Method and device for testing coupling efficiency of photovoltaic grid-connected energy storage type inverter

Abstract

The invention discloses a method for testing the coupling efficiency of a photovoltaic grid-connected energy storage type inverter, and the method comprises the following steps: building a corresponding distributed hardware test topology based on a distributed multi-source cooperation unit, an edge calculation collection terminal, a full-working-condition environment simulation room and a monitoring upper computer; obtaining multi-dimensional original data under different working conditions, performing Kalman denoising and timestamp synchronization processing, and generating a time sequence aligned inverter multi-source test data set; constructing a corresponding inverter basic dimension parameter matrix based on the inverter multi-source test data set, and performing coupling weighting operation and tensor coupling normalization correction calculation to generate a multi-dimension coupling normalization correction weighting coefficient; and obtaining input and output power data of the inverter, carrying out coupling efficiency accounting, and carrying out coupling over-limit secondary correction and testing at the same time, so as to generate a coupling efficiency testing result of the photovoltaic grid-connected energy storage type inverter under the corresponding working condition. According to the invention, the accuracy of testing the coupling efficiency of the inverter can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of inverter testing, in particular, to a method and device for testing coupling efficiency of a photovoltaic grid-connected energy storage type inverter. BACKGROUND

[0002] With the large-scale application of photovoltaic energy storage technology, as the core equipment of "light energy conversion-energy storage scheduling-grid interaction", the energy conversion efficiency of the photovoltaic grid-connected energy storage type inverter needs to be evaluated in multiple dimensions combining "photovoltaic fluctuation, energy storage state, grid operating condition, environmental condition". However, the existing technologies mostly calculate the efficiency based on "photovoltaic power ratio weighting coefficient", but only consider the single dimension of photovoltaic, ignore the coupling influence of dynamic factors such as energy storage cycle attenuation, grid harmonics, component aging on the efficiency, although they cover part of the scenes of inverter testing, but none of them realize the dynamic coupling test of "photovoltaic-energy storage-grid-environment-component" in multiple dimensions, which is difficult to match the complex operating condition requirements of inverters in actual application, thereby reducing the accuracy of inverter coupling efficiency test. SUMMARY

[0003] In view of the deficiencies of the prior art, the present application provides a method for testing coupling efficiency of a photovoltaic grid-connected energy storage type inverter, comprising the following steps: Step S1: based on the distributed multi-source collaborative unit, the edge computing acquisition terminal, the full-condition environment simulation room and the monitoring upper computer, a corresponding distributed hardware test topology is built, and based on the distributed hardware test topology, multi-dimensional original data of the photovoltaic grid-connected energy storage type inverter under different operating conditions are obtained, including photovoltaic-energy storage power data, energy storage charging and discharging data, grid parameter data and temperature monitoring data, and the multi-dimensional original data is subjected to Kalman denoising and time stamp synchronization processing to generate time sequence aligned inverter multi-source test data set; Step S2: based on the inverter multi-source test data set, corresponding photovoltaic power ratio, energy storage SOC value, grid THD content and component temperature rise data are obtained, and a corresponding inverter basic dimension parameter matrix is constructed in combination with the acquired regional light wave intensity index; based on the inverter basic dimension parameter matrix, coupling weighting operation is carried out to generate inverter initial coupling weighting coefficient; Step S3: based on the inverter multi-source test data set, coupling dynamic correction factors are obtained, including energy storage cycle attenuation coefficient, grid harmonic influence coefficient and component aging coefficient, and based on the coupling dynamic correction factors, the inverter initial coupling weighting coefficient is subjected to tensor coupling normalization correction calculation to generate multi-dimensional coupling normalization correction weighting coefficient; Step S4: Obtain the inverter input and output power data, and based on the multi-dimensional coupling normalization correction weighting coefficient and the coupling dynamic correction factor, the inverter input and output power data is coupled to calculate the coupling efficiency value of the photovoltaic grid-connected energy storage type inverter under the current working condition; based on the coupling efficiency values obtained under different working conditions, a multi-dimensional coupling efficiency surface graph is generated, and according to the multi-dimensional coupling efficiency surface graph, coupling over-limit secondary correction and testing are performed to generate the coupling efficiency test results of the photovoltaic grid-connected energy storage type inverter under the corresponding working condition.

[0004] Further, the application also provides a photovoltaic grid-connected energy storage type inverter coupling efficiency testing device for executing the photovoltaic grid-connected energy storage type inverter coupling efficiency testing method as described above, which comprises a distributed multi-source collaborative unit, an edge computing collection terminal, a full working condition environment simulation room, a monitoring host computer and a data storage module. The distributed multi-source collaborative unit comprises a photovoltaic array simulator, an energy storage battery simulator, an intelligent power grid simulation unit and a distributed collaborative unit; the photovoltaic array simulator supports U-P / UI curve output and is used for simulating photovoltaic power under different light; the energy storage battery simulator supports SOC0%-100% dynamic adjustment and cycle attenuation simulation and is used for simulating the charge and discharge state of the energy storage system; the intelligent power grid simulation unit has voltage sag and harmonic injection functions and is used for simulating different power grid working conditions; the distributed collaborative unit serves as an intermediate node and realizes power closed-loop control of photovoltaic, energy storage and power grid; The edge computing collection terminal comprises a multi-channel power analyzer and a temperature monitoring module; the multi-channel power analyzer supports 8-channel voltage / current collection, the sampling rate is adjustable from 100Hz to 1kHz, and is used for obtaining corresponding input and output power data of the photovoltaic grid-connected energy storage type inverter; the temperature monitoring module adopts 128-point collection, the measurement range is-40℃-1000℃, and is used for collecting component temperature rise data; The full working condition environment simulation room comprises an altitude pressure adjustment module, an illumination intensity simulation module and a temperature control module; the altitude pressure adjustment module can cover the pressure conditions of different geographical scenes of plateau and plain; the illumination intensity simulation module can reproduce the illumination changes in different periods and different seasons; the temperature control module can realize temperature adjustment from-40℃ to 85℃, and the temperature fluctuation is controlled within ±1℃; The monitoring host computer communicates with each unit through LAN+4G backup dual link, can complete initial coupling coefficient calculation, dynamic correction factor integration and coupling efficiency operation, and supports finite state machine FSM control logic at the same time; The data storage module adopts a local cache + cloud backup dual storage architecture, and simultaneously integrates a data backtracking retrieval module, can support querying historical data according to a timestamp, a working condition parameter and an inverter model, and thus realizes full-link traceability of testing.

[0005] The application has the beneficial effects that: by constructing a distributed hardware test topology, integrating multi-source collaborative units, edge terminals and other devices, the limitations of relying on single test equipment in the prior art are broken through, and multi-dimensional original data such as photovoltaic- energy storage power, power grid parameters and the like can be comprehensively collected. At the same time, in view of the problems that the original data is easy to be disturbed and the time stamp is dislocated, Kalman noise elimination and time stamp synchronization processing are adopted, random noise in the test process is effectively filtered out, the different dimension data is accurately aligned in the time dimension, and a multi-source test data set with consistent time sequence is generated, which solves the problem of poor data quality in the prior art that only single-dimensional data is collected, and provides a complete and accurate data basis for subsequent multi-dimensional coupling analysis. Secondly, by extracting key parameters such as photovoltaic power ratio and energy storage SOC value based on the multi-source test data set, and introducing regional light wave intensity as a reference to construct a basic dimension parameter matrix, the limitations of the prior art that only uses "photovoltaic power ratio" as a single weighting basis are broken through, the initial coupling weighting coefficient is generated by coupling and weighting operation on the basic dimension parameter matrix, and the originally isolated dimension parameters are converted into weighted indexes with correlation, so that it is no longer limited to a single photovoltaic dimension, can better match the scene in which the inverter is jointly affected by multiple static factors in actual application, reduces the efficiency calculation deviation caused by single weight dimension, and further improves the accuracy of coupling efficiency test. Then, by extracting coupling dynamic correction factors such as energy storage cycle attenuation, power grid harmonic influence and component aging from the multi-source test data set, the dynamic factors completely ignored by the prior art are accurately captured, the analysis gap of the "component" dimension is filled, and the full-dimensional coverage of "photovoltaic- energy storage- power grid- environment- component" is realized. Through tensor coupling normalization correction calculation, the dynamic correction factor and the initial coupling weighting coefficient are deeply integrated, the initial coefficient is dynamically adjusted, the problem that the weight is fixed and cannot match the state change of the inverter in long-term operation due to the failure to consider dynamic factors in the prior art is solved, and the dynamic correction makes the weighting coefficient reflect the dynamic changes in actual operation such as energy storage attenuation and component aging in real time, so that the weight basis is more in line with the real working state of the inverter, avoiding the calculation error of fixed weight under dynamic working conditions, and greatly improving the timeliness and accuracy of the coupling weighting coefficient. Finally, by calculating the input and output power based on the multi-dimensional coupling normalization correction weighting coefficient, the dynamic weight and the actual power data are combined with the coupling dynamic correction factor, the coupling efficiency value under the current working condition can be calculated. By generating a multi-dimensional coupling efficiency surface graph, the efficiency distribution under different working conditions is intuitively presented, which facilitates quick identification of abnormal efficiency intervals; and through coupling over-limit quadratic correction and test, the calculation deviation under extreme working conditions is further eliminated, ensuring the stability of the efficiency result, which realizes the whole process test of multi-dimensional dynamic coupling of "photovoltaic- energy storage- power grid- environment- component", completely matches the complex working condition requirements of the inverter in actual application, and completely solves the problems of incomplete test scene coverage and insufficient consideration of dynamic factors in the prior art, thereby significantly improving the accuracy and reliability of the inverter coupling efficiency test. Attached Figure Description

[0006] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of the coupling efficiency test method for photovoltaic grid-connected energy storage inverters in this embodiment; Figure 2 for Figure 1 A detailed flowchart illustrating the implementation steps of step S1. Figure 3 This is a schematic diagram of the structure of the photovoltaic grid-connected energy storage inverter coupling efficiency testing device in this embodiment. Detailed Implementation

[0007] The following drawings disclose several embodiments of the present invention. For clarity, many practical details will be described in the following description. However, it should be understood that these practical details are not intended to limit the invention. That is, in some embodiments of the invention, these practical details are not essential. Furthermore, for the sake of simplicity, some conventional structures and components will be shown in the drawings in a simple schematic manner.

[0008] To further understand the invention's content, features, and effects, the following embodiments are provided, and detailed descriptions are given below in conjunction with the accompanying drawings: Reference Figure 1 , Figure 1 This is a flowchart of the coupling efficiency testing method for photovoltaic grid-connected energy storage inverters in this embodiment. The coupling efficiency testing method for photovoltaic grid-connected energy storage inverters in this embodiment includes the following steps: Step S1: Construct a corresponding distributed hardware test topology based on the distributed multi-source collaborative unit, edge computing acquisition terminal, full-condition environment simulation room and monitoring host computer, and acquire multi-dimensional raw data of photovoltaic grid-connected energy storage inverter under different operating conditions based on the distributed hardware test topology, including photovoltaic-energy storage power data, energy storage charging and discharging data, grid parameter data and temperature monitoring data, and perform Kalman noise reduction and timestamp synchronization processing on the multi-dimensional raw data to generate a time-aligned inverter multi-source test dataset; In the embodiment of the present application, by constructing a distributed hardware test topology, a distributed multi-source cooperative unit connects a photovoltaic array simulator, an energy storage battery simulator and an intelligent power grid simulation unit through a dedicated line, an edge computing collection terminal accesses the DC input, AC output power interface and IGBT, capacitor temperature monitoring point of the inverter through a shielded cable, a temperature control module (adjustment range -40℃-150℃) and a gas pressure control module (50-110kPa) are arranged inside the full working condition environment simulation room, and a monitoring host computer is connected with each unit through optical fiber and wireless dual channels. Under the condition of environmental temperature 28℃, air pressure 100kPa, illumination 750W / m 2 Working condition, collect photovoltaic-energy storage power data (photovoltaic output 35kW, energy storage discharge 8kW), energy storage charge-discharge data (current 32A, voltage 250V, SOC from 70% to 62%), power grid parameter data (voltage 380V, frequency 50Hz, THD 1.8%), and temperature monitoring data (IGBT temperature 58℃, capacitor temperature 42℃, environmental temperature 28℃). Perform Kalman filtering on the original data, compress the voltage data noise from ±0.6V to ±0.08V, control the temperature data fluctuation from ±1.2℃ to ±0.15℃, align all parameters according to 500μs timestamp, and generate an inverter multi-source test data set containing 10 groups of continuous time sequence records (each group has 300 synchronization points).

[0009] Step S2: Based on the inverter multi-source test data set, the corresponding photovoltaic power ratio, energy storage SOC value, power grid THD content and component temperature rise data are obtained, and the corresponding inverter basic dimension parameter matrix is constructed combined with the acquisition of the local illumination wave intensity index; based on the inverter basic dimension parameter matrix, coupling weighting operation is performed to generate an inverter initial coupling weighting coefficient; In the embodiment of the present application, by extracting photovoltaic power 35kW and energy storage charge-discharge power 8kW from the inverter multi-source test data set, the photovoltaic power ratio 35÷(35+8)=81.4% is calculated; the current capacity 82kWh and the rated capacity 100kWh of the energy storage are extracted, and the energy storage SOC value 82% is calculated; the total effective value of the harmonic voltage 6.84V and the fundamental voltage 380V of the power grid are extracted, and the power grid THD content 6.84÷380=1.8% is calculated; the IGBT temperature 58℃ and the environmental temperature 28℃ are extracted, and the component temperature rise 30K is calculated. The local illumination wave intensity index 260 (unit W / m 2 h), and the five parameters are used as row vectors, and a column vector is constructed every 2 hours from 9:00 to 17:00 as a time node, forming a 5-row and 5-column inverter basic dimension parameter matrix. The entropy weight method is used to assign weights, and the light wave intensity accounts for 32%, the photovoltaic power accounts for 24%, the energy storage SOC value accounts for 20%, the grid THD content accounts for 15%, and the component temperature rise accounts for 9%, and the initial coupling weight coefficient of the inverter is 0.88 by summing the product of the matrix element value and the corresponding weight.

[0010] Step S3: obtaining a coupling dynamic correction factor based on the inverter multi-source test data set, including an energy storage cycle attenuation coefficient, a grid harmonic influence coefficient and a component aging coefficient, and performing tensor coupling normalization correction calculation on the initial coupling weight coefficient of the inverter based on the coupling dynamic correction factor, to generate a multi-dimensional coupling normalization correction weight coefficient; In the embodiment of the application, the energy storage cumulative charge and discharge cycle is extracted from the inverter multi-source test data set for 150 times, the initial rated capacity is 100 kWh, the current actual capacity is 98.2 kWh, the energy storage cycle attenuation coefficient is calculated as 98.2÷100=0.982; the grid THD content is extracted as 1.8%, and the grid harmonic influence coefficient is calculated as 1.0-1.8x0.015=0.973 according to the preset rule (coefficient 1.0 when THD=0, coefficient decreasing by 0.015 for each increase of 1%); the cumulative temperature rise of IGBT is extracted as 450K, the rated life corresponds to the cumulative temperature rise of 8000K, and the component aging coefficient is calculated as 1-(450÷8000)=0.94375. The three coefficients are integrated into a coupling dynamic correction factor [0.982, 0.973, 0.94375], and the weights [0.25, 0.45, 0.3] are assigned, and the factor weight value is calculated as 0.982x0.25+0.973x0.45+0.94375x0.3≈0.2455+0.4379+0.2831≈0.9665. The initial coupling weight coefficient of the inverter is 0.88, and the value is subjected to tensor coupling operation 0.88x0.9665≈0.8505, and then divided by the weight sum (0.211+0.243+0.264+0.282=1.0) of each power test point (20%, 50%, 80%, 100% rated power) to generate a multi-dimensional coupling normalization correction weight coefficient 0.8505.

[0011] Step S4: obtaining inverter input and output power data, and performing coupling efficiency accounting on the inverter input and output power data based on the multi-dimensional coupling normalization correction weight coefficient and the coupling dynamic correction factor, to obtain the coupling efficiency value of the photovoltaic grid-connected energy storage type inverter under the current working condition; generating a multi-dimensional coupling efficiency surface graph based on the coupling efficiency values obtained under different working conditions, and performing coupling over-limit quadratic correction and testing according to the multi-dimensional coupling efficiency surface graph, to generate the coupling efficiency test result of the photovoltaic grid-connected energy storage type inverter under the corresponding working condition.

[0012] In the embodiment of the present application, by acquiring the inverter input power data (photovoltaic 35kW, energy storage discharge 8kW, total input power 43kW), output power data (grid-connected AC power 41kW), setting the integral period 3600 seconds, calculating the total input energy 43kWx1h=43kWh, the total output energy 41kWx1h=41kWh. Based on the multi-dimensional coupling normalization correction weighting coefficient 0.8505, the coupling efficiency calculation formula (41÷43) x 0.8505 x 0.982 x 0.973 x 0.94375 x 100%≈73.2%, the current working condition coupling efficiency value is obtained. Selecting 36 working conditions of photovoltaic power ratio 20%-100% (every 20% one gradient), energy storage SOC value 10%-90% (every 20% one gradient), grid THD content 1%-4% (every 1% one gradient), obtaining the coupling efficiency value of each working condition (78%-89%). Taking the photovoltaic power ratio as the X axis, the energy storage SOC value as the Y axis, and the coupling efficiency value as the Z axis, a multi-dimensional coupling efficiency surface graph is drawn. For the four working condition points in the surface graph with efficiency less than 80%, the grid harmonic influence coefficient weight is adjusted to 0.5 and recalculated, so that the efficiency is improved to more than 80%. Based on the corrected surface graph, the coupling efficiency value, deviation range and qualified identification of each working condition are output, and the photovoltaic grid-connected energy storage type inverter coupling efficiency test result is generated.

[0013] Further, with reference to Figure 2 , Figure 2 For Figure 1 the detailed implementation step flowchart of step S1 in the embodiment, step S1 in the embodiment includes the following steps: Step S11: Based on the distributed multi-source collaborative unit, the edge computing acquisition terminal, the full-condition environment simulation room and the monitoring upper computer, a corresponding distributed hardware test topology is built, wherein the distributed multi-source collaborative unit includes a photovoltaic array simulator, an energy storage battery simulator, an intelligent power grid simulation unit and a distributed collaborative unit, the edge computing acquisition terminal includes a multi-channel power analyzer and a temperature monitoring module, the full-condition environment simulation room is configured with an altitude pressure adjustment module, a light intensity simulation module and a temperature control module; In the embodiment of the present application, a distributed hardware test topology is built based on a distributed multi-source collaborative unit, an edge computing collection terminal, a full-condition environment simulation room and a monitoring host computer. The distributed multi-source collaborative unit includes: a photovoltaic array simulator (output power range 0-50kW, voltage regulation range 200-1000V), an energy storage battery simulator (capacity simulation range 0-100kWh, charge and discharge current ±200A), an intelligent power grid simulation unit (output voltage 380V / 220V, frequency 50Hz±0.1Hz), and a distributed collaborative unit (supporting RS485 / Ethernet communication, control delay <100ms). The edge computing collection terminal includes: a multi-channel power analyzer (measurement accuracy 0.1 level, supporting 4-channel voltage and 4-channel current synchronous collection), and a temperature monitoring module (measurement range -40-150℃, accuracy ±0.5℃, including 8 temperature measurement channels). The full-condition environment simulation room is configured with: an altitude pressure adjustment module (adjustment range 50-110kPa, control accuracy ±0.5kPa), an illumination intensity simulation module (simulation range 0-1200W / m 2 , uniformity >90%), and a temperature control module (control range -20-60℃, accuracy ±0.5℃). Each device is connected to the monitoring host computer through Ethernet to form a complete test topology.

[0014] Step S12: The device initialization instruction is issued by the monitoring host computer to drive the distributed collaborative unit to start the devices in the order of full-condition environment simulation room→energy storage battery simulator→photovoltaic array simulator→intelligent power grid simulation unit; after the parameters of the full-condition environment simulation room are stable, including temperature fluctuation <±1℃, pressure fluctuation <±2kPa and illumination fluctuation <±50W / m 2 , and for a preset duration, the distributed collaborative unit enters the power closed-loop control mode; at the same time, the sampling rate of the multi-channel power analyzer is dynamically adjusted according to the inverter capacity, and the sampling interval of the temperature monitoring module is set to 1-2s / frame; In the embodiment of the present application, the device initialization instruction is issued by the monitoring host computer, the instruction is transmitted to the distributed collaborative unit through Ethernet, and the devices are started in the order of full-condition environment simulation room→energy storage battery simulator→photovoltaic array simulator→intelligent power grid simulation unit. The temperature control module of the full-condition environment simulation room is started first to stabilize the indoor temperature to 25℃; then the altitude pressure adjustment module is started to stabilize the pressure to 101kPa; finally, the illumination intensity simulation module is started to stabilize the illumination to 1000W / m 2 , and the temperature fluctuation is 0.8℃, the pressure fluctuation is 1.2kPa, and the illumination fluctuation is 35W / m 2, all meet the stability requirements, and after a preset duration of 10 minutes, the distributed cooperative unit triggers to enter a power closed-loop control mode. The inverter capacity is 30kW, and the sampling rate of the multi-channel power analyzer is dynamically adjusted according to the capacity: the voltage sampling rate is set to 10kHz, and the current sampling rate is set to 10kHz, ensuring that the sampling frequency is more than 5 times the inverter switching frequency; the temperature monitoring module sampling interval is set to 1.5s / frame, and 8 temperature measurement channels respectively collect the temperatures of the inverter IGBT module, filter inductor, energy storage interface and other key parts, and the sampled data is transmitted back to the monitoring host computer in real time.

[0015] Step S13: Based on the power closed-loop control mode, the distributed hardware test topology is controlled to collect and obtain multi-dimensional original data of the photovoltaic grid-connected energy storage type inverter under different working conditions, including photovoltaic-energy storage power data, energy storage charge and discharge data, grid parameter data and temperature monitoring data. In the embodiment of the application, based on the power closed-loop control mode, the distributed cooperative unit issues control instructions to each device: the photovoltaic array simulator is controlled to output power at 500W / m 2 , 800W / m 2 , 1000W / m 2 in three gears, and the corresponding output voltages are 400V, 600V and 800V; the energy storage battery simulator is controlled to discharge (discharge current 50A) when the photovoltaic power is insufficient, and to charge (charge current 30A) when the photovoltaic power is excessive; the smart grid simulation unit is controlled to maintain the grid voltage at 380V and the frequency at 50Hz. The edge computing collection terminal synchronously collects multi-dimensional original data: the multi-channel power analyzer collects photovoltaic-energy storage power data (photovoltaic output power, energy storage charge and discharge power, resolution 0.1W), grid parameter data (grid voltage, current, power factor, voltage resolution 0.1V, current resolution 0.01A); the temperature monitoring module collects inverter key part temperature data (IGBT temperature, inductor temperature, resolution 0.1℃), covering different working conditions such as illumination, temperature and load change, a total of 10 groups of working condition data, each group of working condition lasting 5 minutes.

[0016] Step S14: Kalman noise reduction and timestamp synchronization processing are performed on the multi-dimensional original data to control the time deviation of each data source within 10ms, and a time sequence aligned inverter multi-source test data set is generated.

[0017] In the embodiment of the present application, by performing Kalman noise reduction processing on the collected multi-dimensional original data: for photovoltaic power data, a Kalman filter model is established, the state equation is set as the power change rate = 0.1 x the power at the previous moment, the observation equation is set as the collected power = the real power + the measurement noise, the noise introduced by the illumination fluctuation is iteratively calculated and eliminated, and the fluctuation of the processed power data is reduced from ± 50W to ± 10W; the temperature data is denoised by the same method, and the fluctuation is reduced from ± 0.5℃ to ± 0.1℃. Time stamp synchronization processing is performed: taking the monitoring host computer system time as the reference, the time stamps of each data source collection time (photovoltaic power data time stamp, energy storage data time stamp, temperature data time stamp) are extracted, the deviation of each time stamp from the reference time is calculated, the data sampling points are adjusted through linear interpolation, the time deviation of the photovoltaic power data and the energy storage data is corrected to 5ms, the time deviation of the temperature data and the power data is corrected to 8ms, and both are controlled within 10ms. Finally, the inverter multi-source test data set with time sequence alignment is generated, which contains power, power grid and temperature data of 10 groups of working conditions, and each group of data contains 300 time-synchronized sampling points.

[0018] Further, step S2 comprises the following steps: Step S21: separating the photovoltaic simulation power subset and the energy storage charge and discharge power subset from the inverter multi-source test data set, and calculating the inverter total input power data according to the real-time photovoltaic output power data in the photovoltaic simulation power subset and the real-time energy storage power data in the energy storage charge and discharge power subset, wherein the inverter total input power data is the algebraic sum of the real-time photovoltaic output power data and the real-time energy storage power data, and when the energy storage is charging, the positive value is added, and when the energy storage is discharging, the negative value is deducted; and calculating the photovoltaic power proportion based on the real-time photovoltaic output power data and the inverter total input power data. In the embodiment of the present application, from the inverter multi-source test data set, the photovoltaic simulation power subset and the energy storage charge and discharge power subset are screened according to the data identifier, the photovoltaic simulation power subset contains real-time photovoltaic output power data every 1.5s / frame, and the energy storage charge and discharge power subset contains real-time energy storage power data every 1.5s / frame. Using the power calculation module, the real-time photovoltaic output power data and the real-time energy storage power data of the same time frame are subjected to algebraic operation, when the real-time energy storage power data is 10kW (indicating charging), the inverter total input power data is the positive value of 10kW added to the real-time photovoltaic output power data 50kW, i.e. 60kW; when the real-time energy storage power data is 8kW (indicating discharging), the inverter total input power data is the negative value of 8kW deducted from the real-time photovoltaic output power data 45kW, i.e. 37kW. Then, through the photovoltaic power proportion calculation formula, the real-time photovoltaic output power data of each time frame is divided by the corresponding inverter total input power data, such as 50kW ÷ 60kW, to obtain the photovoltaic power proportion of 83.3%, and the photovoltaic power proportion data is updated once every 1.5s.

[0019] Step S22: Extract the charging and discharging current, voltage and time series data corresponding to the energy storage battery from the energy storage charging and discharging power subset, and calculate the cumulative charging and discharging capacity data of the energy storage battery based on the charging and discharging current, voltage and time series data, and calculate the energy storage SOC value in combination with the rated capacity data of the energy storage battery; In the embodiment of the present application, by extracting the energy storage battery charging and discharging current, voltage data and corresponding time series data every 1.5s / frame from the energy storage charging and discharging power subset in chronological order, the charging and discharging current data range is 0-50A, and the voltage data range is 220-380V. Using the capacity calculation module, 3600 groups of charging and discharging current data and time interval 1.5s within 1 hour are integrated according to the formula "cumulative charging and discharging capacity = ∫(charging and discharging current × time)", such as 20A for a certain period of time. The integral result is 20A×3600s=72000As=72Ah within 1 hour, and the cumulative charging and discharging capacity data is obtained. The rated capacity data of the energy storage battery is 100Ah, and the energy storage SOC value is calculated by the formula "energy storage SOC value = (rated capacity-cumulative charging and discharging capacity) ÷ rated capacity × 100%", such as (100Ah-72Ah) ÷ 100Ah × 100%, the energy storage SOC value is 28%, and the energy storage SOC value is updated every 1 hour.

[0020] Step S23: Extract the power grid parameter subset from the inverter multi-source test data set, and separate out the power grid voltage waveform data. Fourier transform the power grid voltage waveform data to obtain the fundamental voltage component data and each harmonic voltage component data; calculate the total effective value of the power grid harmonics based on the fundamental voltage component data and each harmonic voltage component data, and generate the power grid THD content through the formula "power grid THD content = total effective value of power grid harmonics ÷ effective value of fundamental voltage component × 100%"; In the embodiment of the present application, the power grid parameter subset is extracted from the inverter multi-source test data set according to the data classification label, and the power grid parameter subset contains power grid voltage waveform data every 1.5s / frame, and the voltage waveform data sampling frequency is 5kHz. Fourier transform module is adopted to perform Fourier decomposition on the power grid voltage waveform data of each time frame to obtain the fundamental voltage component data with a frequency of 50Hz and each harmonic voltage component data with a frequency of 100Hz, 150Hz and 200Hz. The effective value of the fundamental voltage component data is 380V, the effective value of the 100Hz harmonic voltage component is 5V, the effective value of the 150Hz harmonic voltage component is 3V, and the effective value of the 200Hz harmonic voltage component is 2V. The total effective value of the power grid harmonics is calculated by the formula "total effective value of power grid harmonics = √(sum of squares of effective values of each harmonic voltage component)", and the total effective value of the power grid harmonics is calculated as √(5 2 +32 +2 2 ) = 38 = 6.16V. Substituting into the grid THD content calculation formula, 6.16V ÷ 380V x 100%, the grid THD content is 1.62%, and the grid THD content data is updated every 1.5s.

[0021] Step S24: Extracting the temperature monitoring subset from the inverter multi-source test data set, and separating the real-time temperature data and environmental temperature data corresponding to each component, and calculating the real-time temperature rise difference data corresponding to each component based on the real-time temperature data and the environmental temperature data, and generating the component temperature rise data in combination with the sampling time interval of the temperature monitoring module; In the embodiment of the application, from the inverter multi-source test data set, the temperature monitoring subset is extracted according to the component identifier, the temperature monitoring subset contains the real-time temperature data and environmental temperature data of the IGBT module, the reactor, the radiator and other components every 1.5s / frame, the real-time temperature data ranges from 25-80℃, and the environmental temperature data ranges from 20-25℃. Using the temperature rise calculation module, the real-time temperature data of each time frame of the same component is subtracted from the corresponding environmental temperature data, such as the real-time temperature data of the IGBT module 65℃ minus the environmental temperature data 23℃, to obtain the real-time temperature rise difference data of 42℃. In combination with the sampling time interval of the temperature monitoring module 1.5s, the average value of 400 groups of real-time temperature rise difference data within 10 minutes is taken, such as the average value of 400 groups of data is 40℃, to generate the component temperature rise data of 40℃, and the component temperature rise data is updated every 10 minutes.

[0022] Step S25: Obtain the regional light wave intensity parameter and construct the corresponding inverter basic dimension parameter matrix with the photovoltaic power ratio, the energy storage SOC value, the grid THD content and the component temperature rise data, and perform coupling weighting operation based on the inverter basic dimension parameter matrix to generate the initial coupling weighting coefficient of the inverter.

[0023] In the embodiment of the present application, the regional light monitoring system is used to obtain the light wave intensity index of the region once per hour, such as 250, 260, 245, etc. The 1.5s / frame photovoltaic power ratio data obtained in step S21 (taking the hourly average value, such as 80%, 78%, 82%), the 1-hour 1-time energy storage SOC value data obtained in step S22 (such as 30%, 28%, 32%), the 1.5s / frame grid THD content data obtained in step S23 (taking the hourly average value, such as 1.6%, 1.5%, 1.7%), and the 10-minute 1-time component temperature rise data obtained in step S24 (taking the hourly average value, such as 40℃, 38℃, 42℃) are collected. The regional light wave intensity index, photovoltaic power ratio, energy storage SOC value, grid THD content, and component temperature rise data are used as matrix row vectors, and the hourly time unit is used as matrix column vectors to construct a 5-row×24-column inverter basic dimension parameter matrix, and the matrix elements are the specific values of each parameter at the corresponding time point. A weighted operation module is used, and the weights of the parameters are set according to the influence degree of the parameters on the coupling efficiency, such as a regional light wave intensity index weight of 30%, a photovoltaic power ratio weight of 25%, an energy storage SOC value weight of 20%, a grid THD content weight of 15%, and a component temperature rise data weight of 10%. The product sum of the matrix element value×the corresponding row weight×the corresponding column weight is calculated to finally generate the initial coupling weight coefficient of the inverter.

[0024] Further, step S25 includes the following steps: Step S251: Obtain the original light data of the region, including the global light radiation data of satellite remote sensing, the real-time light data of single point of ground distributed sensor, and the historical light statistical data recorded by the weather station, and extract the spectral radiation intensity data of different wave bands from the global light radiation data, and generate the original light space-time distribution matrix in combination with the time continuity of the real-time light data of single point. In the embodiment of the present application, the original light data of the region is obtained, the global light radiation data of satellite remote sensing covers an area of 100km×100km, and the spectral radiation intensity of 300-1100nm wave band (300-400nm is 50W / m 2 , 400-700nm is 400W / m 2 , and 700-1100nm is 200W / m 2 ); the ground distributed sensor records the real-time light data of single point every 10 minutes (8:00 is 300W / m 2 , 9:00 is 500W / m 2 , and 10:00 is 700W / m 2 ); and the weather station provides the historical light statistical data of nearly 5 years (600W / m 2 per day in summer, and 300W / m 2). The intensity of each band is extracted from the global data, and the time continuity of the single-point data is combined to construct a 10*10*12 (space*space*time) original matrix of the illumination space-time distribution with an interval of 1 hour and 10 km, and the matrix element is the comprehensive illumination intensity of the corresponding space-time point.

[0025] Step S252: The original matrix of the illumination space-time distribution is normalized, and the corresponding illumination radiation cumulative value per unit time is calculated based on the normalized original matrix of the illumination space-time distribution, and then the geographic latitude and longitude data of the region are obtained to derive the corresponding illumination incidence angle change parameter of the region; In the embodiment of the present application, the original matrix of the illumination space-time distribution is normalized, and each element value is divided by the maximum value 1000 W / m 2 of the matrix to obtain a normalized matrix in the range of 0-1. The illumination radiation cumulative value per unit time (1 hour) is calculated, such as the period of 8:00-9:00, the normalized value increases from 0.3 to 0.5, and the cumulative value is (0.3+0.5)*1 / 2*1=0.4 (unit normalized cumulative value). The geographic latitude and longitude (30° N, 120° E) of the region is obtained, and the illumination incidence angle change parameter is calculated by a trigonometric function, the incidence angle is 63.5° at 12 o'clock on the winter solstice, 26.5° at 12 o'clock on the summer solstice, and 50° at 12 o'clock on the spring and autumn equinoxes, and an incidence angle change parameter table is generated according to seasons and time periods.

[0026] Step S253: The illumination intensity attenuation coefficient corresponding to different seasons and time periods is calculated according to the illumination incidence angle change parameter and the historical illumination statistical data, and the illumination intensity attenuation coefficient is combined with the real-time illumination radiation intensity data to generate a real-time illumination effective intensity parameter; In the embodiment of the present application, the illumination intensity attenuation coefficient of different seasons and time periods is calculated according to the illumination incidence angle change parameter and the historical illumination statistical data. When the incidence angle is 26.5° at noon in summer, the historical illumination statistical value is 600 W / m 2 , and the attenuation coefficient is 0.9 (600 W / m 2 ÷ theoretical value 667 W / m 2 ) according to the atmospheric attenuation model; when the incidence angle is 63.5° at noon in winter, the historical value is 300 W / m 2 , and the attenuation coefficient is 0.6 (300 W / m 2 ÷ theoretical value 500 W / m 2 ); the incidence angle is 60° at 9 o'clock on the spring and autumn equinoxes, and the attenuation coefficient is 0.75. The real-time illumination radiation intensity data (such as 500 W / m 2 measured at a certain moment) is multiplied by the corresponding attenuation coefficient 0.75, 500*0.75=375 W / m 2 , to generate a real-time illumination effective intensity parameter.

[0027] Step S254: Time domain analysis is performed on the real-time light effective intensity parameter to extract the corresponding light intensity peak value, valley value and fluctuation times per unit time, and the light intensity fluctuation amplitude parameter is calculated, and the light stability index is generated based on the light intensity fluctuation amplitude parameter and the obtained regional weather type data; the real-time light effective intensity parameter, the light stability index and the light incidence angle change parameter are coupled and calculated to obtain the regional light wave intensity parameter; In the embodiment of the application, by performing time domain analysis on the real-time light effective intensity parameter, the light intensity peak value 400 W / m 2 , the valley value 350 W / m 2 , and the fluctuation times 3 times in 1 hour are extracted, and the fluctuation amplitude parameter (400-350) ÷ 375 = 13.3% is calculated. The regional weather type data (60% of sunny days, 30% of cloudy days, and 10% of overcast days) is obtained, and the probability density fitting is performed in combination with the fluctuation amplitude parameter, the sunny day fluctuation amplitude 10% corresponds to the probability 0.8, the cloudy day fluctuation amplitude 20% corresponds to the probability 0.6, and the light stability index 0.7 (weighted average value) is generated. The real-time light effective intensity parameter 375 W / m 2 , the light stability index 0.7, and the light incidence angle change parameter 50° are coupled and calculated, 375×0.7 ÷ 50 = 5.25, and the regional light wave intensity parameter 5.25 is obtained.

[0028] Step S255: The regional light wave intensity parameter is constructed into a corresponding inverter basic dimension parameter matrix with photovoltaic power proportion, energy storage SOC value, power grid THD content and component temperature rise data in different test periods, including daytime peak segment, daytime flat segment and night valley segment, wherein the regional light wave intensity parameter, photovoltaic power proportion, energy storage SOC value, power grid THD content and component temperature rise data are used as matrix row vectors, different test periods are used as matrix column vectors, and elements are specific values of each parameter in different test periods, and the analytic hierarchy process is used to assign weights of each column period and each row parameter in the inverter basic dimension parameter matrix, first, the column period weight is determined according to the inverter running load proportion of each period, and then the row parameter weight is determined according to the influence degree of each parameter on the coupling efficiency; a coupling weighted operation model is constructed based on the matrix element value, the column period weight and the row parameter weight, and the inverter initial coupling weighted coefficient is generated by summing the product of the element value, the row weight and the column weight.

[0029] In the embodiment of the present application, the daytime peak segment (10:00-15:00), the daytime flat segment (8:00-10:00, 15:00-18:00), and the nighttime valley segment (18:00-8:00 the next day) are selected as the test period. The regional light intensity index (peak segment 5.25, flat segment 4.8, valley segment 0), the photovoltaic power ratio (peak segment 80%, flat segment 60%, valley segment 0), the energy storage SOC value (peak segment 70%, flat segment 60%, valley segment 90%), the grid THD content (peak segment 2%, flat segment 3%, valley segment 1%), and the component temperature rise data (peak segment 40K, flat segment 30K, valley segment 20K) are used as the row vector, and the three time periods are used as the column vector to construct a 5-row 3-column inverter basic dimension parameter matrix. According to the peak segment load ratio 50%, the flat segment 30%, and the valley segment 20%, the column weight is determined by the analytic hierarchy process; according to the light intensity index 20%, the photovoltaic power ratio 30%, the energy storage SOC value 20%, the grid THD content 15%, and the component temperature rise 15%, the row weight is determined. The sum of the product of the matrix element value, the row weight, and the column weight (such as the peak segment: 5.25x0.2x0.5+80x0.3x0.5+…) generates the initial coupling weight coefficient of the inverter 0.86.

[0030] Further, step S3 includes the following steps: Step S31: Extracting energy storage charge and discharge time sequence data from the inverter multi-source test data set, including charge and discharge start time, end time, current change curve, and voltage fluctuation data; calculating the single cycle duration based on the charge and discharge start and end time, determining the effective charge and discharge capacity combined with the current change curve and voltage fluctuation data, generating the energy storage cycle depth parameter by accumulating the ratio of the effective charge and discharge capacity to the rated capacity of the energy storage; and deriving the energy storage cycle attenuation coefficient according to the energy storage cycle depth parameter and the cycle number statistical data; In the embodiment of the present application, the energy storage charge and discharge timing data is screened out from the inverter multi-source test data set according to the time stamp, including the starting time (such as 8:00, 10:30) and the termination time (such as 9:30, 12:00) of each charge and discharge, 1 group of current change curve (range 0-50A) and voltage fluctuation data (range 220-380V) per second. The time calculation module is used to subtract the starting time from the termination time, such as 9:30-8:00, to obtain the single cycle length of 90 minutes. Through the capacity calculation module, the current change curve is integrated with the time interval of 1 second. If the current is stable at 30A in a certain cycle, the integral result of 90 minutes is 30A x 5400s = 162000As = 162Ah, which is determined as the effective charge and discharge capacity. The rated capacity of the energy storage is 200Ah, and the calculation result of 162Ah ÷ 200Ah = 0.81 is obtained. The energy storage cycle depth parameter 0.81 is generated. The cycle depth parameters of 50 cycles are counted. If the effective charge and discharge capacity of the 50th cycle is 150Ah, the calculation result of 150Ah ÷ 200Ah = 0.75 is obtained. Then, through (0.81-0.75) ÷ 50 = 0.0012, the energy storage cycle attenuation coefficient 0.0012 is derived.

[0031] Step S32: separating the power grid parameter subset from the inverter multi-source test data set and extracting the power grid voltage harmonic spectrum data, the fundamental frequency stability data and the voltage sag recovery time data; performing inverse Fourier transform on the power grid voltage harmonic spectrum data to obtain harmonic time domain waveform features, calculating a harmonic energy proportion parameter in combination with the fundamental frequency stability data; calculating an interference coefficient on the inverter efficiency based on the harmonic energy proportion parameter and the voltage sag recovery time data, and obtaining a power grid harmonic influence coefficient after normalization processing of the interference coefficient; In the embodiment of the present application, the power grid parameter subset is separated from the inverter multi-source test data set, and 1 group of power grid voltage harmonic spectrum data (containing 1-20 harmonics), fundamental frequency stability data (50±0.2Hz) and voltage sag recovery time data (0.5-2s) per 1.5s are extracted. The harmonic spectrum data is processed by the inverse Fourier transform module to obtain the harmonic time domain waveform features (such as the peak value of 10th harmonic waveform 3V). The ratio of fundamental power to total power is calculated. If the fundamental power is 980W and the total power is 1000W, the harmonic energy proportion parameter is 0.02. The interference coefficient on the inverter efficiency is calculated to be 0.052 by 0.02 x 0.6 + 0.1 x 0.4 = 0.052, corresponding to a voltage sag recovery time of 1s. The interference coefficient is divided by the maximum value 0.2 (preset), 0.052 ÷ 0.2 = 0.26, and the normalized processing obtains the power grid harmonic influence coefficient 0.26.

[0032] Step S33: Extracting a temperature monitoring subset from the inverter multi-source test data set, including real-time temperature, temperature rise rate and temperature fluctuation period data of IGBT and capacitor key components, calculating the component relative temperature rise parameter based on the real-time temperature data combined with the environmental temperature data, and calculating the component thermal cycle damage factor based on the temperature rise rate and temperature fluctuation period data, and comparing the component thermal cycle damage factor with the rated life parameter corresponding to the component to generate the component aging coefficient; In the embodiment of the application, by extracting a temperature monitoring subset from the inverter multi-source test data set, containing IGBT and capacitor, every 1.5s1 group real-time temperature (IGBT range 40-80℃, capacitor range 30-60℃), temperature rise rate (0.5-2℃ / min) and temperature fluctuation period (10-30min). Take the environmental temperature 25℃, the IGBT real-time temperature 60℃, 60℃-25℃=35℃, calculate the component relative temperature rise parameter 35℃. Based on the temperature rise rate 1℃ / min and the fluctuation period 20min, through the formula "thermal cycle damage factor=(temperature rise rate×fluctuation period)÷1000", 1×20÷1000=0.02, the component thermal cycle damage factor 0.02 is calculated. Get the IGBT rated life parameter 10000h, the capacitor rated life 8000h, compare the damage factor 0.02, if the cumulative damage reaches 0.02, the life consumption is 200h, 200h÷10000h=0.02, generate the component aging coefficient 0.02.

[0033] Step S34: Integrating the energy storage cycle attenuation coefficient, the power grid harmonic influence coefficient and the component aging coefficient into a coupling dynamic correction factor in the form of a row vector, and distributing the factor weight vector to the corresponding influence degree of the photovoltaic grid-connected energy storage type inverter based on the coupling dynamic correction factor, and performing tensor product calculation based on the coupling dynamic correction factor, the factor weight vector and the initial coupling weighting coefficient of the inverter to obtain the multi-dimensional coupling correction weighting coefficient of the inverter; In the embodiment of the application, by integrating the energy storage cycle attenuation coefficient 0.0012, the power grid harmonic influence coefficient 0.26 and the component aging coefficient 0.02 into a row vector [0.0012, 0.26, 0.02] form, a coupling dynamic correction factor is generated. According to the influence degree, the factor weight vector [0.1, 0.6, 0.3] (the power grid has the greatest influence, followed by the component aging, and finally the energy storage attenuation). Get the initial coupling weighting coefficient of the inverter 0.85, and perform tensor product calculation: (0.0012×0.1+0.26×0.6+0.02×0.3)×0.85= (0.00012+0.156+0.006)×0.85=0.16212×0.85≈0.1378, get the multi-dimensional coupling correction weighting coefficient of the inverter 0.1378.

[0034] Step S35: selecting different power test points and obtaining the inverter multi-dimensional coupling correction weighting coefficients corresponding to the power test points, and performing coupling normalization correction calculation based on the inverter multi-dimensional coupling correction weighting coefficients corresponding to the different power test points to generate multi-dimensional coupling normalization correction weighting coefficients.

[0035] In the embodiment of the application, 20%, 50%, 80%, and 100% rated power (100 kW rated power, corresponding to 20 kW, 50 kW, 80 kW, and 100 kW) are selected as different power test points (5%, 10%, 15%, 25%, 30%, 50%, 75%, and 100% can also be selected). The inverter multi-dimensional coupling correction weighting coefficients corresponding to each test point are obtained: 0.12 at 20 kW, 0.1378 at 50 kW, 0.15 at 80 kW, and 0.16 at 100 kW. The sum of all test point coefficients is calculated: 0.12+0.1378+0.15+0.16=0.5678. Through coupling normalization correction calculation (the ratio of each coefficient to the sum), 0.12÷0.5678≈0.211 at 20 kW, 0.1378÷0.5678≈0.243 at 50 kW, 0.15÷0.5678≈0.264 at 80 kW, and 0.16÷0.5678≈0.282 at 100 kW. The multi-dimensional coupling normalization correction weighting coefficients are 0.211, 0.243, 0.264, and 0.282, respectively.

[0036] Further, step S4 includes the following steps: Step S41: obtaining inverter input and output power data, including input power data and output power data, wherein the input power data includes photovoltaic input power output by a photovoltaic array simulator and charge and discharge power of an energy storage battery simulator, and the output power data is inverter grid-connected alternating current power. In the embodiment of the application, the input and output power data is extracted from the inverter multi-source test data set. The input power data includes photovoltaic input power output by a photovoltaic array simulator (range 0-100 kW, recorded once every 1.5 s, such as 45 kW and 58 kW) and charge and discharge power of an energy storage battery simulator (positive when charging, range 0-30 kW; negative when discharging, range -30-0 kW, recorded once every 1.5 s, such as 12 kW and -7 kW). The output power data is inverter grid-connected alternating current power (range 0-100 kW, recorded once every 1.5 s, such as 43 kW and 54 kW). The timestamps of all input and output power data are strictly corresponding to form a continuous power time sequence.

[0037] Step S42: Time integration processing is performed on the inverter input and output power data to calculate the corresponding total input energy and total output energy in a unit period of 5-10s as the integration period, and the photovoltaic weight component, the energy storage weight component, the grid weight component, the temperature rise weight component and the region weight component are disassembled from the multi-dimensional coupling normalization correction weight coefficient, and the photovoltaic power ratio coefficient and the energy storage power ratio coefficient are calculated based on the photovoltaic input power and the charging and discharging power, and the total input energy is corrected once based on the photovoltaic weight component, the energy storage weight component and the photovoltaic power ratio coefficient and the energy storage power ratio coefficient to generate a first corrected input energy; In the embodiment of the application, by performing time integration processing on the inverter input and output power data, setting the integration period to 8s, and multiplying the photovoltaic input power, the energy storage charging and discharging power, and the grid-connected alternating current power in each period by the time interval 1.5s, the total input energy (such as photovoltaic input energy 360kJ, energy storage charging energy 90kJ, and total input energy 450kJ) and the total output energy (such as 430kJ) in a unit period are obtained. The photovoltaic weight component 0.5, the energy storage weight component 0.2, the grid weight component 0.2, the temperature rise weight component 0.05, and the region weight component 0.05 are disassembled from the multi-dimensional coupling normalization correction weight coefficient (0.211 when 20kW, 0.243 when 50kW, 0.264 when 80kW, and 0.282 when 100kW, and the value corresponding to 50kW is 0.243). The photovoltaic power ratio coefficient is 45kW÷(45kW+12kW)=0.789, and the energy storage power ratio coefficient is 12kW÷(45kW+12kW)=0.211. The first corrected input energy is generated by the first correction formula "first corrected input energy=total input energy×(photovoltaic weight component×photovoltaic power ratio coefficient+energy storage weight component×energy storage power ratio coefficient)", that is, 450kJ×(0.5×0.789+0.2×0.211)=196.515kJ.

[0038] Step S43: Obtain the grid working condition coefficient, the temperature rise deviation coefficient, and the region light fluctuation coefficient, and perform secondary correction on the first corrected input energy by introducing the grid weight component, the temperature rise weight component, and the region weight component to generate the corrected input energy, specifically, corrected input energy=first corrected input energy×(1+grid weight component×grid working condition coefficient+temperature rise weight component×temperature rise deviation coefficient+region weight component×region light fluctuation coefficient), and at the same time, compared with the standard input energy range of the same power level inverter, if it exceeds the range ±10%, return to step S42 to disassemble the weight component again; In the embodiment of the present application, the grid operating condition coefficient 0.03 (grid voltage fluctuation 3%), temperature rise deviation coefficient 0.02 (component temperature rise exceeds the standard 2%), and regional light fluctuation coefficient 0.01 (light intensity fluctuation 1%) are obtained. Substituting the quadratic correction formula "corrected input energy = first corrected input energy × (1 + grid weight component × grid operating condition coefficient + temperature rise weight component × temperature rise deviation coefficient + regional weight component × regional light fluctuation coefficient)", that is, 196.515 kJ × (1 + 0.2 × 0.03 + 0.05 × 0.02 + 0.05 × 0.01) = 196.515 kJ × 1.0075 = 198.089 kJ. The standard input energy range of the same power level inverter is 180-220 kJ, and 198.089 kJ is within the range, so there is no need to return to step S42; if the calculation result is 245 kJ, which exceeds the range by 10%, the weight components need to be re-decomposed.

[0039] Step S44: Based on the corrected input energy and the total output energy, and combined with the coupling dynamic correction factor, the coupling efficiency is calculated to obtain the corresponding coupling efficiency value of the photovoltaic grid-connected energy storage type inverter under the current operating condition; In the embodiment of the present application, based on the corrected input energy 198.089 kJ and the total output energy 430 kJ, combined with the coupling dynamic correction factor (energy storage cycle attenuation coefficient 0.95, grid harmonic influence coefficient 0.98, and component aging coefficient 0.9), the coupling efficiency calculation formula "coupling efficiency value = (total output energy ÷ corrected input energy) × energy storage cycle attenuation coefficient × grid harmonic influence coefficient × component aging coefficient × 100%" is used, that is, (430 ÷ 198.089) × 0.95 × 0.98 × 0.9 × 100% ≈ 2.171 × 0.8379 × 100% ≈ 181.9%, to obtain the coupling efficiency value under the current operating condition (note: the example data is only for demonstration of the calculation logic, and the actual efficiency should be ≤100%).

[0040] Step S45: Based on the coupling efficiency values obtained under different operating conditions, a multi-dimensional coupling efficiency surface graph is generated, and coupling over-limit quadratic correction and testing are performed according to the multi-dimensional coupling efficiency surface graph to generate the coupling efficiency test results of the photovoltaic grid-connected energy storage type inverter under the corresponding operating condition.

[0041] In the embodiment of the present application, by selecting 27 working conditions of photovoltaic power ratio 30%-90% (every 20% gradient), energy storage SOC value 20%-100% (every 20% gradient), and grid THD content 1%-5% (every 1% gradient), the coupling efficiency value (range 85%-98%) corresponding to each working condition is obtained. The multi-dimensional coupling efficiency surface graph is drawn with photovoltaic power ratio as X-axis, energy storage SOC value as Y-axis, and coupling efficiency value as Z-axis. The three working condition points with efficiency lower than 85% in the surface graph are superimposed and modified twice, and the coupling dynamic correction factor is adjusted to recalculate the efficiency, so that it reaches more than 85%. Based on the modified surface graph, the coupling efficiency value, qualified grade and corresponding parameter data of each working condition are output, and the photovoltaic grid-connected energy storage type inverter coupling efficiency test result is generated.

[0042] Further, step S45 includes the following steps: Step S451: By extracting different working condition parameters, including photovoltaic power ratio, energy storage SOC value, grid THD content, component temperature rise data and regional light wave intensity parameter, and the coupling efficiency value under the corresponding working condition, a working condition parameter-coupling efficiency mapping data set is constructed; the correlation coefficient between each working condition parameter is calculated based on the working condition parameter-coupling efficiency mapping data set, and the core working condition parameters which have significant influence on coupling efficiency are screened out through the correlation coefficient, and then the dimension division basis of the surface graph is generated in combination with the value range of the core working condition parameters; In the embodiment of the present application, 500 groups of different working condition parameters are extracted from the previous test data, the photovoltaic power ratio range is 30%-90% (every 10% gradient), the energy storage SOC value range is 20%-100% (every 20% gradient), the grid THD content range is 0.5%-3% (every 0.5% gradient), the component temperature rise data range is 20-60℃ (every 10℃ gradient), and the regional light wave intensity parameter range is 200-300 (every 20 gradient). At the same time, the coupling efficiency value (range 85%-98%) under the corresponding working condition is extracted to construct the working condition parameter-coupling efficiency mapping data set. The correlation coefficient formula is used to calculate the correlation coefficient between each working condition parameter and the coupling efficiency by using the correlation degree calculation module. The photovoltaic power ratio correlation coefficient is 0.82, the energy storage SOC value is 0.75, the grid THD content is-0.68, the component temperature rise data is-0.62, and the regional light wave intensity parameter is 0.55. The photovoltaic power ratio and the energy storage SOC value with absolute value greater than 0.7 of the correlation coefficient are selected as the core working condition parameters. In combination with the value range of the core working condition parameters, the photovoltaic power ratio is divided into 6 intervals of 30%-40%, 40%-50%, …, 80%-90%, the energy storage SOC value is divided into 4 intervals of 20%-40%, 40%-60%, …, 80%-100%, and the dimension division basis of the surface graph is generated.

[0043] Step S452: data completion is performed on the working condition parameter-coupling efficiency mapping dataset to fill in the missing working condition-efficiency data points; based on the completed dataset, an initial multi-dimensional coupling surface model is constructed with the core working condition parameters as the coordinate axes, including the photovoltaic power ratio as the X-axis, the energy storage SOC value as the Y-axis, and the coupling efficiency value as the Z-axis; the slope change of adjacent data points in the initial multi-dimensional coupling surface model is calculated to generate a surface smoothness parameter, and if the surface smoothness parameter is lower than a preset threshold value, the initial multi-dimensional coupling surface model is optimized through a Gaussian filtering algorithm to obtain a multi-dimensional coupling efficiency surface graph; In the embodiment of the present application, through the inspection of the working condition parameter-coupling efficiency mapping dataset, it is found that 5 groups of working condition-efficiency data corresponding to the photovoltaic power ratio of 50%-60% and the energy storage SOC value of 40%-60% are missing. The interpolation completion method is adopted, the adjacent data points (photovoltaic power ratio of 40%-50%, energy storage SOC value of 40%-60% efficiency of 92%; photovoltaic power ratio of 60%-70%, energy storage SOC value of 40%-60% efficiency of 93%; photovoltaic power ratio of 50%-60%, energy storage SOC value of 20%-40% efficiency of 91%; photovoltaic power ratio of 50%-60%, energy storage SOC value of 60%-80% efficiency of 94%) are taken as the basis, and the missing data point efficiency value of 92.5% is calculated through bilinear interpolation to complete the data completion. Based on the completed dataset, the initial multi-dimensional coupling surface model is constructed with the X-axis photovoltaic power ratio (30%-90%), the Y-axis energy storage SOC value (20%-100%), and the Z-axis coupling efficiency value (85%-98%). The slope of adjacent data points in the initial model (such as X=30%, Y=20% efficiency 85%; X=40%, Y=20% efficiency 87%) is calculated as (87-85) / (40-30)=0.2, the slope change of all adjacent points is calculated to generate a surface smoothness parameter of 0.15. The preset smoothness threshold is 0.2, and since 0.15 is lower than the threshold, the Gaussian filtering algorithm is adopted, the filtering standard deviation is set to 1.2, the initial model is smoothed to obtain a multi-dimensional coupling efficiency surface graph, and the slope change of adjacent regions of the surface is uniform.

[0044] Step S453: the peak region and the step-down region corresponding to the coupling efficiency are extracted from the multi-dimensional coupling efficiency surface graph, the efficiency average of the peak region and the minimum efficiency of the step-down region are calculated, and the efficiency fluctuation amplitude parameter is generated based on the difference between the two; the coupling efficiency overrun determination threshold is determined in combination with the rated efficiency range of the inverter, and the overrun working condition points in the surface graph that exceed the threshold range are marked through the comparison between the efficiency fluctuation amplitude parameter and the coupling efficiency overrun determination threshold; In the embodiment of the present application, by observing from the multi-dimensional coupling efficiency surface graph, the coupling efficiency of the region with photovoltaic power proportion of 70%-80% and energy storage SOC value of 80%-100% is concentrated in 96%-98%, which is determined as the peak region, and the efficiency average of 15 groups of data in this region is 97.2%; the coupling efficiency of the region with photovoltaic power proportion of 30%-40% and energy storage SOC value of 20%-40% is concentrated in 85%-87%, which is determined as the sudden drop region, and the minimum efficiency in this region is 85.3%. The difference between the two is 97.2%-85.3%=11.9%, and the efficiency fluctuation amplitude parameter 11.9% is generated. Given the inverter rated efficiency range of 88%-98%, the coupling efficiency over-limit determination threshold is determined to be lower than 88% or higher than 98%. By comparing all working condition points in the surface graph, it is found that there are 8 working condition points in the region with photovoltaic power proportion of 30%-40% and energy storage SOC value of 20%-30% whose efficiency is lower than 88%, which are marked as over-limit working condition points.

[0045] Step S454: Root cause analysis is performed on the marked over-limit working condition points to extract the working condition parameter data and coupling dynamic correction factors corresponding to the over-limit working condition points, calculate the deviation rate of the correction factors under the over-limit working condition; based on the correlation between the deviation rate and the over-limit efficiency value, the contribution degree parameter of each correction factor to the efficiency over-limit is calculated; the adjustment amplitude of the correction factor is determined according to the contribution degree parameter, the coupling dynamic correction factor is iteratively optimized; the optimized correction factor is re-substituted into the calculation to obtain the optimized coupling efficiency value; the multi-dimensional coupling efficiency surface graph is updated based on the optimized coupling efficiency value, and whether there is an over-limit working condition point is checked again, if there is still, the step is repeated until the coupling efficiency of all working condition points in the surface graph is within the rated efficiency range; In the embodiment of the application, by extracting the working condition parameters (such as photovoltaic power ratio 32%, energy storage SOC value 25%, etc.) corresponding to the eight over-limit working condition points and the coupling dynamic correction factors (energy storage cycle attenuation coefficient 0.0015, power grid harmonic influence coefficient 0.3, element aging coefficient 0.025), the correction factor deviation rate is calculated, the energy storage cycle attenuation coefficient standard value is 0.0012, the deviation rate (0.0015-0.0012) / 0.0012=25%; the power grid harmonic influence coefficient standard value is 0.25, the deviation rate (0.3-0.25) / 0.25=20%; the element aging coefficient standard value is 0.02, the deviation rate (0.025-0.02) / 0.02=25%. Based on the correlation between the deviation rate and the over-limit efficiency value (such as 86.5%, over-limit 1.5%), the contribution degree parameters of each correction factor are calculated by the contribution degree formula, the energy storage cycle attenuation coefficient contribution degree is 0.3, the power grid harmonic influence coefficient is 0.4, and the element aging coefficient is 0.3. According to the contribution degree, the correction factor is adjusted, the adjustment range of the power grid harmonic influence coefficient is 20%×0.4=8%, and the adjustment is 0.3×(1-8%)=0.276; the adjustment range of the energy storage cycle attenuation coefficient is 25%×0.3=7.5%, and the adjustment is 0.0015×(1-7.5%)=0.0013875; the adjustment range of the element aging coefficient is 25%×0.3=7.5%, and the adjustment is 0.025×(1-7.5%)=0.023125. The optimized correction factor is substituted into the coupling efficiency calculation formula, and the optimized efficiency value (such as 86.5% to 88.2%) is obtained, and the multi-dimensional coupling efficiency curve is updated. Again, it is found that there are still two over-limit working condition points, and the above steps are repeated, and after the second adjustment of the correction factor, the efficiency of all working condition points is within the range of 88%-98%.

[0046] Step S455: Perform coupling efficiency test based on the multi-dimensional coupling efficiency curve after over-limit correction to generate coupling efficiency test results of the photovoltaic grid-connected energy storage type inverter under corresponding working conditions.

[0047] In the embodiment of the present application, 20 typical test operating points (covering different photovoltaic power ratios, energy storage SOC value intervals) are selected based on the multi-dimensional coupling efficiency surface map after the limit is modified, such as photovoltaic power ratio 45%, energy storage SOC value 50%; photovoltaic power ratio 65%, energy storage SOC value 70%, etc. For each test operating point, the corresponding coupling efficiency value is read from the surface map, and the inverter input power and output power data under the operating condition are collected by combining the actual test equipment, and the efficiency formula "coupling efficiency = output power / input power*100%" is used for verification, such as input power 50kW, output power 48.5kW, calculated efficiency 97%, and the error with the read value 97.2% is 0.2%, which meets the error requirement. The test and verification results of 20 operating points are summarized to generate the coupling efficiency test result report of the photovoltaic grid-connected energy storage type inverter under the corresponding operating condition, including operating condition parameters, coupling efficiency value, verification error and other information.

[0048] Further, the step S453 of extracting the peak region and the step-down region corresponding to the coupling efficiency from the multi-dimensional coupling efficiency surface map includes the following steps: The three-dimensional data of the multi-dimensional coupling efficiency surface map is subjected to grid processing, the surface is divided into a plurality of uniform grid cells, and the coupling efficiency value and the corresponding operating condition coordinates of each grid cell are extracted; the overall efficiency average of the surface is calculated based on the coupling efficiency value of the grid cell, and the efficiency deviation matrix is generated by combining the deviation of the coupling efficiency value of each grid cell from the average. In the embodiment of the present application, the three-dimensional data (X-axis photovoltaic power ratio 30%-90%, Y-axis energy storage SOC value 20%-100%, Z-axis coupling efficiency 85%-98%) of the multi-dimensional coupling efficiency surface map is subjected to grid processing, the X-axis is divided into 12 intervals at an interval of 5% by using a grid division tool, the Y-axis is divided into 16 intervals at an interval of 5%, and 12*16=192 uniform grid cells are formed, each of which corresponds to a working condition coordinate range of 5%*5%. The coupling efficiency value of each grid cell is extracted, such as X=30%-35%, Y=20%-25% cell efficiency 85.2%, X=70%-75%, Y=80%-85% cell efficiency 97.8%, etc. The efficiency values of the 192 grid cells are summed and divided by 192 by using the average calculation module to obtain the overall efficiency average of the surface 92.5%. The deviation of the efficiency value of each grid cell from the average is calculated, such as 85.2%-92.5%=-7.3%, 97.8%-92.5%=5.3%, and the deviation values are arranged in order of grid coordinates to generate a 12-row 16-column efficiency deviation matrix.

[0049] Further, the grid cells with positive deviation are screened from the efficiency deviation matrix, the standard deviation of the efficiency values of these cells is calculated, and the grid cells with the standard deviation of the efficiency values less than a preset stability threshold are classified as potential peak value cells; based on the working condition coordinates of the potential peak value cells, the spatial distance between the cells is calculated, the potential peak value cells with the spatial distance less than a preset clustering threshold are merged into a peak value candidate region, and the maximum efficiency value and the number of cells of each peak value candidate region are counted to generate a peak value region candidate list; In the embodiment of the present application, 48 grid cells with positive deviation are screened from the efficiency deviation matrix, and the efficiency values (in the range of 92.6%-97.8%) of these cells are extracted. The standard deviation of the 48 efficiency values is 1.8 by using a standard deviation calculation module to calculate the standard deviation according to the standard deviation formula. The 48 cells with the standard deviation 1.8 less than the threshold value are all classified as potential peak value cells based on the preset stability threshold 2.0. Based on the working condition coordinates (such as X=65%-70%, Y=75%-80%; X=70%-75%, Y=80%-85%, etc.) of the potential peak value cells, the spatial distance between the cells is calculated by using the Euclidean distance formula, such as the interval of 5% in the X direction and the interval of 5% in the Y direction between the above two cells, the spatial distance is √(5%²+5%²)≈7.07%. The cells with the spatial distance 7.07% less than the threshold value are merged to form 3 peak value candidate regions based on the preset clustering threshold 8%. The maximum efficiency value and the number of cells of each candidate region are counted, such as the maximum value 97.8% of region 1 containing 8 cells, the maximum value 96.5% of region 2 containing 6 cells, and the maximum value 95.2% of region 3 containing 5 cells, to generate a peak value region candidate list.

[0050] Further, the efficiency change rate of the grid cells around each candidate region in the peak value region candidate list is extracted, which is specifically the efficiency gradient along the photovoltaic power ratio dimension and the energy storage SOC value dimension, and if the efficiency change rate is negative, it is determined that the candidate region is an effective peak value region. In the embodiment of the present application, the effectiveness of the three candidate regions in the peak region candidate list is verified one by one. Taking region 1 (X=65%-80%, Y=75%-90%) as an example, the efficiency values of the surrounding grid cells (X=60%-65%, Y=75%-90%; X=80%-85%, Y=75%-90%; X=65%-80%, Y=70%-75%; X=65%-80%, Y=90%-95%) are extracted. The efficiency gradient along the photovoltaic power ratio dimension (X axis) is calculated, the left side peripheral cell efficiency is 95.5%, the region 1 edge cell efficiency is 96.8%, and the gradient is (95.5%-96.8%) / 5%=-0.26% / %; the right side peripheral cell efficiency is 95.0%, the region 1 edge cell efficiency is 97.2%, and the gradient is (95.0%-97.2%) / 5%=-0.44% / %. The efficiency gradient along the energy storage SOC value dimension (Y axis) is calculated, the lower peripheral cell efficiency is 95.8%, the region 1 edge cell efficiency is 96.5%, and the gradient is (95.8%-96.5%) / 5%=-0.14% / %; the upper peripheral cell efficiency is 95.3%, the region 1 edge cell efficiency is 97.0%, and the gradient is (95.3%-97.0%) / 5%=-0.34% / %. The efficiency change rates in the four directions are all negative values, and region 1 is determined as an effective peak region; regions 2 and 3 are verified in the same way, and both meet the efficiency change rate being negative, and are determined as effective peak regions.

[0051] Further, the grid cells with negative deviation are screened out from the efficiency deviation matrix, the efficiency difference between the efficiency values of these cells and adjacent cells is calculated, the cells with an efficiency difference absolute value greater than a preset mutation threshold are marked as sudden drop boundary cells; based on the working condition coordinates of the sudden drop boundary cells, a sudden drop region boundary contour is constructed, and then according to whether the efficiency values of the grid cells in the sudden drop region boundary contour are continuously lower than the efficiency average value, a sudden drop candidate region is determined, the efficiency minimum value and the number of cells of each sudden drop candidate region are counted, and a sudden drop region candidate list is generated; In the embodiment of the present application, by screening 62 grid cells with negative deviation from the efficiency deviation matrix, the efficiency values of these cells and their adjacent cells (up, down, left and right) are extracted. The efficiency difference values of the cells and adjacent cells are calculated, such as X=30%-35%, Y=20%-25% cell efficiency 85.2%, adjacent X=35%-40%, Y=20%-25% cell efficiency 88.5%, difference absolute value 3.3%; adjacent X=30%-35%, Y=25%-30% cell efficiency 89.0%, difference absolute value 3.8%. The preset mutation threshold is 3.0%, and the cell with a difference absolute value greater than the threshold is marked as a sudden drop boundary cell. Based on the working condition coordinates of all sudden drop boundary cells (such as X=30%-35%, Y=20%-25%; X=30%-35%, Y=25%-30%; X=35%-40%, Y=20%-25%, etc.), the boundary fitting tool is used to connect these coordinate points to construct the sudden drop region boundary contour (X=30%-45%, Y=20%-35%). The efficiency values of the 24 grid cells in the contour are checked, which are all lower than the overall efficiency average 92.5% (range 85.2%-91.8%), and are determined as sudden drop candidate regions. The minimum efficiency of the region is 85.2% with 24 cells, and it is found that the cell efficiency in another contour (X=45%-50%, Y=20%-25%) is also lower than the average, forming a second sudden drop candidate region, with a minimum value of 86.5% and 5 cells, generating a sudden drop region candidate list.

[0052] Further, the effectiveness of each candidate region in the sudden drop region candidate list is verified, the efficiency change rate of the grid cells in the candidate region is extracted, and if the absolute value of the efficiency change rate in at least one dimension is greater than the preset sudden drop gradient threshold, the candidate region is determined as an effective sudden drop region.

[0053] In the embodiment of the present application, the effectiveness of two candidate regions in the sudden drop area candidate list is verified. Taking the first candidate region (X=30%-45%, Y=20%-35%) as an example, the efficiency values of all grid cells in the region are extracted, and the efficiency change rate along the photovoltaic power ratio dimension (X axis) is calculated. From X=30%-35% to X=40%-45%, the efficiency increases from 85.2% to 91.8%, and the change rate is (91.8%-85.2%) / (15%)≈0.44% / %, and the absolute value is 0.44% / %. The preset sudden drop gradient threshold is 0.3% / %, and the absolute value of the efficiency change rate in this dimension is greater than the threshold. The efficiency change rate along the energy storage SOC value dimension (Y axis) is calculated, from Y=20%-25% to Y=30%-35%, the efficiency increases from 85.2% to 90.5%, and the change rate is (90.5%-85.2%) / (15%)≈0.35% / %, and the absolute value is also greater than the threshold, so the candidate region is determined as an effective sudden drop region. The second candidate region (X=45%-50%, Y=20%-25%) is verified, and the efficiency along the photovoltaic power ratio dimension increases from 86.5% to 92.3% (change rate (92.3%-86.5%) / 5%≈1.16% / %), and the absolute value is greater than the threshold, so it is determined as an effective sudden drop region.

[0054] Further, the coupling efficiency test in step S455 is specifically a coupling efficiency grading test of the photovoltaic grid-connected energy storage type inverter under the corresponding working condition based on the multi-dimensional coupling efficiency surface graph after the super limit correction. If the coupling efficiency value under the corresponding working condition is >85% and the grid THD content is ≤5% and the corresponding power factor is ≥0.9, it is determined as first-class qualified; if the coupling efficiency value under the corresponding working condition is >85% and 5%<grid THD content≤8%, it is determined as second-class qualified; if the coupling efficiency value under the corresponding working condition is ≤85% or the component temperature rise data is >80K, it is determined as unqualified; at the same time, when the cumulative value of the component temperature rise data is 600-650K, the coupling efficiency-residual life curve of the photovoltaic grid-connected energy storage type inverter is generated, and the coupling efficiency test result of the photovoltaic grid-connected energy storage type inverter under the corresponding working condition is generated in combination with the grading result.

[0055] In the embodiment of the present application, 15 typical working conditions are selected for coupling efficiency classification test based on the multi-dimensional coupling efficiency surface map after transfinite correction. In a certain working condition, the coupling efficiency is 89%, the grid THD content is 3.2%, and the power factor is 0.92, which meets > 85% and ≤ 5% and ≥ 0.9, and is determined as first-class qualified. In another working condition, the coupling efficiency is 87% and the grid THD content is 6.5%, which meets > 85% and 5% < ≤ 8%, and is determined as second-class qualified. In a certain working condition, the coupling efficiency is 84%, which is directly determined as unqualified. In another working condition, the coupling efficiency is 86% but the component temperature rise is 82K, which is determined as unqualified. When the cumulative value of the component temperature rise is 600K, the remaining life of the inverter is 8000h; when it is 620K, the remaining life is 7500h; and when it is 650K, the remaining life is 7000h. The coupling efficiency-remaining life curve is drawn by corresponding these data, and the qualified grade, efficiency value and remaining life data of each working condition are summarized by combining the classification results and the curve data, to generate the coupling efficiency test results of the photovoltaic grid-connected energy storage type inverter under the corresponding working conditions.

[0056] Further, the present application also provides a photovoltaic grid-connected energy storage type inverter coupling efficiency test device, Figure 3 , Figure 3 is a structural schematic diagram of the photovoltaic grid-connected energy storage type inverter coupling efficiency test device in the embodiment, which is used for executing the photovoltaic grid-connected energy storage type inverter coupling efficiency test method as described above. The photovoltaic grid-connected energy storage type inverter coupling efficiency test device comprises a distributed multi-source collaborative unit, an edge computing collection terminal, a full-working-condition environment simulation room, a monitoring host computer and a data storage module. The distributed multi-source collaborative unit comprises a photovoltaic array simulator, an energy storage battery simulator, an intelligent power grid simulation unit and a distributed collaborative unit. The photovoltaic array simulator supports U-P / UI curve output and is used for simulating photovoltaic power under different illuminations. The energy storage battery simulator supports SOC 0%-100% dynamic adjustment and cycle attenuation simulation and is used for simulating the charge and discharge states of the energy storage system. The intelligent power grid simulation unit has voltage sag and harmonic injection functions and is used for simulating different power grid working conditions. The distributed collaborative unit serves as an intermediate node and realizes power closed-loop control of photovoltaic, energy storage and power grid. In the embodiment of the present application, in the distributed multi-source collaborative unit, the photovoltaic array simulator is set to have an output voltage range of 0-1500V and an output current of 0-60A. Through the built-in U-P / UI curve generation function, the photovoltaic array simulator simulates the output voltage of 600V and the output current of 10A (power of 6kW) under the illumination intensity of 200W / m 2 2 ​The photovoltaic power characteristic of the output voltage 800V and the current 50A (power 40kW) is output. The energy storage battery simulator is set to the rated capacity 100kWh, and the dynamic change from 10% (10kWh) to 100% (100kWh) is realized through an SOC adjustment program. When charging, the power is continuously input at 5kW, and the SOC is increased by 5% per hour. When discharging, the power is output at 8kW, and the SOC is decreased by 8% per hour. Meanwhile, the cyclic attenuation simulation function is started, and the capacity is attenuated by 1% after completing 100 charging and discharging cycles. After the 100th cycle, the rated capacity is reduced to 99kWh. The smart grid simulation unit is set to the fundamental wave voltage 380V and the frequency 50Hz. The voltage is reduced from 380V to 190V (reduced by 50%) within 0.5s through the voltage sudden drop module to realize fault simulation. The 3rd and 5th harmonics are injected through the harmonic injection module, so that the THD content of the power grid reaches 5%. The distributed collaborative unit collects the power data of the photovoltaic, energy storage and power grid in real time through the RS485 communication interface. When the photovoltaic output power is 40kW and the power grid load is 35kW, the energy storage is triggered to discharge 5kW, so that the total output power is matched with the load, and the power closed-loop control is realized.

[0057] The edge computing acquisition terminal comprises a multi-channel power analyzer and a temperature monitoring module. The multi-channel power analyzer supports 8-channel voltage / current acquisition with a sampling rate of 100Hz-1kHz adjustable, and is used to acquire input and output power data of the photovoltaic grid-connected energy storage type inverter. The temperature monitoring module adopts 128-point acquisition with a measurement range of-40℃-1000℃, and is used to acquire component temperature rise data. In the edge computing acquisition terminal, the multi-channel power analyzer is connected with the photovoltaic input terminal, the energy storage input terminal and the grid-connected output terminal of the inverter, and a total of 6 channels (2-way photovoltaic voltage / current, 2-way energy storage voltage / current and 2-way grid-connected voltage / current) are started. According to the rated power 50kW of the inverter, the sampling rate is set to 500Hz, and the voltage and current data are acquired once every 2ms. The photovoltaic input power, the energy storage charging and discharging power and the grid-connected output power are calculated in real time through a power calculation formula (power=voltage*current*power factor), and one set of average power data is generated every 1s. The temperature monitoring module adopts K-type thermocouple sensors, which are attached to the surfaces of 15 key components of the inverter IGBT module, the reactor and the electrolytic capacitor, and the remaining 113 acquisition points are reserved. The measurement interval is set to 1s / frame. When the temperature of the IGBT module rises to 70℃, the temperature value is recorded in real time and compared with the ambient temperature 25℃, and the temperature rise data of 45K are calculated. All temperature data are transmitted to the edge computing unit through wired transmission for local caching.

[0058] The full-condition environment simulation room comprises an altitude air pressure adjusting module, an illumination intensity simulation module and a temperature control module; the altitude air pressure adjusting module can cover the air pressure conditions of different geographical scenes corresponding to the plateau and the plain; the illumination intensity simulation module can reproduce the illumination changes in different periods and different seasons; the temperature control module can realize the temperature adjustment of-40 DEG C to 85 DEG C, and the temperature fluctuation is controlled within ±1 DEG C; In the embodiment of the application, the altitude air pressure adjusting module of the full-condition environment simulation room works cooperatively with a vacuum pump and a gas supplement valve to simulate the air pressure of 101 kPa at an altitude of 0 m (the plain), the air pressure of 70 kPa at an altitude of 3000 m (the plateau) and the air pressure of 50 kPa at an altitude of 5000 m, and the air pressure adjusting accuracy is controlled within ±2 kPa. The illumination intensity simulation module adopts an LED array light source, and realizes the adjustment of the illumination intensity from 0 W / m 2 (night) to 1200 W / m 2 (midday strong light) through a light adjusting controller, simulates the seasonal and periodical changes of the illumination of 300 W / m 2 at 8 o'clock in the morning in spring, 1000 W / m 2 at 12 o'clock in the noon in summer and 200 W / m2 at 4 o'clock in the afternoon in winter. The temperature control module realizes the temperature adjustment of-40 DEG C (severe cold) to 85 DEG C (high temperature) through a heating pipe and a refrigerating unit, sets the target temperature as 25 DEG C during the test, feeds back the indoor temperature in real time through a temperature sensor, starts the heating pipe when the temperature drops to 24.2 DEG C, starts the refrigerating unit when the temperature rises to 25.8 DEG C, strictly controls the temperature fluctuation within ±1 DEG C, and ensures the stability of the test environment.

[0059] The monitoring host computer communicates with each unit through LAN+4G standby dual link, can complete the initial coupling coefficient calculation, dynamic correction factor integration and coupling efficiency operation, and simultaneously supports the finite state machine FSM control logic; In the embodiment of the application, the monitoring host computer establishes a main communication connection with the distributed collaborative unit, the edge computing collection terminal and the full-condition environment simulation room through a LAN link (rate 1 Gbps), and simultaneously enables a 4G backup link (downlink rate 100 Mbps), and when the LAN link is interrupted, the 4G link is automatically switched within 500 ms. The host computer runs a coupling coefficient calculation program, inputs the photovoltaic weight 0.5, the energy storage weight 0.2, the grid weight 0.2 and the regional weight 0.1, combines the collected photovoltaic power ratio 80% and the energy storage SOC 60%, and calculates the initial coupling weight coefficient 0.82; through a dynamic correction factor integration module, the energy storage cycle attenuation coefficient 0.95, the grid harmonic influence coefficient 0.98 and the element aging coefficient 0.9 are inputted to generate the comprehensive correction factor 0.837; and then the coupling efficiency formula (coupling efficiency = output power / input power x comprehensive correction factor x 100%) is substituted to obtain the current working condition coupling efficiency 92.5%. At the same time, the host computer controls the test process according to the state sequence of “device initialization → working condition issuing → data collection → efficiency calculation → report output” based on the finite state machine FSM logic, and needs to meet the trigger conditions of power stability (fluctuation < 2%) and temperature stability (fluctuation < 1°C) before each state switching.

[0060] The data storage module adopts a local cache + cloud backup dual storage architecture, and simultaneously integrates a data backtracking retrieval module, can support querying historical data according to a time stamp, a working condition parameter and an inverter model, and thus realizes test full-link traceability.

[0061] In the embodiment of the application, the local cache unit of the data storage module adopts an industrial-grade solid state disk (capacity 1 TB), and the power data and the temperature data transmitted by the edge computing collection terminal are packaged and stored at a 10s period, each data packet contains a time stamp, working condition parameters (photovoltaic power, energy storage SOC and grid THD) and test data, the cache capacity supports 100 hours of rolling coverage, and after full capacity, the earliest non-abnormal data is automatically deleted. The cloud backup unit synchronizes the local cache data to a cloud server (capacity 10 TB) at 23 o'clock every day through an encrypted network (AES-256 encryption), realizes double data storage. The data backtracking retrieval module provides a multi-condition query function, inputs the time stamp “2024-05-20 14:30-15:00”, can retrieve all test data in the period; inputs the working condition parameters “photovoltaic power 30-40 kW, energy storage SOC 50-60%”, can filter out the test records matching the working conditions; inputs the inverter model “INV-50kW-01”, can retrieve all historical test data of the inverter of the model, and realizes test full-link traceability.

[0062] The above merely illustrates the embodiments of the present application but should not be taken as limitations. For those skilled in the art, the present application can have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. within the spirit and principles of the present application shall be included in the scope of the claims of the present application.

Claims

1. A method for testing the coupling efficiency of a photovoltaic grid-connected energy storage type inverter, characterized in that, Comprise the following steps: Step S1: Based on the distributed multi-source collaborative unit, edge computing acquisition terminal, full working condition environment simulation room and monitoring host computer, the corresponding distributed hardware test topology is built, and the multi-dimensional original data of the photovoltaic grid-connected energy storage type inverter under different working conditions is obtained based on the distributed hardware test topology, including photovoltaic-energy storage power data, energy storage charging and discharging data, power grid parameter data and temperature monitoring data, and the multi-dimensional original data is processed by Kalman noise reduction and time stamp synchronization to generate the time sequence aligned inverter multi-source test data set; Step S2: Based on the inverter multi-source test data set, the corresponding photovoltaic power ratio, energy storage SOC value, power grid THD content and component temperature rise data are obtained, and the corresponding inverter basic dimension parameter matrix is constructed combined with the obtained regional light wave intensity index; Based on the inverter basic dimension parameter matrix, coupling weighted operation is carried out to generate the initial coupling weighted coefficient of the inverter; Step S3: Based on the inverter multi-source test data set, the coupling dynamic correction factor is obtained, including the energy storage cycle attenuation coefficient, the power grid harmonic influence coefficient and the component aging coefficient, and the initial coupling weighted coefficient of the inverter is calculated based on the coupling dynamic correction factor, and the multi-dimensional coupling normalization correction weighted coefficient is generated; Step S4: Obtain the inverter input and output power data, and based on the multi-dimensional coupling normalization correction weighted coefficient combined with the coupling dynamic correction factor, the coupling efficiency of the inverter input and output power data is calculated to obtain the coupling efficiency value of the photovoltaic grid-connected energy storage type inverter under the current working condition; Based on the coupling efficiency value obtained under different working conditions, a multi-dimensional coupling efficiency surface graph is generated, and coupling over-limit quadratic correction and test are carried out according to the multi-dimensional coupling efficiency surface graph to generate the coupling efficiency test result of the photovoltaic grid-connected energy storage type inverter under the corresponding working condition.

2. The method for testing the coupling efficiency of a photovoltaic grid-connected energy storage type inverter according to claim 1, characterized in that, Step S1 includes the following steps: Step S11: Based on the distributed multi-source collaborative unit, edge computing acquisition terminal, full working condition environment simulation room and monitoring host computer, the corresponding distributed hardware test topology is built, wherein the distributed multi-source collaborative unit comprises photovoltaic array simulator, energy storage battery simulator, smart grid simulation unit and distributed collaborative unit, the edge computing acquisition terminal comprises multi-channel power analyzer and temperature monitoring module, the full working condition environment simulation room is provided with altitude pressure adjusting module, light intensity simulation module and temperature control module; Step S12: by monitoring the host computer issued device initialization instruction to drive distributed collaborative unit according to the sequence of full working condition environment simulation room→ energy storage battery simulator→ photovoltaic array simulator→ smart grid simulation unit to start the device; after the full working condition environment simulation room parameters stable, including temperature fluctuation <±1℃, pressure fluctuation <±2kPa and light fluctuation <±50W / m 2 , and continue for a predetermined length of time, then trigger the distributed collaborative unit to enter the power closed loop control mode; at the same time, according to the inverter capacity dynamic adjustment of multi-channel power analyzer corresponding sampling rate, and the temperature monitoring module sampling interval is set to 1-2s / frame; Step S13: Based on the power closed-loop control mode, the multi-dimensional original data of the photovoltaic grid-connected energy storage type inverter under different working conditions is collected and obtained by controlling the distributed hardware test topology, including photovoltaic-energy storage power data, energy storage charging and discharging data, power grid parameter data and temperature monitoring data; Step S14: The multi-dimensional original data is processed by Kalman noise reduction and time stamp synchronization to control the time deviation of each data source within 10ms, and the time sequence aligned inverter multi-source test data set is generated.

3. The method for testing the coupling efficiency of a photovoltaic grid-connected energy storage type inverter according to claim 1, characterized in that, Step S2 includes the following steps: Step S21: separate the photovoltaic simulation power subset and the energy storage charging and discharging power subset from the inverter multi-source test data set, and calculate the total input power data of the inverter according to the real-time photovoltaic output power data in the photovoltaic simulation power subset and the real-time energy storage power data in the energy storage charging and discharging power subset, wherein the total input power data of the inverter is the algebraic sum of the real-time photovoltaic output power data and the real-time energy storage power data, and when the energy storage is charging, the positive value is superimposed, and when the energy storage is discharging, the negative value is deducted; calculate the photovoltaic power ratio based on the real-time photovoltaic output power data and the total input power data of the inverter; Step S22: extract the charging and discharging current, voltage and time sequence data corresponding to the energy storage battery from the energy storage charging and discharging power subset, and calculate the cumulative charging and discharging capacity data of the energy storage battery based on the charging and discharging current, voltage and time sequence data, and calculate the energy storage SOC value in combination with the rated capacity data of the energy storage battery; Step S23: extract the power grid parameter subset from the inverter multi-source test data set, separate the power grid voltage waveform data, and perform Fourier transform processing on the power grid voltage waveform data to obtain the fundamental voltage component data and each harmonic voltage component data; calculate the total effective value of the power grid harmonics based on the fundamental voltage component data and each harmonic voltage component data, and generate the power grid THD content through the power grid THD content = total effective value of the power grid harmonics / fundamental voltage component effective value x 100%; Step S24: extract the temperature monitoring subset from the inverter multi-source test data set, separate the real-time temperature data and environmental temperature data corresponding to each component, and calculate the real-time temperature rise difference value data of the component based on the real-time temperature data and the environmental temperature data, and generate the component temperature rise data in combination with the sampling time interval of the temperature monitoring module; Step S25: obtain the regional light intensity parameter and construct the corresponding inverter basic dimension parameter matrix in combination with the photovoltaic power ratio, the energy storage SOC value, the power grid THD content and the component temperature rise data, and perform coupling weighting operation based on the inverter basic dimension parameter matrix to generate the initial coupling weighting coefficient of the inverter.

4. The method for testing the coupling efficiency of a photovoltaic grid-connected energy storage type inverter according to claim 3, characterized in that, Step S25 includes the following steps: Step S251: obtain the original light data of the region, including the global light radiation data of satellite remote sensing, the single-point real-time light data of ground distributed sensors and the historical light statistical data recorded by the weather station, extract the spectral radiation intensity data of different wave bands from the global light radiation data, and generate the light space-time distribution original matrix in combination with the time continuity of the single-point real-time light data; Step S252: standardize the light space-time distribution original matrix, calculate the corresponding light radiation cumulative value per unit time based on the standardized light space-time distribution original matrix, and obtain the geographic latitude and longitude data of the region to derive the light incidence angle change parameter of the region; Step S253: calculate the light intensity attenuation coefficient corresponding to different seasons and different time periods according to the light incidence angle change parameter and the historical light statistical data, and combine the light intensity attenuation coefficient with the real-time light radiation intensity data to generate the real-time light effective intensity parameter; Step S254: Time domain analysis is performed on the real-time light effective intensity parameter to extract the corresponding light intensity peak value, valley value and fluctuation times per unit time, and the light intensity fluctuation amplitude parameter is calculated, and the light stability index is generated based on the light intensity fluctuation amplitude parameter and the obtained regional weather type data; the real-time light effective intensity parameter, the light stability index and the light incidence angle change parameter are coupled and calculated to obtain the regional light wave intensity parameter; Step S255: The regional light wave intensity parameter is constructed into a corresponding inverter basic dimension parameter matrix with photovoltaic power proportion, energy storage SOC value, power grid THD content and component temperature rise data in different test periods, including daytime peak period, daytime flat period and night valley period, wherein the regional light wave intensity parameter, photovoltaic power proportion, energy storage SOC value, power grid THD content and component temperature rise data are used as matrix row vectors, different test periods are used as matrix column vectors, and the specific values of each parameter in different test periods are used as elements; the analytic hierarchy process is used to assign weights to each column period weight and each row parameter weight in the inverter basic dimension parameter matrix; the column period weight is determined according to the inverter running load proportion of each period, and the row parameter weight is determined according to the influence degree of each parameter on the coupling efficiency; a coupling weighted operation model is constructed based on the matrix element value, column period weight and row parameter weight to generate the initial coupling weighted coefficient of the inverter by summing the product of element value, row weight and column weight.

5. The method for testing the coupling efficiency of photovoltaic grid-connected energy storage type inverter according to claim 1, characterized in that, Step S3 includes the following steps: Step S31: Extracting energy storage charge and discharge time sequence data from the inverter multi-source test data set, including charge and discharge start time, end time, current change curve and voltage fluctuation data; calculating the single cycle duration based on the charge and discharge start and end time, determining the effective charge and discharge capacity combining the current change curve and voltage fluctuation data, and generating the energy storage cycle depth parameter by accumulating the ratio of effective charge and discharge capacity to rated capacity; then deriving the energy storage cycle attenuation coefficient according to the energy storage cycle depth parameter and cycle number statistical data; Step S32: Separating the power grid parameter subset from the inverter multi-source test data set and extracting power grid voltage harmonic frequency spectrum data, fundamental frequency stability data and voltage step-down recovery time data; obtaining harmonic time domain waveform characteristics by inverse Fourier transform of the power grid voltage harmonic frequency spectrum data, and calculating the harmonic energy proportion parameter combining the fundamental frequency stability data; calculating the interference coefficient to the inverter efficiency based on the harmonic energy proportion parameter and the voltage step-down recovery time data, and obtaining the power grid harmonic influence coefficient after normalization processing of the interference coefficient; Step S33: Extracting temperature monitoring subset from the inverter multi-source test data set, including real-time temperature, temperature rise rate and temperature fluctuation period data of IGBT and capacitor key components, calculating the component relative temperature rise parameter based on the real-time temperature data combining the environmental temperature data, and calculating the component thermal cycle damage factor based on the temperature rise rate and temperature fluctuation period data, and generating the component aging coefficient by comparing and analyzing the component thermal cycle damage factor with the rated life parameter corresponding to the component; Step S34: The energy storage cycle attenuation coefficient, the power grid harmonic influence coefficient, and the element aging coefficient are integrated in the form of a row vector to generate a coupled dynamic correction factor. The corresponding influence degree distribution factor weight vector of the photovoltaic grid-connected energy storage type inverter is distributed based on the coupled dynamic correction factor. Meanwhile, the tensor product calculation is performed based on the coupled dynamic correction factor, the factor weight vector, and the initial coupling weighting coefficient of the inverter to obtain the multi-dimensional coupling correction weighting coefficient of the inverter. Step S35: Different power test points are selected, and the multi-dimensional coupling correction weighting coefficient corresponding to the power test points is obtained. The multi-dimensional coupling normalization correction calculation is performed based on the multi-dimensional coupling correction weighting coefficient corresponding to the different power test points to generate the multi-dimensional coupling normalization correction weighting coefficient.

6. The method for testing the coupling efficiency of a photovoltaic grid-connected energy storage type inverter according to claim 5, characterized in that, Step S4 includes the following steps: Step S41: Obtain the inverter input and output power data, including input power data and output power data, wherein the input power data includes the photovoltaic input power output by the photovoltaic array simulator and the charge and discharge power of the energy storage battery simulator, and the output power data is the inverter grid-connected alternating current power; Step S42: Perform time integration processing on the inverter input and output power data to calculate the total input energy and total output energy in a unit period with 5-10s as the integration period. The photovoltaic weight component, energy storage weight component, power grid weight component, temperature rise weight component, and region weight component are extracted from the multi-dimensional coupling normalization correction weighting coefficient. The photovoltaic power ratio coefficient and the energy storage power ratio coefficient are calculated based on the photovoltaic input power and the charge and discharge power. The total input energy is corrected once based on the photovoltaic weight component, the energy storage weight component, and the photovoltaic power ratio coefficient and the energy storage power ratio coefficient to generate the first correction input energy; Step S43: Obtain the power grid working condition coefficient, the temperature rise deviation coefficient, and the region light fluctuation coefficient. The first correction input energy is corrected twice by introducing the power grid weight component, the temperature rise weight component, and the region weight component to generate the corrected input energy. Specifically, the corrected input energy = the first correction input energy × (1 + the power grid weight component × the power grid working condition coefficient + the temperature rise weight component × the temperature rise deviation coefficient + the region weight component × the region light fluctuation coefficient). At the same time, the corrected input energy is compared with the standard input energy range of the inverter of the same power level. If it exceeds the range ±10%, return to step S42 to re-extract the weight component; Step S44: Calculate the coupling efficiency based on the corrected input energy and the total output energy and the coupled dynamic correction factor to obtain the coupling efficiency value of the photovoltaic grid-connected energy storage type inverter under the current working condition; Step S45: Generate a multi-dimensional coupling efficiency surface graph based on the coupling efficiency values obtained under different working conditions. Perform coupling over-limit secondary correction and testing based on the multi-dimensional coupling efficiency surface graph to generate the coupling efficiency test result of the photovoltaic grid-connected energy storage type inverter under the corresponding working condition.

7. The method for testing the coupling efficiency of a photovoltaic grid-connected energy storage type inverter according to claim 6, characterized in that, Step S45 includes the following steps: Step S451: Construct a working condition parameter-coupling efficiency mapping dataset by extracting different working condition parameters, including photovoltaic power ratio, energy storage SOC value, grid THD content, component temperature rise data, and regional light wave intensity parameters, and the coupling efficiency value under the corresponding working condition; calculate the correlation coefficient between each working condition parameter based on the working condition parameter-coupling efficiency mapping dataset, and filter out the core working condition parameters that have a significant impact on the coupling efficiency through the correlation coefficient, and then generate a curved surface dimension division basis combined with the value range of the core working condition parameters; Step S452: Data completion is performed on the working condition parameter-coupling efficiency mapping dataset to fill in the missing working condition-efficiency data points; based on the completed dataset, an initial multi-dimensional coupling surface model is constructed with the core working condition parameters as the coordinate axes, including X-axis for photovoltaic power ratio, Y-axis for energy storage SOC value, and Z-axis for coupling efficiency value; the slope change of adjacent data points in the initial multi-dimensional coupling surface model is calculated to generate a curved surface smoothness parameter, and if the curved surface smoothness parameter is lower than a preset threshold, the initial multi-dimensional coupling surface model is optimized through a Gaussian filtering algorithm to obtain a multi-dimensional coupling efficiency curved surface graph; Step S453: Extract the peak region and step-down region corresponding to the coupling efficiency from the multi-dimensional coupling efficiency curved surface graph, calculate the efficiency average of the peak region and the minimum efficiency of the step-down region, and generate an efficiency fluctuation amplitude parameter based on the difference between the two; determine the coupling efficiency overrun judgment threshold value in combination with the inverter rated efficiency range, and then mark the overrun working condition points in the curved surface graph that exceed the threshold range through comparison between the efficiency fluctuation amplitude parameter and the coupling efficiency overrun judgment threshold value; Step S454: Root cause analysis is performed on the marked overrun working condition points to extract the working condition parameter data and coupling dynamic correction factor corresponding to the overrun working condition points, and calculate the deviation rate of the correction factor under the overrun working condition; calculate the contribution degree parameter of each correction factor to the efficiency overrun based on the correlation between the deviation rate and the overrun efficiency value; determine the adjustment amplitude of the correction factor according to the contribution degree parameter, and iteratively optimize the coupling dynamic correction factor; re-substitute the optimized correction factor into the calculation to obtain the optimized coupling efficiency value; update the multi-dimensional coupling efficiency curved surface graph based on the optimized coupling efficiency value, and check again whether there are overrun working condition points, and if there are still, repeat the step until the coupling efficiency of all working condition points in the curved surface graph is within the rated efficiency range; Step S455: Based on the multi-dimensional coupling efficiency curved surface graph after the overrun correction, perform coupling efficiency testing to generate the coupling efficiency test result of the photovoltaic grid-connected energy storage type inverter under the corresponding working condition.

8. The method for testing the coupling efficiency of a photovoltaic grid-connected energy storage type inverter according to claim 7, characterized in that, The step S453 of extracting the peak region and step-down region corresponding to the coupling efficiency from the multi-dimensional coupling efficiency curved surface graph includes the following steps: Grid processing is performed on the three-dimensional data of the multi-dimensional coupling efficiency curved surface graph, the curved surface is divided into a plurality of uniform grid cells, and the coupling efficiency value and corresponding working condition coordinates of each grid cell are extracted; the overall efficiency average of the curved surface is calculated based on the coupling efficiency value of the grid cell, and the efficiency deviation matrix is generated by combining the coupling efficiency value and the average deviation of each grid cell; Screening grid cells with positive deviation from the efficiency deviation matrix, calculating the standard deviation of the efficiency values of these cells, and classifying grid cells with standard deviation of efficiency values less than a preset stability threshold as potential peak value cells; based on the operating condition coordinates of the potential peak value cells, calculating the spatial distance between cells, and merging potential peak value cells with a spatial distance less than a preset clustering threshold into peak value candidate regions, while simultaneously counting the maximum efficiency value and the number of cells of each peak value candidate region to generate a peak value region candidate list; Verifying the effectiveness of each candidate region in the peak value region candidate list, extracting the efficiency change rate of the grid cells around the candidate region, specifically the efficiency gradient along the photovoltaic power ratio dimension and the energy storage SOC value dimension, and if the efficiency change rate is negative, determining that the candidate region is an effective peak value region; Screening grid cells with negative deviation from the efficiency deviation matrix, calculating the efficiency difference between these cells and adjacent cells, and marking cells with an absolute value of the efficiency difference greater than a preset mutation threshold as sudden drop boundary cells; based on the operating condition coordinates of the sudden drop boundary cells, constructing a sudden drop region boundary contour, and then determining a sudden drop candidate region according to whether the efficiency values of the grid cells within the sudden drop region boundary contour are continuously lower than the average efficiency, and counting the minimum efficiency value and the number of cells of each sudden drop candidate region to generate a sudden drop region candidate list; Verifying the effectiveness of each candidate region in the sudden drop region candidate list, extracting the efficiency change rate of the grid cells within the candidate region, and if the absolute value of the efficiency change rate in at least one dimension is greater than a preset sudden drop gradient threshold, determining that the candidate region is an effective sudden drop region.

9. The method for testing the coupling efficiency of a photovoltaic grid-connected energy storage type inverter according to claim 7, characterized in that, The coupling efficiency test in step S455 is a coupling efficiency grading test of the photovoltaic grid-connected energy storage inverter under the corresponding operating condition based on the multi-dimensional coupling efficiency surface graph after the super-limit correction, and if the coupling efficiency value under the corresponding operating condition is > 85% and the grid THD content is ≤ 5% and the corresponding power factor is ≥ 0.9, it is determined to be first-class qualified; if the coupling efficiency value under the corresponding operating condition is > 85% and 5% < the grid THD content ≤ 8%, it is determined to be second-class qualified; if the coupling efficiency value under the corresponding operating condition is ≤ 85% or the component temperature rise data is > 80K, it is determined to be unqualified; at the same time, when the cumulative value corresponding to the component temperature rise data is 600-650K, the coupling efficiency-residual life curve is generated with the residual life of the photovoltaic grid-connected energy storage inverter, and the coupling efficiency test result of the photovoltaic grid-connected energy storage inverter under the corresponding operating condition is generated in combination with the grading result.

10. A device for testing the coupling efficiency of a photovoltaic grid-connected energy storage type inverter, characterized in that it comprises: The device for performing the method for testing the coupling efficiency of the photovoltaic grid-connected energy storage inverter as claimed in any one of claims 1-9 comprises a distributed multi-source collaborative unit, an edge computing collection terminal, a full-condition environment simulation room, a monitoring host computer, and a data storage module: The distributed multi-source cooperative unit comprises a photovoltaic array simulator, an energy storage battery simulator, an intelligent power grid simulation unit and a distributed cooperative unit; the photovoltaic array simulator supports U-P / UI curve output and is used for simulating photovoltaic power under different illuminations; the energy storage battery simulator supports SOC 0%-100% dynamic adjustment and cycle attenuation simulation and is used for simulating the charge and discharge state of an energy storage system; the intelligent power grid simulation unit has voltage sag and harmonic injection functions and is used for simulating different power grid working conditions; and the distributed cooperative unit serves as an intermediate node and realizes power closed-loop control of photovoltaic, energy storage and power grid; The edge computing collection terminal comprises a multi-channel power analyzer and a temperature monitoring module; the multi-channel power analyzer supports 8-channel voltage / current collection, has a sampling rate of 100Hz-1kHz and can be adjusted, and is used for acquiring input and output power data of a photovoltaic grid-connected energy storage type inverter; the temperature monitoring module adopts 128-point collection, has a measurement range of-40℃-1000℃, and is used for collecting component temperature rise data; The full working condition environment simulation room comprises an altitude pressure adjustment module, an illumination intensity simulation module and a temperature control module; the altitude pressure adjustment module can cover the pressure conditions of different geographical scenes of a plateau and a plain; the illumination intensity simulation module can reproduce illumination changes in different periods and different seasons; and the temperature control module can realize temperature adjustment of-40℃-85℃ and control temperature fluctuation within ±1℃; The monitoring host computer communicates with each unit through LAN+4G backup dual links, can complete initial coupling coefficient calculation, dynamic correction factor integration and coupling efficiency operation, and simultaneously supports finite state machine FSM control logic; The data storage module adopts a local cache+cloud backup dual storage architecture, simultaneously integrates a data backtracking retrieval module, can support historical data query according to a time stamp, a working condition parameter and an inverter model, and thus realizes test full-link traceability.

Images