, which relates to the field of photovoltaic power generation. In this solution, historical data of photovoltaic inverters within a preset time period are obtained; health indicators are calculated based on historical data, and an assessment system model is established based on each health indicator. The health indicators used to establish the assessment system model include at least two of indicators for characterizing fault conditions, indicators for characterizing power generation conditions and indicators for characterizing equipment performance conditions; the health score of the photovoltaic inverter is calculated based on each health indicator and the assessment system model, so that the health of the inverter can be more accurately assessed;

[0004] Most of the inspection tasks of photovoltaic modules are still carried out manually. Some newly built power stations with a higher degree of intelligence use drone inspections. Although this inspection method can perform visible fault detection according to instructions, it is often unable to undertake a large number of inspection tasks due to its insufficient battery life, and cannot identify faults that are not shown in the image. The equipment investment and maintenance costs are also high, and it has certain limitations. Summary of the invention

[0005] In order to solve the above problems, the present invention provides an inverter health assessment system.

[0006] An inverter health assessment system, comprising:

[0007] A data collection module, wherein the data collection module is used to collect real-time data of the inverter;

[0008] A dust accumulation assessment module, which is used to decompose and calculate the loss amount of the photovoltaic power station caused by dust accumulation based on the inverter;

[0009] A health assessment module, the health assessment module comprising an assessment unit and a prediction unit, the assessment unit being used to assess the working performance of the inverter in real time according to the data collected by the data collection module and the data of the dust accumulation assessment module;

[0010] First, the characteristic factors of the inverter working performance are obtained, specifically:

[0011] Obtain the real-time power factor P of the inverter;

[0012] Obtain the real-time efficiency value E of the inverter;

[0013] Obtain the real-time temperature abnormality T of the inverter;

[0014] Obtain the real-time current unbalance value Y of the inverter;

[0015] Obtain the real-time voltage imbalance value U of the inverter;

[0016] The acquired influencing factors are standardized, specifically:

[0017] According to the formula Calculate and obtain the standardized i-th characteristic factor X i , where β i represents the i-th characteristic factor, μ i represents the average value of the i-th characteristic factor, σ i represents the standard deviation of the i-th characteristic factor;

[0018] According to the formula Get the real-time evaluation value A of the inverter, where n is the number of characteristic factors, B i is a preset weight coefficient corresponding to each characteristic factor, where g is the energy loss factor of the photovoltaic power station where the inverter is located in the current time period;

[0019] A threshold value of the real-time evaluation value is set in advance, and the real-time evaluation value A of the inverter and the threshold value of the real-time evaluation value are connected;

[0020] If the real-time evaluation value A of the inverter is greater than the threshold of the real-time evaluation value, the working state of the inverter is judged to be abnormal, and an alarm signal is generated and transmitted to the mobile terminal of the maintenance personnel; if the real-time evaluation value A of the inverter is less than or equal to the threshold of the real-time evaluation value, a signal is generated and transmitted to the prediction unit;

[0021] The prediction unit is used to predict the degradation of inverter performance over time based on the real-time data collected by the data collection module.

[0022] Preferably, the health assessment module further includes a diagnosis unit, and the diagnosis unit is used to identify the fault of the inverter after the assessment unit determines that the working state of the inverter is abnormal, specifically:

[0023] The inverter fault types are set in advance as electrical fault, thermal fault and mechanical fault, and the three types are marked as fault type 1, fault type 2 and fault type 3 respectively;

[0024] According to the formula Calculate and obtain the given input H i The probability value Q(y=K|H i ), where H i represents the i-th input vector, Z k is the network output (logit) corresponding to fault type K, e is the base of the natural logarithm, and 3 is because the number of fault types is 3;

[0025] Get the probability of fault type 1

[0026] Get the probability of fault type 2

[0027] Get the probability of fault type 3

[0028] The probability of fault type 1 Q(y=1|H i ), the probability of fault type 2 Q(y=2|H i ) and the probability of fault type 3 Q(y=3|H i ) for comparison, and the fault type with the highest probability is determined as the current fault type of the inverter, and is transmitted to the maintenance personnel's mobile terminal through a signal.

[0029] Preferably, the i-th input vector H i Represents the data characteristic vector input to the diagnostic unit, and the specific acquisition method is as follows:

[0030] For fault type 1 electrical fault, the input data characteristic vector is H1;

[0031] For fault type 2 thermal fault, the input data feature vector is H2;

[0032] For fault type 3 mechanical fault, the input data characteristic vector is H3;

[0033] According to the formula: H1=(P*ω1)+(E*ω2)+(T*ω3)+(Y*ω4)+(U*ω5);

[0034] H2=(P*φ1)+(E*φ2)+(T*φ3)+(Y*φ4)+(U*φ5);

[0035] H3=(P*γ1)+(E*γ2)+(T*γ3);

[0036] H1, H2 and H3 are obtained, wherein ω1, ω2, ω3, ω4 and ω5, φ1, φ2, φ3, φ4 and φ5, and γ1, γ2 and γ3 are all preset proportional coefficients.

[0037] Preferably, the specific working mode of the prediction unit is as follows:

[0038] Set a time period M in advance, in days;

[0039] Obtain the average evaluation value M1 of the inverter in the past M days;

[0040] Obtain the frequency M2 of the inverter evaluation value being less than the evaluation value threshold value in the past M days;

[0041] Obtain the difference M3 between the time when the inverter evaluation value was less than the evaluation value threshold in the past M days and the current time;

[0042] The predicted evaluation value S(abnormal|N) of the inverter tomorrow is obtained by calculation, where δ1, δ2 and δ3 are preset proportional coefficients, e is the base of the natural logarithm, and ε is the historical maintenance impact factor;

[0043] A threshold value of the predicted evaluation value is set. If the predicted evaluation value S (abnormal|N) of the inverter tomorrow is greater than the threshold value of the predicted evaluation value, it is determined that the state of the inverter tomorrow will be abnormal, and a signal is transmitted to the mobile terminal of the maintenance personnel.

[0044] Preferably, the historical maintenance impact factor ε is obtained as follows:

[0045] Obtain the historical working time value D1 of the inverter;

[0046] Obtain the maintenance data value D2 of the inverter;

[0047] Obtain the predicted environmental data value D3 of the inverter;

[0048] According to the formula ε=(D1*θ1)+(D2*θ2)+(D3*θ3), the historical maintenance impact factor ε is obtained, where θ1, θ2 and θ3 are preset proportional coefficients.

[0049] Preferably, the specific working mode of the dust accumulation assessment module is as follows:

[0050] Obtain the theoretical power generation energy E1 of the photovoltaic power station where the inverter is located;

[0051] Obtain the actual power generation energy E2 of the photovoltaic power station where the inverter is located in the current time period;

[0052] According to the formula The energy loss factor g of the photovoltaic power station where the inverter is located in the current time period is obtained.

[0053] Preferably, the specific method of obtaining the theoretical power generation energy E1 of the photovoltaic power station is as follows:

[0054] Obtain the total area A of the photovoltaic module where the inverter is located;

[0055] Obtain the standard power generation efficiency η of the photovoltaic module where the inverter is located;

[0056] Obtain the irradiance I1 of the photovoltaic module where the inverter is located;

[0057] According to the formula E1=A*η*I1*t, the theoretical power generation energy E1 of the photovoltaic power station is obtained, where t is time in hours.

[0058] Preferably, the dust accumulation evaluation module is also used to evaluate a single area where the inverter is located, to evaluate whether the dust accumulation energy loss in the area where the inverter is located is abnormal, specifically:

[0059] Take the inverter as the center and draw a circle with a preset radius to obtain the unit area of ​​the inverter and the area A of the unit area. i ;

[0060] According to the formula E3 = A i *η*I1*t*K1, the dust loss energy E3 in the inverter area is obtained, where A i represents the unit area where the i-th inverter is located, and K1 is the dust accumulation coefficient of the area;

[0061] A dust accumulation loss energy threshold is set in advance. If the dust accumulation loss energy E3 in the area where the inverter is located is greater than the dust accumulation loss energy threshold, a signal is generated and transmitted to the mobile phone terminal of the maintenance personnel.

[0062] Preferably, the specific method of obtaining the dust accumulation coefficient K1 is as follows:

[0063] According to the formula The dust accumulation coefficient K1 is obtained by calculation, where K1 represents the dust shielding rate.

[0064] Preferably, the inverter cleaning day is obtained in the following manner:

[0065] Select a reference time period, clean the solar inverter in the absence of precipitation, observe the change in power generation after cleaning, record the power generation data every day, analyze the power generation data, and determine the time when the benefits of the cleaning effect reach the critical point, that is, after r days after cleaning, the power generation no longer increases significantly, and r days is defined as the cleaning failure period;

[0066] The cumulative power increase after cleaning is obtained, and the cleaning benefit value is obtained by dividing the cumulative power increase by the cleaning cost;

[0067] t clean =argmax(cleaning benefit value(1-rainfall impact factor));

[0068] If the cleaning benefit value is greater than 1 and t clean If it is less than the cleaning failure period r, the inverter cleaning day is equal to t clean , if t clean If it is greater than the cleaning failure period r, then the inverter cleaning day is r days after the last cleaning.

[0069] Beneficial effects: Diagnosing PV panels through data analysis can not only effectively identify inefficient equipment operation, but also accurately determine the type of fault based on data characteristics, and quantitatively evaluate the operating efficiency and health of the power station;

[0070] Taking the inverter as the unit, the power loss caused by dust accumulation in the photovoltaic power station is decomposed and calculated. Based on the actual power generation power of the photovoltaic inverter and the irradiation, temperature and other data measured at the power station site, the dust accumulation in different areas of the power station can be automatically evaluated separately, accurately reflecting the dust accumulation in the photovoltaic power station and the power generation loss affected by dust. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0072] like Figure 1 As shown, an inverter health assessment system includes:

[0073] It should be noted that the stable, continuous and efficient power generation of photovoltaic power stations can guarantee the benefits, but various equipment problems and external environmental influences lead to frequent failures, which reduce the power generation efficiency of photovoltaic power generation systems. Among them, the inverter is the core component of the photovoltaic power generation unit, and its fault diagnosis and inefficiency identification method has become a key technology. This technology is also a key research direction to help photovoltaic power stations refine their operation and maintenance control and improve the energy efficiency of photovoltaic power generation systems.

[0074] Most of the inspection tasks of photovoltaic modules are still carried out manually. Some newly built power stations with a high degree of intelligence use drone inspections. Although this inspection method can perform visible fault troubleshooting according to instructions, it is often unable to undertake a large number of inspection tasks due to its insufficient endurance, and cannot identify faults that are not shown in the image. The equipment investment and maintenance costs are also high, which has certain limitations. This technical solution diagnoses photovoltaic modules through data analysis methods, which can not only efficiently identify the inefficient operation of equipment, but also accurately determine the type of faults based on data characteristics, and quantitatively evaluate the operating efficiency and health of the power station;

[0075] A data collection module, which is used to collect real-time data of the inverter; it should be noted that in this embodiment, the real-time data includes but is not limited to power generation, voltage, current, temperature, irradiance, etc.;

[0076] A dust accumulation assessment module, which is used to decompose and calculate the loss amount of the photovoltaic power station caused by dust accumulation based on the inverter as a unit; it should be noted that by separately evaluating the dust accumulation conditions in different areas of the photovoltaic power station, the dust accumulation conditions of the photovoltaic power station and the power generation loss affected by the dust are accurately reflected;

[0077] A health assessment module, the health assessment module comprising an assessment unit and a prediction unit, the assessment unit being used to assess the working performance of the inverter in real time according to the data collected by the data collection module and the data of the dust accumulation assessment module;

[0078] First, characteristic factors of the inverter working performance are obtained, specifically: It should be noted that the characteristic factors of the inverter working performance are various influencing factors that affect the inverter working performance when the inverter is working. By collecting this information, it is convenient to evaluate the health status of the inverter when it is working;

[0079] The real-time power factor P of the inverter is obtained; it should be noted that the power factor is an indicator to measure the efficiency of the inverter, which indicates the ratio of the actual power output by the inverter to the apparent power. A power factor close to 1 indicates that the inverter efficiency is high. In this embodiment, the value of the real-time power factor P can be the actual power of the inverter divided by the apparent power of the inverter. The actual power, also called active power, is the power actually output by the inverter, in watts. The apparent power is the product of the voltage and current in the circuit, in volt-amperes.

[0080] The real-time efficiency value E of the inverter is obtained; it should be noted that the efficiency represents the ability of the inverter to convert the input solar energy into output electrical energy. In this embodiment, the value of the real-time efficiency value E can be the actual power output by the current inverter divided by the input power. The input power is the energy radiated by the solar energy. The calculation method is irradiance (unit is watt per square meter, W / m 2 ) multiplied by the area of ​​the photovoltaic panel (in square meters, m 2 );

[0081] The real-time temperature anomaly T of the inverter is obtained; it should be noted that the temperature anomaly is a relative indicator, which is used to measure the deviation degree between the current inverter temperature and the historical average temperature of the inverter;

[0082] The real-time current unbalance value Y of the inverter is obtained; it should be noted that, in this embodiment, the current unbalance value is the current difference value between multiple inputs of the inverter;

[0083] The real-time voltage imbalance value U of the inverter is obtained; it should be noted that, in this embodiment, the voltage imbalance value is the voltage difference value between multiple inputs of the inverter;

[0084] The acquired influencing factors are standardized, specifically:

[0085] According to the formula Calculate and obtain the standardized i-th characteristic factor X i , where β i represents the i-th characteristic factor, μ i represents the average value of the i-th characteristic factor, σ i represents the standard deviation of the i-th characteristic factor; it should be noted that in order to reduce the error caused by the characteristic factors in subsequent processing, the acquired characteristic factors need to be standardized before processing. For example, the real-time power factor P is substituted to obtain

[0086] According to the formula Get the real-time evaluation value A of the inverter, where n is the number of characteristic factors, B i is the preset weight coefficient corresponding to each characteristic factor, where g is the energy loss factor of the photovoltaic power station where the inverter is located in the current time period; it should be noted that in this embodiment, the number of n is 6, β i They are β1, β2, β3, β4, β5 and β6 respectively. In this embodiment, the values ​​of β1, β2, β3, β4, β5 and β6 may be 0.127, 0.096, 0.238, 0.276, 0.078 and 0.185;

[0087] It should also be noted that the real-time working status of the inverter can be diagnosed by using data analysis methods to comprehensively consider the real-time working parameters of the inverter and various influencing factors of the working environment, which is conducive to the subsequent efficient identification of whether the overall photovoltaic equipment has inefficient operation and to quantitatively evaluate the operating efficiency and health of the power station;

[0088] A threshold value of the real-time evaluation value is set in advance, and the real-time evaluation value A of the inverter and the threshold value of the real-time evaluation value are connected;

[0089] If the real-time evaluation value A of the inverter is greater than the threshold of the real-time evaluation value, the working state of the inverter is judged to be abnormal, and an alarm signal is generated and transmitted to the mobile terminal of the maintenance personnel. If the real-time evaluation value A of the inverter is less than or equal to the threshold of the real-time evaluation value, a signal is generated and transmitted to the prediction unit. It should be noted that the maintenance personnel are pre-set staff members, and the mobile terminal of the maintenance personnel is electrically connected to the health assessment module. When it is judged that the working state of the inverter is abnormal, the alarm signal is transmitted to the maintenance personnel in a short time. By real-time monitoring and evaluating the working state of the inverter, the inverter that may affect the working efficiency of the photovoltaic power station can be maintained at the first time to reduce possible losses.

[0090] The prediction unit is used to predict the degradation of inverter performance over time based on the real-time data collected by the data collection module.

[0091] As an optional embodiment: the health assessment module further includes a diagnosis unit, and the diagnosis unit is used to identify the fault of the inverter after the assessment unit determines that the working state of the inverter is abnormal, specifically: It should be noted that in the actual maintenance process, even if the working state of the inverter is abnormal, it is obtained at the first time, but because there are many types of faults of the inverter, if the fault type cannot be known, it will also cause the subsequent maintenance time to be prolonged;

[0092] The inverter fault types are set in advance as electrical fault, thermal fault and mechanical fault, and the three types are marked as fault type 1, fault type 2 and fault type 3 respectively;

[0093] According to the formula Calculate and obtain the given input H i The probability value Q(y=K|H i ), where H i represents the i-th input vector, Z k is the network output (logit) corresponding to fault type K, e is the base of the natural logarithm, and 3 is because the number of fault types is 3; it should be noted that in this embodiment, Z k It is calculated by a feedforward neural network. The specific steps are as follows:

[0094] Input feature vector: First, the feature vector H containing the real-time data of the inverter i Input to the neural network;

[0095] Forward propagation: feature vector H i A forward pass through the layers of a neural network typically involves:

[0096] Hidden layer: one or more hidden layers, which may include convolutional layers and activation functions;

[0097] Fully connected layers: After the hidden layers, there may be one or more fully connected layers to integrate and process the feature representations passed from the hidden layers;

[0098] Output layer: Finally, the feature vector is passed through a final fully connected layer, which usually has no activation function or uses a linear activation function (for logits) to produce the original output value;

[0099] Get logits: The output of the last fully connected layer is logitsZ k

[0100] Each Z k It can be considered as a non-standardized “score” or “measure” of the inverter belonging to fault category K by the model;

[0101] Get the probability of fault type 1

[0102] Get the probability of fault type 2

[0103] Get the probability of fault type 3

[0104] The probability of fault type 1 Q(y=1|H i ), the probability of fault type 2 Q(y=2|H i ) and the probability of fault type 3 Q(y=3|H i ) for comparison, the fault type with the highest probability is determined as the current fault type of the inverter, and transmitted to the mobile terminal of the maintenance personnel through a signal. It should be noted that, for example, the input feature vectors H1, H2 and H3 are 0.85, 0.9 and -01.5, assuming that the output Z k They are Z1=1, Z1=0.5 and Z1=0.3 respectively;

[0105] So

[0106]

[0107]

[0108] It should also be noted that the possible fault type of the inverter can be obtained in a short time. After determining that the inverter has a fault, the fault type of your inverter can be preliminarily determined and transmitted to the maintenance personnel. This allows the maintenance personnel to have a preliminary understanding of the fault before maintenance, reducing the impact of inverter maintenance on the overall photovoltaic power station.

[0109] As an optional embodiment: the i-th input vector H i Represents the data characteristic vector input to the diagnostic unit, and the specific acquisition method is as follows:

[0110] For fault type 1 electrical fault, the input data characteristic vector is H1;

[0111] For fault type 2 thermal fault, the input data feature vector is H2;

[0112] For fault type 3 mechanical fault, the input data characteristic vector is H3;

[0113] According to the formula: H1 = (P*ω1) + (E*ω2) + (T*ω3) + (Y*ω4) + (U*ω5); It should be noted that in the present embodiment, if the inverter has an electrical fault, P may be close to or lower than the normal value, because the electrical fault may cause unstable power output, E may drop significantly because the fault increases energy loss, T may be higher than the normal value, because overload or short circuit may cause overheating, Y and U may have larger values, indicating an imbalance in the circuit; It should also be noted that in the present embodiment, the values ​​of ω1, ω2, ω3, ω4 and ω5 may be 0.192, 0.098, 0.187, 0.091 and 0.24;

[0114] H2=(P*φ1)+(E*φ2)+(T*φ3)+(Y*φ4)+(U*φ5); It should be noted that, in this embodiment, if the inverter has a thermal fault problem, P may decrease slightly because the temperature increase may affect the inverter performance, E may decrease because high temperature may cause a decrease in efficiency, T will be significantly higher than the normal value, which is a direct indicator of thermal fault, and Y and U may change, but may not be the main feature; It should also be noted that in this embodiment, the values ​​of φ1, φ2, φ3, φ4 and φ5 can be 0.092, 0.073, 0.719, 0.091 and 0.025;

[0115] H3=(P*γ1)+(E*γ2)+(T*γ3); It should be noted that, in this embodiment, if a mechanical failure occurs in the inverter, P and E may not be immediately affected unless the failure causes overheating; T will increase over time because a fan failure may cause poor heat dissipation; Y and U may not be affected unless the mechanical failure indirectly affects the electrical performance; It should also be noted that, in this embodiment, the values ​​of γ1, γ2 and γ3 may be 0.276, 0.298 and 0.426;

[0116] H1, H2 and H3 are obtained, wherein ω1, ω2, ω3, ω4 and ω5, φ1, φ2, φ3, φ4 and φ5, and γ1, γ2 and γ3 are all preset proportional coefficients.

[0117] As an optional embodiment: the specific working mode of the prediction unit is as follows:

[0118] Set a time period M in advance, in days;

[0119] Obtain the average evaluation value M1 of the inverter in the past M days;

[0120] The frequency M2 of the inverter evaluation value being less than the evaluation value threshold value in the past M days is obtained; it should be noted that, in this embodiment, the value of M2 is the time of abnormal evaluation value in the past M days divided by the total time in M ​​days;

[0121] Obtain the difference M3 between the time when the inverter evaluation value was less than the evaluation value threshold in the past M days and the current time;

[0122] The predicted evaluation value S (abnormal|N) of the inverter tomorrow is obtained by calculation, wherein δ1, δ2 and δ3 are preset proportional coefficients, e is the base of the natural logarithm, and ε is the historical maintenance impact factor; it should be noted that in this embodiment, the value of e is 2.71828, and the values ​​of δ1, δ2 and δ3 are 0.283, 0.387 and 0.340;

[0123] A threshold value of the predicted evaluation value is set. If the predicted evaluation value S (abnormal|N) of the inverter tomorrow is greater than the threshold value of the predicted evaluation value, it is judged that the state of the inverter tomorrow will be abnormal, and a signal is transmitted to the mobile terminal of the maintenance personnel. It should be noted that by combining the historical data and maintenance data of the inverter to predict the future working state of the inverter, maintenance can be performed in advance when the inverter is about to have working problems, thereby avoiding losses to the photovoltaic power station caused by problems with the inverter.

[0124] As an optional embodiment: the historical maintenance impact factor ε is obtained as follows:

[0125] Obtain the historical working time value D1 of the inverter; it should be noted that, in this embodiment, the value of the historical working time value D1 is obtained by comparing the accumulated operating time of the inverter since installation with the standard working time of the inverter by the staff, and is any integer from 1 to 10;

[0126] The maintenance data value D2 of the inverter is obtained; it should be noted that, in this embodiment, the value of the maintenance data value D2 is obtained by the staff referring to the maintenance record of the inverter, including the time and type of cleaning and replacement of components, and is any integer from 1 to 10;

[0127] The predicted environmental data value D3 of the inverter is obtained; it should be noted that, in this embodiment, the value of the historical working time value D1 depends on the staff's reference to the weather forecast for environmental factors that may affect the performance of the inverter tomorrow, such as temperature, humidity, and irradiance, which is any integer from 1 to 10;

[0128] According to the formula ε=(D1*θ1)+(D2*θ2)+(D3*θ3), the historical maintenance impact factor ε is obtained, where θ1, θ2 and θ3 are preset proportional coefficients. It should be noted that in this embodiment, the values ​​of θ1, θ2 and θ3 can be 0.283, 0.389 and 0.328; it should also be noted that during the prediction process, the inverter will also be affected by the working time or maintenance data and environmental data of the historical inverter. By comprehensively calculating this impact value and combining it with the predicted data, the prediction result can be simulated more accurately.

[0129] As an optional embodiment: the specific working mode of the dust accumulation assessment module is as follows:

[0130] It should be noted that the loss calculation of traditional photovoltaic power stations does not analyze the cause of the loss, but only calculates the total power loss of the photovoltaic power station. In addition, during the operation and maintenance of photovoltaic power stations, there is a lack of scientific and reasonable data evaluation methods for measuring the degree of dust accumulation. The main method is to maintain standard battery panels or standard strings in the power station, clean the standard panels or strings separately, and then make an estimate based on the operation data after cleaning and the operation data of equipment in other uncleaned areas. When different areas in the station are affected by special conditions such as roads and emissions and have different degrees of local dust accumulation, it will be difficult to make an accurate estimate and evaluation.

[0131] This technical solution uses the inverter as a unit to decompose and calculate the power loss caused by dust accumulation in the photovoltaic power station. Based on the actual power generation of the photovoltaic inverter and the radiation, temperature and other data measured on site, it can automatically evaluate the dust accumulation in different areas of the power station and accurately reflect the dust accumulation in the photovoltaic power station and the power generation loss affected by dust.

[0132] Obtain the theoretical power generation energy E1 of the photovoltaic power station where the inverter is located;

[0133] The actual power generation energy E2 of the photovoltaic power station where the inverter is located in the current time period is obtained; it should be noted that the actual power generation energy E2 is obtained by measuring the actual electric energy output by the inverter;

[0134] According to the formula The energy loss factor g of the photovoltaic power station where the inverter is located in the current time period is obtained.

[0135] As an optional embodiment: the specific method of obtaining the theoretical power generation energy E1 of the photovoltaic power station is as follows:

[0136] Obtain the total area A of the photovoltaic module where the inverter is located;

[0137] Obtain the standard power generation efficiency η of the photovoltaic module where the inverter is located;

[0138] Obtain the irradiance I1 of the photovoltaic module where the inverter is located;

[0139] According to the formula E1=A*η*I1*t, the theoretical power generation energy E1 of the photovoltaic power station is obtained, where t is time in hours.

[0140] As an optional embodiment: the dust accumulation evaluation module is also used to evaluate a single area where the inverter is located, to evaluate whether the dust accumulation energy loss in the area where the inverter is located is abnormal, specifically:

[0141] Take the inverter as the center and draw a circle with a preset radius to obtain the unit area of ​​the inverter and the area A of the unit area. i ;

[0142] According to the formula E3 = A i *η*I1*t*K1, the dust loss energy E3 in the inverter area is obtained, where A i represents the unit area where the i-th inverter is located, and K1 is the dust accumulation coefficient of the area;

[0143] A dust accumulation loss energy threshold is set in advance. If the dust accumulation loss energy E3 in the area where the inverter is located is greater than the dust accumulation loss energy threshold, a signal is generated and transmitted to the mobile phone terminal of the maintenance personnel.

[0144] As an optional embodiment: the specific method of obtaining the dust accumulation coefficient K1 is as follows:

[0145] According to the formula The dust accumulation coefficient K1 is obtained by calculation. K1 represents the dust shielding rate. It should be noted that through the above steps, the power loss due to dust accumulation and the improved benefits after cleaning can be quantitatively calculated to guide the effective cleaning of photovoltaic stations and avoid blind cleaning decisions. The dust accumulation coefficient K1 can be determined by the dust shielding rate and the cleaning effect, reflecting the contribution of cleaning to the improvement of power generation.

[0146] As an optional embodiment: the inverter cleaning day is specifically obtained in the following manner:

[0147] Select a reference time period, clean the solar inverter in the absence of precipitation, observe the change in power generation after cleaning, record the power generation data every day, analyze the power generation data, and determine the time when the benefits of the cleaning effect reach the critical point, that is, after r days after cleaning, the power generation no longer increases significantly, and r days is defined as the cleaning failure period;

[0148] The cumulative power increase after cleaning is obtained, and the cumulative power increase is divided by the cleaning cost to obtain the cleaning benefit value; it should be noted that two groups of inverters are selected, one group is the inverters cleaned daily (group a), and the other group is the inverters not cleaned this time (group b).

[0149] A linear relationship is established based on the daily power generation of inverter group a and the daily power generation of inverter group b using a linear fitting method.

[0150] The theoretical power generation of inverter group B in the cleaning state is calculated through the linear fitting coefficient.

[0151] Calculate the difference between the actual power generation of the inverters in group b before cleaning and the theoretical power generation in the cleaning state, and obtain the cumulative power increase after cleaning;

[0152] It should also be noted that cleaning costs include labor, water resources, equipment depreciation, etc.;

[0153] t clean =argmax(cleaning benefit value(1-rainfall impact factor)); It should be noted that the rainfall impact factor can be a coefficient adjusted according to precipitation to reduce the impact of rainfall on cleaning benefits, which is obtained by staff based on electrical forecasts;

[0154] If the cleaning benefit value is greater than 1 and t clean If it is less than the cleaning failure period r, the inverter cleaning day is equal to t clean , if t clean If it is greater than the cleaning failure period r, then the inverter cleaning day is r days after the last cleaning.

[0155] How it works

[0156] Diagnosing PV panels through data analysis can not only effectively identify inefficient equipment operation, but also accurately determine the type of fault based on data characteristics, and quantitatively evaluate the operating efficiency and health of the power station.

[0157] Taking the inverter as the unit, the power loss caused by dust accumulation in the photovoltaic power station is decomposed and calculated. Based on the actual power generation power of the photovoltaic inverter and the irradiation, temperature and other data measured at the power station site, the dust accumulation in different areas of the power station can be automatically evaluated separately, accurately reflecting the dust accumulation in the photovoltaic power station and the power generation loss affected by dust.

[0158] The above are only preferred implementations of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technical staff in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of this template.

Claims

1. An inverter health assessment system, It is characterized in that include: A data collection module, wherein the data collection module is used to collect real-time data of the inverter; A dust accumulation assessment module, which is used to decompose and calculate the loss amount of the photovoltaic power station caused by dust accumulation based on the inverter; A health assessment module, the health assessment module comprising an assessment unit and a prediction unit, the assessment unit being used to assess the working performance of the inverter in real time according to the data collected by the data collection module and the data of the dust accumulation assessment module; First, the characteristic factors of the inverter working performance are obtained, specifically: Get the real-time power factor of the inverter ; Get the real-time efficiency value of the inverter ; Get the real-time temperature anomaly of the inverter ; Get the real-time current imbalance value of the inverter ; Get the real-time voltage imbalance value of the inverter ; The acquired influencing factors are standardized, specifically: According to the formula , calculate and obtain the standardized characteristic factors ,in Indicates characteristic factors, Indicates The average value of the characteristic factors, Indicates The standard deviation of the characteristic factors; According to the formula , get the real-time evaluation value of the inverter ,in is the number of characteristic factors, is the preset weight coefficient corresponding to each characteristic factor, where is the energy loss factor of the photovoltaic power station where the inverter is located in the current time period; A threshold value of the real-time evaluation value is set in advance, and the real-time evaluation value A of the inverter and the threshold value of the real-time evaluation value are connected; If the real-time evaluation value of the inverter If the real-time evaluation value is greater than the threshold, the inverter is judged to be in an abnormal working state, and an alarm signal is generated and transmitted to the maintenance personnel’s mobile terminal. is less than or equal to the threshold of the real-time evaluation value, a signal is generated and transmitted to the prediction unit; The prediction unit is used to predict the degradation of inverter performance over time based on the real-time data collected by the data collection module; The health assessment module further includes a diagnosis unit, which is used to identify the fault of the inverter after the assessment unit determines that the working state of the inverter is abnormal, specifically: The inverter fault types are set in advance as electrical fault, thermal fault and mechanical fault, and the three types are marked as fault type 1, fault type 2 and fault type 3 respectively; According to the formula , calculate and obtain the given input When the inverter is in the fault category The probability value ,in, Indicates input vector, Corresponds to the fault type Network output , e is the base of the natural logarithm, and 3 is because the number of fault types is 3; Get the probability of fault type 1 ; Get the probability of fault type 2 ; Get the probability of fault type 3 ; The probability of failure type 1 , the probability of failure type 2 and the probability of failure type 3 After comparison, the fault type with the highest probability is determined as the current fault type of the inverter, and the signal is transmitted to the mobile terminal of the maintenance personnel; The said Input vector Represents the data characteristic vector input to the diagnostic unit, and the specific acquisition method is as follows: For fault type 1 electrical fault, the input data feature vector is ; For fault type 2 thermal fault, the input data feature vector is ; For fault type 3 mechanical fault, the input data feature vector is ; According to the formula: ; ； ； Get and ,in and and , and All are preset proportional factors; The specific working method of the prediction unit is as follows: Set a time period in advance , the unit is day; Get the past Average rating of inverters for the day ; Get the past Frequency of inverter evaluation value being less than the evaluation value threshold ; Get the past The difference between the time when the inverter evaluation value is less than the evaluation value threshold and the current time ; , calculate and obtain the predicted evaluation value of the inverter tomorrow ,in is the preset proportionality factor, e is the base of the natural logarithm, Maintaining impact factors for history; A threshold value for the predicted evaluation value is set. If the predicted evaluation value of the inverter tomorrow If the value is greater than the threshold of the predicted evaluation value, it is determined that the inverter status will be abnormal tomorrow, and a signal is transmitted to the maintenance personnel's mobile terminal.

2. The inverter health assessment system according to claim 1, It is characterized in that The historical maintenance impact factor The way to obtain is as follows: Get the historical working time value of the inverter ; Get the maintenance data value of the inverter ; Get the predicted environmental data value of the inverter ; According to the formula , get the historical maintenance impact factor ,in is the preset scaling factor.

3. The inverter health assessment system according to claim 2, It is characterized in that The specific working mode of the dust accumulation assessment module is as follows: Get the theoretical power generation energy of the photovoltaic power station where the inverter is located ; Get the actual power generation energy of the photovoltaic power station where the inverter is located in the current time period ; According to the formula , obtain the energy loss factor of the photovoltaic power station where the inverter is located in the current time period .

4. The inverter health assessment system according to claim 3, It is characterized in that Theoretical power generation energy of the photovoltaic power station The specific way to obtain is as follows: Get the total area of ​​the photovoltaic panels where the inverter is located ; Get the standard power generation efficiency of the photovoltaic module where the inverter is located ; Get the irradiance of the photovoltaic module where the inverter is located ; According to the formula , get the theoretical power generation energy of the photovoltaic power station ,in The time is in hours.

5. The inverter health assessment system according to claim 4, It is characterized in that The dust accumulation evaluation module is also used to evaluate a single area where the inverter is located, to evaluate whether the dust accumulation energy loss in the area where the inverter is located is abnormal, specifically: Take the inverter as the center and draw a circle with a preset radius to obtain the unit area of ​​the inverter and the area of ​​the unit area. ; According to the formula , obtain the dust loss energy in the area where the inverter is located ,in Indicates The unit area where each inverter is located, is the dust accumulation coefficient of the area; A dust accumulation energy loss threshold is set in advance. If the dust accumulation energy loss in the area where the inverter is located is If the energy loss due to dust accumulation is greater than the threshold, a signal is generated and transmitted to the maintenance personnel's mobile phone terminal.

6. The inverter health assessment system according to claim 5, It is characterized in that The dust accumulation coefficient The specific way to obtain is as follows: According to the formula , calculate and obtain the dust accumulation coefficient ,in Indicates the dust shielding rate.

7. The inverter health assessment system according to claim 6, It is characterized in that The specific method for obtaining the inverter cleaning day is as follows: Select a reference time period, clean the solar inverter in the absence of precipitation, observe the change in power generation after cleaning, record the power generation data every day, analyze the power generation data, and determine the time when the benefits of the cleaning effect reach the critical point, that is, after r days after cleaning, the power generation no longer increases significantly, and r days is defined as the cleaning failure period; The cumulative power increase after cleaning is obtained, and the cleaning benefit value is obtained by dividing the cumulative power increase by the cleaning cost; ； If the cleaning benefit value is greater than 1, and is less than the cleaning failure period r, then the inverter cleaning day is equal to ,if If it is greater than the cleaning failure period r, then the inverter cleaning day is r days after the last cleaning.

Images