Photovoltaic inverter state identification system and method based on four fusible terminals

Abstract

The invention discloses a photovoltaic inverter state recognition system and method based on a four-fusible terminal, and relates to the technical field of inverter data analysis, and the system comprises an inverter state recognition module, a to-be-referenced data collection module, a reference data selection module, and a state prediction and adjustment module. The inverter state identification module identifies and monitors the state of the distributed photovoltaic inverter, and performs early warning processing when monitoring that the state of the inverter is abnormal. Historical operation data of a to-be-referenced inverter is acquired through a to-be-referenced data acquisition module, bearing load information and historical optimization information of a current inverter and the to-be-referenced inverter are acquired respectively, to-be-referenced data are modeled through a reference data selection module, and target reference data are screened out according to model matching. And the state prediction and adjustment module performs state prediction and adjustment processing on the current inverter according to the target reference data, so that the accuracy of predicting the service life of the distributed photovoltaic inverter is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of inverter data analysis, and in particular to a photovoltaic inverter state identification system and method based on four fusion terminals. Background Art

[0002] Distributed photovoltaic inverters are a key device in solar power generation systems. Their working principle is to convert the direct current generated by each solar panel into alternating current, thereby realizing the utilization of solar power, maximizing the power generation efficiency of solar cells and improving the overall efficiency of the entire power generation system. Real-time monitoring of the status data of distributed photovoltaic inverters can help timely discover and handle abnormal operating conditions of the inverter to reduce the risk of inverter failure. It can also analyze and predict the service life of the inverter through the monitored data. The data monitoring of distributed photovoltaic inverters can be realized through the four-integrated terminal, which refers to a terminal device with observable, controllable, adjustable and measurable functions; By accurately evaluating the service life of the inverter, regular inspections and replacement of key components can be arranged before the inverter approaches the end of its expected life, avoiding service interruptions due to sudden failures. When predicting the service life of the inverter, the historical status data of some inverter devices of the same model can be used as a reference to predict the service life of the current device to improve the accuracy of the prediction results. However, in the prior art, only the historical data of devices of the same model are selected as a reference without pre-screening. Since the working environments of different inverters are different, the accuracy of the service life prediction results cannot be guaranteed, affecting the subsequent inspection work. Summary of the invention

[0003] The object of the present invention is to provide a photovoltaic inverter state identification system and method based on four fusion terminals to solve the problems raised in the prior art.

[0004] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: A photovoltaic inverter state identification system based on four fusion terminals, the system comprising an inverter state identification module, a reference data acquisition module, a reference data selection module and a state prediction and adjustment module; The inverter status recognition module is used to perform status recognition and monitoring of the distributed photovoltaic inverter, and an early warning process is performed when an abnormal inverter status is detected; The referenced data acquisition module is used to collect historical operation data of the referenced inverter and respectively collect load information and historical optimization information of the current inverter and the referenced inverter; The reference data selection module performs modeling on the reference data, establishes a load analysis model and an optimization analysis model for the current inverter and the reference inverter, respectively, and selects the target reference data based on model matching; The state prediction and adjustment module performs state prediction and adjustment processing on the current inverter according to the target reference data.

[0005] Preferably, the inverter state identification module includes a conversion efficiency monitoring unit and a used time monitoring unit; The conversion efficiency monitoring unit is used to monitor the conversion efficiency of the current inverter during operation by using the four-integrated terminal, set a conversion efficiency reduction threshold, and send a warning signal when it is detected that the conversion efficiency reduction value of the current inverter exceeds the threshold. The current inverter refers to the inverter whose service life needs to be predicted at present. The conversion efficiency of the inverter is obtained by dividing the AC output efficiency of the monitored inverter during operation by the DC input efficiency. The conversion efficiency reduction value is equal to the conversion efficiency after the reduction minus the conversion efficiency before the reduction. The four "can" in the four-integrated terminal refer to observable, measurable, adjustable and controllable. Among them, observable refers to the panoramic visualization of low-voltage distributed photovoltaic statistical data, operating status, regulation and control, and abnormal alarm; measurable refers to the minute-level collection of low-voltage distributed photovoltaic user data, the 15-minute load data collection of all low-voltage distributed photovoltaic users, and the prediction of low-voltage distributed photovoltaic power generation load; adjustable refers to the flexible adjustment of low-voltage distributed photovoltaic power and voltage; controllable refers to the rigid controllability of all low-voltage distributed photovoltaic users. Using the four-integrated terminals to monitor the status of distributed photovoltaic inverters can effectively improve the timeliness and accuracy of abnormal data detection and implementation of treatment measures, and can effectively ensure the reliable operation of photovoltaic inverters.

[0006] The used time of the current inverter is obtained through the used time monitoring unit, and the used time of the current inverter is obtained as t.

[0007] Preferably, the reference data acquisition module includes an operation data acquisition unit, a load acquisition unit and an optimization information acquisition unit; The operation data collection unit collects historical operation data of an inverter that is the same model as the current inverter and has been discontinued, wherein the historical operation data refers to the conversion efficiency data of the inverter when it was previously in operation, and the inverter that is the same model as the current inverter and has been discontinued is used as a reference inverter, where discontinued means that the inverter has been discontinued due to expiration of its service life; The load acquisition unit is used to acquire the load carried by the reference inverter before the usage time reaches t, and the usage time of the reference inverter is counted when the corresponding load is collected. The load acquisition unit is also used to collect the load carried by the current inverter before the usage time reaches t, and the usage time of the current inverter is counted when the corresponding load is collected; The optimization information acquisition unit is used to obtain the number of times the reference inverter is optimized before the usage time reaches t, and the usage time of the reference inverter is counted each time it is optimized. The optimization information acquisition unit is also used to collect the number of times the current inverter is optimized before the usage time reaches t, and the usage time of the current inverter is counted each time it is optimized. The optimization of the inverter refers to the optimization of the components of the inverter, for example: optimizing the DC filter, the reactor, the transformer and other components. After optimizing these components, the conversion efficiency of the inverter can be improved.

[0008] Preferably, the reference data selection module includes a reference data integration unit, an environmental data comparison unit and a target data screening unit; The referenced data integration unit integrates the referenced data, establishes a historical load analysis model and an optimization analysis model of the referenced inverter and uses them as a first load analysis model and a first optimization analysis model respectively, and establishes a historical load analysis model and an optimization analysis model of the current inverter and uses them as a second load analysis model and a second optimization analysis model respectively; By means of the environmental data comparison unit, according to the offset comparison between the models of the reference inverter and the current inverter, a working environment matching coefficient between the reference inverter and the current inverter is analyzed; The target data screening unit screens out historical operating data of the target inverter as target reference data.

[0009] Preferably, the state prediction and adjustment module includes a service life prediction unit and an optimization replacement processing unit; Predicting the service life of the current inverter by referring to the target reference data through the service life prediction unit; The optimization and replacement processing unit is used to optimize or replace the inverter before the service life of the current inverter reaches the predicted value.

[0010] Preferably, the load set carried by a random reference inverter before the usage time reaches t is collected as H={H1,H2,...H m}, the set of used time of the reference inverter when the corresponding load is collected is T={T1,T2,...T m}, the corresponding usage time of the reference inverter each time it is optimized before the usage time reaches t is F={F1,F2,...F g}, g represents the number of times the corresponding reference inverter is optimized before the usage time reaches t, and the load set carried by the current inverter before the usage time reaches t is h={h1,h2,...h n}, the usage time set of the current inverter when the corresponding load is collected is T ’ ={T1 ’ ,T2 ’ ,...T n ’}, the used time of the current inverter each time it is optimized before the used time reaches t is f={f1,f2,...f k}, where k represents the number of times the current inverter is optimized before the usage time reaches t, m represents the number of times a random reference inverter carries a load, and n represents the number of times the current inverter carries a load.

[0011] Preferably, the reference data is integrated into the first, second, third and fourth training samples, and the first training sample is {(H1, T1), (H2, T2), ... (H m , T m )}, the second training sample is {(h1,T1 ’ ), (h2,T2 ’ ),...(h n , T n ’ )}, the third training sample is {(1, F1), (2, F2), ... (g, F g )}, the fourth training sample is {(1, f1), (2, f2), ... (k, f k )}, after performing a straight line fitting on the first training sample, the first load analysis model is established: , a1 and b1 represent the bias and intercept of the first load analysis model respectively, x1 represents the variable representing the load in the first load analysis model, y1 represents the variable representing the used time in the first load analysis model, and the second load analysis model is established after linear fitting of the second training sample: , a2 and b2 represent the bias and intercept of the second load analysis model respectively, x2 represents the variable representing the load in the second load analysis model, y2 represents the variable representing the used time in the second load analysis model, and the first optimization analysis model is established after linear fitting of the third training sample: , C1 and D1 represent the bias and intercept of the first optimization analysis model respectively, X1 represents the variable in the first optimization analysis model indicating the number of times the inverter is optimized, Y1 represents the variable in the first optimization analysis model indicating the used time when the inverter is optimized, and the second optimization analysis model is established after linear fitting of the fourth training sample: , C2 and D2 represent the bias and intercept of the second optimization analysis model respectively, X2 represents the variable in the second optimization analysis model indicating the number of times the inverter is optimized, and Y2 represents the variable in the second optimization analysis model indicating the used time when the inverter is optimized; When predicting the service life of an inverter, the accuracy of the service life prediction results can be improved by referring to the historical parameters of equipment of the same model. Considering that the working environment of the inverter will be different, although the inverters are of the same model, there may be large differences in the working environment between some inverters and the current inverter, which in turn leads to large differences in historical operating parameters. For example, there are large differences in the load data carried during operation, and some inverters may optimize components during use, and the conversion efficiency after optimization also fluctuates. Therefore, by collecting the historical usage data of the inverter to be referenced, the historical usage data is modeled and integrated, and the data presented by the model is compared to select the inverter data to be referenced whose working environment data matches the working environment data of the current inverter as the target reference data, and the remaining service life of the current inverter is predicted based on the target reference data, which effectively improves the accuracy of the prediction results of the inverter service life.

[0012] Preferably, the working environment matching coefficient W between a random reference inverter and the current inverter is calculated according to the following formula: W=log a (|a2-a1|+|C2-C1|),0 <a<1; Wherein, a represents the base of the log function, analyzes and compares the working environment matching coefficients between different reference inverters and the current inverter, selects the reference inverter with the highest working environment matching coefficient with the current inverter as the target inverter, and selects the historical operation data of the target inverter as the target reference data; The matching situation between the current inverter and the reference inverter is analyzed by comparing the model biases. The model bias can intuitively reflect the trend change between the working environment parameters of the two inverters, thus improving the referenceability of the selected target reference data.

[0013] Preferably, the conversion efficiency of the target inverter when it is running after the used time t is obtained, and the conversion efficiency of the target inverter when the used time reaches t is obtained. ’When the conversion efficiency decreases and the conversion efficiency decrease value exceeds the conversion efficiency decrease threshold set by the system, the remaining service life of the current inverter is predicted to be: t ’ -t.

[0014] Preferably, when the remaining service life of the current inverter reaches t ’ -t Before optimizing the inverter components or replacing the inverter.

[0015] Photovoltaic inverter state identification method based on four fusion terminals: the method comprises: S1: Perform status identification and monitoring on distributed photovoltaic inverters, and take early warning measures when abnormal inverter status is detected; S2: Collect historical operation data of the reference inverter and respectively collect load information and historical optimization information of the current inverter and the reference inverter; S3: Modeling the reference data, establishing a load analysis model and an optimization analysis model for the current inverter and the reference inverter respectively, and selecting target reference data based on model matching; S4: Predict and adjust the state of the current inverter according to the target reference data.

[0016] Compared with the prior art, the present invention has the following beneficial effects: The present invention can improve the accuracy of the service life prediction results by referring to the historical parameters of the equipment of the same model. Taking into account that the working environment of the inverter will be different, although the inverters are of the same model, there may be some inverters whose working environment is greatly different from that of the current inverter, which in turn leads to large differences in historical operating parameters. For example: there are large differences in the load data carried during operation, and some inverters may optimize components during use, and the conversion efficiency after optimization also fluctuates. Therefore, by collecting the historical usage data of the inverter to be referenced, the historical usage data is modeled and integrated, and the data presented by the model is compared to screen out the inverter data to be referenced whose working environment data matches the working environment data of the current inverter with a high degree of match as the target reference data, and the remaining service life of the current inverter is predicted based on the target reference data, thereby effectively improving the accuracy of the prediction results of the inverter service life. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a structural schematic diagram of a photovoltaic inverter state identification system based on four fusion terminals of the present invention; Figure 2 The present invention is a flow chart of a photovoltaic inverter state identification method based on four fusion terminals. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0019] Example 1: Figure 1 As shown, this embodiment provides a photovoltaic inverter state identification system based on four fusion terminals, and the system includes: an inverter state identification module, a reference data acquisition module, a reference data selection module and a state prediction and adjustment module; the inverter state identification module is used to perform state identification monitoring on the distributed photovoltaic inverter, and an early warning process is performed when the inverter state is abnormal; the reference data acquisition module is used to collect the historical operation data of the reference inverter and respectively collect the load information and historical optimization information of the current inverter and the reference inverter; the reference data is modeled through the reference data selection module, and a load analysis model and an optimization analysis model are respectively established for the current inverter and the reference inverter, and the target reference data is screened out according to the model matching; the state prediction and adjustment module is used to perform state prediction and adjustment processing on the current inverter according to the target reference data.

[0020] The inverter state identification module includes a conversion efficiency monitoring unit and a usage time monitoring unit; the conversion efficiency monitoring unit is used to monitor the conversion efficiency of the current inverter during operation using the four-integrated terminal, set a conversion efficiency decrease threshold, and send a warning signal when it is detected that the conversion efficiency decrease value of the current inverter exceeds the threshold. The current inverter refers to the inverter whose service life needs to be predicted at present. The conversion efficiency of the inverter is obtained by dividing the monitored AC output efficiency of the inverter during operation by the DC input efficiency. The conversion efficiency decrease value is equal to the conversion efficiency after the decrease minus the conversion efficiency before the decrease; the usage time of the current inverter is obtained through the usage time monitoring unit, and the usage time of the current inverter is obtained as t.

[0021] The reference data collection module includes an operation data collection unit, a load collection unit and an optimization information collection unit; the operation data collection unit is used to collect historical operation data of the inverter that has been discontinued and has the same model as the current inverter, and the historical operation data refers to the conversion efficiency data of the inverter in the past operation. The inverter that has been discontinued and has the same model as the current inverter is used as the reference inverter. Discontinued means that the inverter has been discontinued due to expiration of its service life; the load carried by the reference inverter before the usage time reaches t is obtained through the load collection unit, and the usage time of the reference inverter when the corresponding load is collected is counted. The load collection unit is also used to collect Collect the load carried by the current inverter before the usage time reaches t, and count the usage time of the current inverter when the corresponding load is collected; obtain the number of times the reference inverter is optimized before the usage time reaches t through the optimization information collection unit, and count the usage time of the reference inverter each time it is optimized. The optimization information collection unit is also used to collect the number of times the current inverter is optimized before the usage time reaches t, and count the usage time of the current inverter each time it is optimized. The optimization of the inverter refers to the optimization of the components of the inverter, for example: optimizing the DC filter, reactor, transformer and other components. After optimizing these components, the conversion efficiency of the inverter can be improved.

[0022] The reference data selection module includes a reference data integration unit, an environmental data comparison unit and a target data screening unit; the reference data integration unit integrates the reference data, establishes a historical load analysis model and an optimization analysis model of the reference inverter and uses them as the first load analysis model and the first optimization analysis model respectively, and establishes a historical load analysis model and an optimization analysis model of the current inverter and uses them as the second load analysis model and the second optimization analysis model respectively; the environmental data comparison unit analyzes the working environment matching coefficient between the reference inverter and the current inverter based on the offset comparison between the models of the reference inverter and the current inverter; and the target data screening unit screens out the historical operation data of the target inverter as the target reference data.

[0023] The state prediction and adjustment module includes a service life prediction unit and an optimization replacement processing unit; the service life prediction unit predicts the service life of the current inverter with reference to the target reference data; the optimization replacement processing unit optimizes or replaces the inverter before the service life of the current inverter reaches the predicted value.

[0024] The load set carried by a random reference inverter before the usage time reaches t is collected as H={H1,H2,...H m}, the set of used time of the reference inverter when the corresponding load is collected is T={T1,T2,...Tm}, the corresponding usage time of the reference inverter each time it is optimized before the usage time reaches t is F={F1,F2,...F g}, g represents the number of times the corresponding reference inverter is optimized before the usage time reaches t, and the load set carried by the current inverter before the usage time reaches t is h={h1,h2,...h n}, the usage time set of the current inverter when the corresponding load is collected is T ’ ={T1 ’ ,T2 ’ ,...T n ’}, the used time of the current inverter each time it is optimized before the used time reaches t is f={f1,f2,...f k}, where k represents the number of times the current inverter is optimized before the usage time reaches t, m represents the number of times a random reference inverter carries a load, and n represents the number of times the current inverter carries a load.

[0025] The reference data is integrated into the first, second, third and fourth training samples, and the first training sample is {(H1, T1), (H2, T2), ... (H m , T m )}, the second training sample is {(h1,T1 ’ ), (h2,T2 ’ ),...(h n , T n ’ )}, the third training sample is {(1, F1), (2, F2), ... (g, F g )}, the fourth training sample is {(1, f1), (2, f2), ... (k, f k )}, after performing a straight line fitting on the first training sample, the first load analysis model is established: , a1 and b1 represent the bias and intercept of the first load analysis model respectively, x1 represents the variable representing the load in the first load analysis model, y1 represents the variable representing the used time in the first load analysis model, where, , , and establish the second load analysis model after performing a straight line fitting on the second training sample: , a2 and b2 represent the bias and intercept of the second load analysis model respectively, x2 represents the variable representing the load in the second load analysis model, y2 represents the variable representing the used time in the second load analysis model, and the first optimization analysis model is established after linear fitting of the third training sample: , C1 and D1 represent the bias and intercept of the first optimization analysis model respectively, X1 represents the variable in the first optimization analysis model indicating the number of times the inverter is optimized, Y1 represents the variable in the first optimization analysis model indicating the used time when the inverter is optimized, and the second optimization analysis model is established after linear fitting of the fourth training sample: , C2 and D2 represent the bias and intercept of the second optimization analysis model respectively, X2 represents the variable in the second optimization analysis model indicating the number of times the inverter is optimized, and Y2 represents the variable in the second optimization analysis model indicating the used time when the inverter is optimized. The bias and intercept solution methods of each model are the same.

[0026] The working environment matching coefficient W between a random reference inverter and the current inverter is calculated according to the following formula: W=log a (|a2-a1|+|C2-C1|),0 <a<1; Wherein, a represents the base of the log function, analyzes and compares the working environment matching coefficients between different reference inverters and the current inverter, selects the reference inverter with the highest working environment matching coefficient with the current inverter as the target inverter, and selects the historical operation data of the target inverter as the target reference data.

[0027] Get the conversion efficiency of the target inverter when it is running after the used time t, and get the conversion efficiency of the target inverter when the used time reaches t ’ When the conversion efficiency decreases and the conversion efficiency decrease value exceeds the conversion efficiency decrease threshold set by the system, the remaining service life of the current inverter is predicted to be: t ’ -t, when the remaining service life of the current inverter reaches t ’ -t Before optimizing the inverter components or replacing the inverter.

[0028] Example 2: Figure 2 As shown, this embodiment provides a photovoltaic inverter state identification method based on four fusion terminals, the method comprising: S1: Perform status identification and monitoring on distributed photovoltaic inverters, and take early warning measures when abnormal inverter status is detected; S2: Collect historical operation data of the reference inverter and respectively collect load information and historical optimization information of the current inverter and the reference inverter; S3: Modeling the reference data, establishing a load analysis model and an optimization analysis model for the current inverter and the reference inverter respectively, and selecting target reference data based on model matching; S4: predicting and adjusting the state of the current inverter according to the target reference data; For example: a1 in the first load analysis model established for three random reference inverters is obtained to be 0.25, 0.20 and 0.12 respectively, a2 in the second load analysis model established for the current inverter is 0.22, C1 in the first optimization analysis model established for the corresponding three reference inverters is 3.21, 4.50 and 3.56 respectively, C2 in the second optimization analysis model established for the current inverter is 3.54, the base of the log function is set to 0.2, and the working environment matching coefficients between the three reference inverters and the current inverter are calculated to be 0.63, 0.01 and 1.32 respectively. The reference inverter with the highest working environment matching coefficient with the current inverter is the third inverter. The third inverter is used as the target inverter, and the conversion efficiency of the target inverter when it is running after 15 years of use is obtained. It is obtained that the conversion efficiency of the target inverter decreases when it has been used for 22 years and the conversion efficiency decrease value exceeds the conversion efficiency decrease threshold set by the system. The remaining service life of the current inverter is predicted to be: 7 years.

[0029] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A photovoltaic inverter status identification system based on four fusion terminals, characterized by: It includes an inverter state identification module, a reference data acquisition module, a reference data selection module and a state prediction and adjustment module; The inverter status recognition module is used to perform status recognition and monitoring of the distributed photovoltaic inverter, and an early warning process is performed when an abnormal inverter status is detected; The referenced data acquisition module is used to collect historical operation data of the referenced inverter and respectively collect load information and historical optimization information of the current inverter and the referenced inverter; The reference data selection module performs modeling on the reference data, establishes a load analysis model and an optimization analysis model for the current inverter and the reference inverter, respectively, and selects the target reference data based on model matching; The state prediction and adjustment module performs state prediction and adjustment processing on the current inverter according to the target reference data.

2. The photovoltaic inverter state identification system based on four fusion terminals according to claim 1 is characterized in that: The inverter state identification module includes a conversion efficiency monitoring unit and a used time monitoring unit; The conversion efficiency monitoring unit is used to monitor the conversion efficiency of the current inverter during operation using the four fusion terminals, set a conversion efficiency drop threshold, and send an early warning signal when it is detected that the conversion efficiency drop value of the current inverter exceeds the threshold; The used time of the current inverter is obtained through the used time monitoring unit, and the used time of the current inverter is obtained as t.

3. The photovoltaic inverter state identification system based on four fusion terminals according to claim 2 is characterized in that: The reference data acquisition module includes an operation data acquisition unit, a load acquisition unit and an optimization information acquisition unit; The operation data collection unit collects historical operation data of an inverter that is the same model as the current inverter and has been discontinued, wherein the historical operation data refers to the conversion efficiency data of the inverter when it was previously in operation, and the inverter that is the same model as the current inverter and has been discontinued is used as a reference inverter; The load acquisition unit is used to acquire the load carried by the reference inverter before the usage time reaches t, and the usage time of the reference inverter is counted when the corresponding load is collected. The load acquisition unit is also used to collect the load carried by the current inverter before the usage time reaches t, and the usage time of the current inverter is counted when the corresponding load is collected; The optimization information acquisition unit is used to obtain the number of times the reference inverter is optimized before the usage time reaches t, and the usage time of the reference inverter is counted each time it is optimized. The optimization information acquisition unit is also used to collect the number of times the current inverter is optimized before the usage time reaches t, and the usage time of the current inverter is counted each time it is optimized.

4. The photovoltaic inverter state identification system based on four fusion terminals according to claim 3 is characterized in that: The reference data selection module includes a reference data integration unit, an environmental data comparison unit and a target data screening unit; The referenced data integration unit integrates the referenced data, establishes a historical load analysis model and an optimization analysis model of the referenced inverter and uses them as a first load analysis model and a first optimization analysis model respectively, and establishes a historical load analysis model and an optimization analysis model of the current inverter and uses them as a second load analysis model and a second optimization analysis model respectively; By means of the environmental data comparison unit, according to the offset comparison between the models of the reference inverter and the current inverter, a working environment matching coefficient between the reference inverter and the current inverter is analyzed; The target data screening unit screens out historical operating data of the target inverter as target reference data.

5. The photovoltaic inverter state identification system based on four fusion terminals according to claim 4 is characterized in that: The state prediction and adjustment module includes a service life prediction unit and an optimization replacement processing unit; Predicting the service life of the current inverter by referring to the target reference data through the service life prediction unit; The optimization and replacement processing unit is used to optimize or replace the inverter before the service life of the current inverter reaches the predicted value.

6. The photovoltaic inverter state identification system based on four fusion terminals according to claim 5 is characterized in that: The load set carried by a random reference inverter before the usage time reaches t is collected as H={H1,H2,...H m }, the set of used time of the reference inverter when the corresponding load is collected is T={T1,T2,...T m }, the corresponding usage time of the reference inverter each time it is optimized before the usage time reaches t is F={F1,F2,...F g }, g represents the number of times the corresponding reference inverter is optimized before the usage time reaches t, and the load set carried by the current inverter before the usage time reaches t is h={h1,h2,...h n }, the usage time set of the current inverter when the corresponding load is collected is T ’ ={T1 ’ ,T2 ’ ,...T n ’ }, the used time of the current inverter each time it is optimized before the used time reaches t is f={f1,f2,...f k }, where k represents the number of times the current inverter is optimized before the usage time reaches t, m represents the number of times a random reference inverter carries a load, and n represents the number of times the current inverter carries a load.

7. The photovoltaic inverter state identification system based on four fusion terminals according to claim 6 is characterized in that: The reference data is integrated into the first, second, third and fourth training samples, and the first training sample is {(H1, T1), (H2, T2), ... (H m , T m )}, the second training sample is {(h1,T1 ’ ), (h2,T2 ’ ),...(h n , T n ’ )}, the third training sample is {(1, F1), (2, F2), ... (g, F g )}, the fourth training sample is {(1, f1), (2, f2), ... (k, f k )}, after performing a straight line fitting on the first training sample, the first load analysis model is established: , a1 and b1 represent the bias and intercept of the first load analysis model respectively, x1 represents the variable representing the load in the first load analysis model, y1 represents the variable representing the used time in the first load analysis model, and the second load analysis model is established after linear fitting of the second training sample: , a2 and b2 represent the bias and intercept of the second load analysis model respectively, x2 represents the variable representing the load in the second load analysis model, y2 represents the variable representing the used time in the second load analysis model, and the first optimization analysis model is established after linear fitting of the third training sample: , C1 and D1 represent the bias and intercept of the first optimization analysis model respectively, X1 represents the variable in the first optimization analysis model indicating the number of times the inverter is optimized, Y1 represents the variable in the first optimization analysis model indicating the used time when the inverter is optimized, and the second optimization analysis model is established after linear fitting of the fourth training sample: , C2 and D2 represent the bias and intercept of the second optimization analysis model respectively, X2 represents the variable in the second optimization analysis model indicating the number of times the inverter is optimized, and Y2 represents the variable in the second optimization analysis model indicating the used time when the inverter is optimized.

8. The photovoltaic inverter state identification system based on four fusion terminals according to claim 7 is characterized in that: The working environment matching coefficient W between a random reference inverter and the current inverter is calculated according to the following formula: W=log a (|a2-a1|+|C2-C1|),0<a<1; Wherein, a represents the base of the log function, analyzes and compares the working environment matching coefficients between different reference inverters and the current inverter, selects the reference inverter with the highest working environment matching coefficient with the current inverter as the target inverter, and selects the historical operation data of the target inverter as the target reference data.

9. The photovoltaic inverter state identification system based on four fusion terminals according to claim 8 is characterized in that: Get the conversion efficiency of the target inverter when it is running after the used time t, and get the conversion efficiency of the target inverter when the used time reaches t ’ When the conversion efficiency decreases and the conversion efficiency decrease value exceeds the conversion efficiency decrease threshold set by the system, the remaining service life of the current inverter is predicted to be: t ’ -t, when the remaining service life of the current inverter reaches t ’ -t Before optimizing the inverter components or replacing the inverter.

10. A photovoltaic inverter state identification method based on four fusion terminals, applied to a photovoltaic inverter state identification system based on four fusion terminals according to any one of claims 1 to 9, characterized in that: The method comprises: S1: Perform status identification and monitoring on distributed photovoltaic inverters, and take early warning measures when abnormal inverter status is detected; S2: Collect historical operation data of the reference inverter and respectively collect load information and historical optimization information of the current inverter and the reference inverter; S3: Modeling the reference data, establishing a load analysis model and an optimization analysis model for the current inverter and the reference inverter respectively, and selecting target reference data based on model matching; S4: Predict and adjust the state of the current inverter according to the target reference data.

Images