Method and apparatus for controlling a photovoltaic inverter circuit

Abstract

The application relates to a control method and device of a photovoltaic inverter circuit, which comprises the following steps: collecting operation data and environment data of the photovoltaic inverter in real time as a first data set; inputting the data in the first data set into a first intelligent agent to output a reference current and a reference voltage; inputting the data in the first data set into a second intelligent agent to output a harmonic compensation voltage component; inputting the data in the first data set into a third intelligent agent to output the health degree of a power switching device; and determining a first PWM duty ratio and a second PWM duty ratio based on at least part of the first data set, the reference current, the reference voltage, the harmonic compensation voltage component and the health degree. By adjusting the first PWM duty ratio and the second PWM duty ratio based on different maximum power point tracking, grid distortion and the health degree of the power switching device, the maximum power point tracking, the grid distortion and the health degree of the power switching device can be synergistically optimized.

Description

Technical Field

[0001] This application relates to the field of photovoltaic inverter technology, and in particular to a control method and device for a photovoltaic inverter circuit. Background Technology

[0002] With the development of photovoltaic power generation devices, the performance of three-phase grid-connected photovoltaic inverters, as the interface device between photovoltaic arrays and the power grid, is directly related to power generation revenue and the safe and stable operation of the power grid.

[0003] In traditional technologies, proportional-integral controllers are often used for control. However, when photovoltaic inverters face multiple control objectives simultaneously, it is difficult to achieve multi-objective collaborative optimization. Summary of the Invention

[0004] Therefore, it is necessary to provide a control method, device, computer equipment, computer-readable storage medium, and computer program product for photovoltaic inverter circuits that can achieve multi-objective collaborative optimization, in order to address the above-mentioned technical problems.

[0005] In a first aspect, this application provides a control method for a photovoltaic inverter circuit, wherein the photovoltaic inverter circuit includes a boost converter circuit and an inverter circuit, the boost converter circuit being electrically connected to a photovoltaic array, and the inverter circuit being electrically connected to the boost converter circuit; the method includes:

[0006] Real-time acquisition of photovoltaic inverter operating data and environmental data is used as the primary dataset;

[0007] The data in the first dataset is input into the first intelligent agent, which outputs the reference current and reference voltage for the current operating condition; the reference current and reference voltage refer to the current and voltage corresponding to the target maximum power point under the current operating condition, respectively.

[0008] The data in the first dataset is input into the second intelligent agent, which outputs the harmonic compensation voltage component of the current operating condition.

[0009] The data in the first dataset is input into the third intelligent agent, and the health status of the power switching devices in the photovoltaic inverter circuit is output.

[0010] Based on at least a portion of the first dataset, the reference current, the reference voltage, the harmonic compensation voltage component, and the health status, the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit are determined.

[0011] In one embodiment, the first dataset includes the current power generation, the total harmonic distortion rate of the grid after harmonic compensation voltage component compensation, and the grid-side voltage;

[0012] Determining the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit based on at least a portion of the first dataset, the reference current, the reference voltage, the harmonic compensation voltage component, and the health status includes:

[0013] The reference PWM duty cycle of the boost converter circuit is determined based on the reference voltage.

[0014] Based on the power generation, the total harmonic distortion rate of the grid, the grid-side voltage, and the health status, the output weights of the first agent, the second agent, and the third agent are determined; the sum of the output weights of the three agents is 1.

[0015] The deduction coefficient is determined based on the health score, and the health score is positively correlated with the deduction coefficient; the deduction coefficient is greater than or equal to 0 and less than or equal to 1.

[0016] The first PWM duty cycle of the inverter circuit is determined based on the output weight, the reference current, the derating factor, and the harmonic compensation voltage component.

[0017] The second PWM duty cycle of the boost converter circuit is determined based on the output weight, the reference PWM duty cycle, and the derating factor.

[0018] In one embodiment, determining the first PWM duty cycle of the inverter circuit based on the output weight, the reference current, the derating factor, and the harmonic compensation voltage component includes:

[0019] The target current is determined based on the reference current, the derating factor, and the output weight of the third agent;

[0020] The first PWM duty cycle is determined based on the target current and the harmonic compensation voltage component.

[0021] In one embodiment, determining the target current based on the reference current, the derating factor, and the output weights of the third agent includes:

[0022] The target current is calculated based on the following formula:

[0023]

[0024] in, Indicates the reference current. Indicates the target current. This represents the output weights of the third agent. This represents the reduction coefficient.

[0025] In one embodiment, determining the second PWM duty cycle of the boost converter circuit based on the output weight, the reference PWM duty cycle, and the derating factor includes:

[0026] Based on the output weights of the first agent, the second agent, the third agent, the reference pulse width modulation signal, and the derating factor, the second PWM duty cycle is determined using the following formula:

[0027]

[0028] in, Indicates the reference PWM duty cycle. Indicates the duty cycle of the second PWM. This represents the output weight of the first agent. This represents the output weights of the third agent. This represents the reduction coefficient. This represents the output weight of the second agent.

[0029] In one embodiment, determining the output weights of the first agent, the second agent, and the third agent based on the power generation, the total harmonic distortion rate connected to the grid, the grid-side voltage, and the health status includes:

[0030] The power generation, total harmonic distortion rate, grid-side voltage, and health status under the current operating conditions are used as the state inputs for the reinforcement learning algorithm.

[0031] The reward function of the reinforcement learning algorithm is set based on the power generation, the total harmonic distortion rate connected to the grid, and the health status.

[0032] The output weights of the first agent, the second agent, and the third agent are used as the output of the reinforcement learning algorithm.

[0033] In one embodiment, the first dataset includes photovoltaic-side current, photovoltaic-side voltage, ambient temperature, and ambient humidity;

[0034] The step of inputting data from the first dataset into the first intelligent agent and outputting the reference current and reference voltage for the current operating condition includes:

[0035] The photovoltaic-side current, photovoltaic-side voltage, ambient temperature, and ambient humidity are input into a first intelligent agent, which determines the reference current and reference voltage based on a fuzzy logic control algorithm; and / or

[0036] The first dataset includes grid-side voltage and grid-side current;

[0037] The step of inputting data from the first dataset into the second intelligent agent and outputting the harmonic compensation voltage component of the current operating condition includes:

[0038] The grid-side voltage and grid-side current are input into the second intelligent agent, which determines the harmonic compensation voltage component based on a trained first neural network model; and / or

[0039] The first dataset includes the on-state voltage drop, off-state overvoltage, and casing temperature of all power switching devices in the photovoltaic inverter circuit;

[0040] The step of inputting data from the first dataset into the third intelligent agent and outputting the health status of the power switching devices in the photovoltaic inverter circuit includes:

[0041] The on-state voltage drop, off-state overvoltage, and casing temperature of all power switching devices in the photovoltaic inverter circuit are input to a third intelligent agent. Based on the trained second neural network model, the third intelligent agent outputs the initial health status of each power switching device. Based on the initial health status of each power switching device, the third intelligent agent determines the health status of the power switching devices in the photovoltaic inverter circuit.

[0042] In one embodiment, upon obtaining an initial health status, the third agent determines the health status of the power switching devices in the photovoltaic inverter circuit based on the initial health status of each power switching device, including:

[0043] The minimum initial health value among all power switching devices is taken as the health value of the power switching devices in the photovoltaic inverter circuit.

[0044] In one embodiment, before determining the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit based on at least a portion of the first dataset, the reference current, the reference voltage, the harmonic compensation voltage component, and the health status, the method further includes:

[0045] If the health status is lower than a preset threshold, an early warning signal will be issued;

[0046] If the health status is greater than or equal to a preset threshold, proceed to the step of determining the first PWM duty cycle and the second PWM duty cycle.

[0047] Secondly, this application also provides a control device for a photovoltaic inverter circuit, the photovoltaic inverter circuit including a boost converter circuit and an inverter circuit, the boost converter circuit being electrically connected to a photovoltaic array, and the inverter circuit being electrically connected to the boost converter circuit; the device includes:

[0048] The data acquisition module is used to acquire real-time operating data and environmental data of the photovoltaic inverter as the first dataset;

[0049] The first intelligent agent is used to output the reference current and reference voltage of the current operating condition based on the data in the first dataset; the reference current and the reference voltage refer to the current and voltage corresponding to the target maximum power point under the current operating condition, respectively.

[0050] The second intelligent agent is used to output the harmonic compensation voltage component of the current operating condition based on the data in the first dataset.

[0051] A third intelligent agent is used to output the health status of the power switching devices in the photovoltaic inverter circuit based on the data in the first dataset.

[0052] The data processing module is used to determine the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit based on at least a portion of the first dataset, the reference current, the reference voltage, the harmonic compensation voltage component, and the health status.

[0053] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described method.

[0054] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described method.

[0055] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described method.

[0056] The aforementioned control method, apparatus, computer equipment, computer-readable storage medium, and computer program product for photovoltaic inverter circuits, by acquiring the reference current and reference voltage corresponding to the maximum power point under the current operating condition, the harmonic compensation voltage component under the current operating condition, and the health of the power switching devices in the photovoltaic inverter circuit, and determining the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit based on the reference current, reference voltage, harmonic compensation voltage component, and health, achieves multi-objective consideration of maximum power point tracking, grid distortion, and the health of the power switching devices. By adjusting the first PWM duty cycle and the second PWM duty cycle based on different maximum power point tracking, grid distortion, and the health of the power switching devices, coordinated optimization of maximum power point tracking, grid distortion, and the health of the power switching devices can be achieved. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0058] Figure 1 This is an application environment diagram of the control method for a photovoltaic inverter circuit in one embodiment.

[0059] Figure 2 This is a schematic diagram of a photovoltaic inverter circuit in one embodiment.

[0060] Figure 3 This is one of the flowcharts illustrating the control method of a photovoltaic inverter circuit in one embodiment.

[0061] Figure 4 This is a second flowchart illustrating the control method of a photovoltaic inverter circuit in one embodiment.

[0062] Figure 5 This is the third flowchart illustrating the control method of a photovoltaic inverter circuit in one embodiment.

[0063] Figure 6 This is the fourth flowchart illustrating the control method of a photovoltaic inverter circuit in one embodiment.

[0064] Figure 7 This is the fifth flowchart illustrating the control method of a photovoltaic inverter circuit in one embodiment.

[0065] Figure 8 This is one of the structural block diagrams of the control device for a photovoltaic inverter circuit in one embodiment.

[0066] Figure 9 This is the second structural block diagram of the control device for a photovoltaic inverter circuit in one embodiment.

[0067] Figure 10 This is an internal structural diagram of a computer device in one embodiment.

[0068] Figure label:

[0069] Photovoltaic inverter circuit: 102; Photovoltaic module: 104; Controller: 106; Boost converter circuit: 108; Inverter circuit: 110; Filter circuit: 112; Data acquisition module: 202; First intelligent agent: 204; Second intelligent agent: 206; Third intelligent agent: 208; Data processing module: 210; Early warning module: 212. Detailed Implementation

[0070] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0071] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0072] The control method for photovoltaic inverter circuits provided in this application embodiment can be applied to, for example... Figure 1 The photovoltaic system includes a photovoltaic inverter circuit 102, a photovoltaic module 104, and a controller 106. The photovoltaic module 104 is electrically connected to the photovoltaic inverter circuit 102, which is used for electrical connection to the power grid. The controller 106 is used to control the photovoltaic module 104 and the photovoltaic inverter circuit 102. The controller 106 can adopt a heterogeneous architecture of a digital signal processor (DSP) and a field-programmable gate array (FPGA). The DSP is used to execute various algorithms in the control method of the photovoltaic inverter circuit, and the FPGA is used to acquire a first dataset and generate pulse signals based on the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit.

[0073] In one exemplary embodiment, this application provides a control method for a photovoltaic inverter circuit, see reference. Figure 2 The photovoltaic inverter circuit includes a boost converter circuit 108 and an inverter circuit 110. The boost converter circuit 108 is electrically connected to the photovoltaic array, and the inverter circuit 110 is electrically connected to the boost converter circuit 108. The photovoltaic inverter circuit can convert the direct current generated by the photovoltaic array into alternating current to supply power to the power grid.

[0074] The photovoltaic array can be an array of photovoltaic modules, and the boost converter circuit 108 can be a boost converter circuit, which can boost a lower DC voltage to a higher stable voltage. The inverter circuit 110 can be a three-phase three-level neutral point clamped inverter circuit. Neutral point clamping (NPC) uses a specific topology to clamp the neutral point potential of the bridge arms in the three-phase three-level neutral point clamped inverter circuit to the neutral point voltage to prevent damage due to overvoltage. The boost converter circuit and the three-phase three-level neutral point clamped inverter circuit can use existing circuit structures, which will not be described in detail here.

[0075] See Figure 2 The photovoltaic inverter circuit may also include a filter circuit 112. The filter circuit 112 can be connected between the inverter circuit 110 and the power grid. The filter circuit 112 can be used to convert the pulse wave output by the inverter circuit 110 into a sine wave and feed it into the power grid. No restrictions are placed on the structure of the filter circuit 112 here.

[0076] Applying this method to Figure 1 Taking the controller in the example as an illustration, please refer to... Figure 3 The method includes steps S102 and S110, wherein:

[0077] Step S102: Real-time acquisition of photovoltaic inverter operation data and environmental data as the first dataset.

[0078] Operational data may include data corresponding to the electrical signals generated during the operation of the photovoltaic inverter. Environmental data may include data corresponding to the environment in which the photovoltaic inverter operates.

[0079] Operational data can be acquired through a sampling circuit, while environmental data can be acquired through sensors. Both the sampling circuit and sensors can acquire data based on a preset sampling frequency; for example, the preset sampling frequency is 10kHz, but no limitation is imposed on the sampling frequency here.

[0080] The controller can preprocess the data in the first dataset to eliminate noise interference. For example, the data in the first dataset can be processed by moving average filtering. No restrictions are placed on the preprocessing method.

[0081] Step S104: Input the data in the first dataset into the first intelligent agent and output the reference current and reference voltage of the current operating condition.

[0082] Here, the reference current and reference voltage refer to the current and voltage corresponding to the target maximum power point under the current operating conditions, respectively. It is understood that the photovoltaic array in this application will automatically track the maximum power point. If the operating data collected by the sampling circuit is not the operating data for the maximum power point, it indicates that the photovoltaic array is tracking the maximum power point and is currently near it.

[0083] Step S106: Input the data in the first dataset into the second agent and output the harmonic compensation voltage component of the current operating condition.

[0084] Harmonic compensation voltage component is a portion of the voltage actively output by the inverter circuit to eliminate harmonics in the grid-connected current. It is an additional harmonic compensation voltage component superimposed on the original fundamental voltage, which generates a harmonic current in the inductor that is equal in magnitude but opposite in direction to the original grid harmonic current, thus achieving mutual cancellation. If no harmonic compensation voltage component is output for superposition, the photovoltaic inverter can only rely on the fundamental current loop to control the output current. Since the fundamental current loop cannot actively cancel harmonic components in the grid voltage, the absence of a harmonic compensation voltage component may lead to higher distortion of the current waveform after grid connection.

[0085] Step S108: Input the data from the first dataset into the third intelligent agent and output the health status of the power switching devices in the photovoltaic inverter circuit.

[0086] The health status of a power switching device is an indicator that characterizes the degree of health of the power switching device. The higher the health status, the healthier the power switching device.

[0087] Power switching devices are important components in photovoltaic inverter circuits, directly affecting the output quality and efficiency of photovoltaic arrays. Once the performance of power switching devices degrades, it may increase the on-resistance and losses, causing the power switching devices to overheat severely. Furthermore, it may cause waveform distortion, which may lead to failure to meet the grid-connected harmonic standards, and may even cause thermal runaway due to uneven current distribution.

[0088] Step S110: Based on at least a portion of the first dataset, reference current, reference voltage, harmonic compensation voltage component, and health status, determine the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit.

[0089] The degree of grid distortion is highly correlated with the first PWM duty cycle of the inverter circuit. Specifically, the first PWM duty cycle of the inverter circuit determines the modulation wave containing harmonic compensation voltage components. These harmonic compensation voltage components can cancel harmonics in the grid, thereby reducing the degree of grid distortion. Maximum power point tracking (MPPT) is highly correlated with the second PWM duty cycle of the boost converter circuit. Specifically, the boost converter circuit is connected between the photovoltaic array and the DC bus. By changing the second PWM duty cycle, the output voltage of the photovoltaic array can be changed, thus achieving MPPT. The health of the power switching devices is related to both the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit. When the health of the power switching devices is low, power needs to be limited to prioritize the protection of the power switching devices. At this time, harmonics may be high, and MPPT may not be achieved, thus affecting both the first and second PWM duty cycles. By determining the first and second PWM duty cycles based on reference current, reference voltage, harmonic compensation voltage components, and health, multi-objective collaborative optimization of MPPT, grid distortion, and power switching devices can be achieved.

[0090] The aforementioned control method for the photovoltaic inverter circuit obtains the reference current and reference voltage corresponding to the maximum power point under the current operating condition, the harmonic compensation voltage component under the current operating condition, and the health of the power switching devices in the photovoltaic inverter circuit. Based on the reference current, reference voltage, harmonic compensation voltage component, and health, it determines the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit. This achieves multi-objective consideration of maximum power point tracking, grid distortion, and the health of the power switching devices. By adjusting the first PWM duty cycle and the second PWM duty cycle based on different maximum power point tracking, grid distortion, and the health of the power switching devices, it is possible to achieve coordinated optimization of maximum power point tracking, grid distortion, and the health of the power switching devices.

[0091] In one embodiment, the first dataset includes photovoltaic-side current, photovoltaic-side voltage, ambient temperature, and ambient humidity.

[0092] Among them, photovoltaic side current and photovoltaic side voltage are operational data, while ambient temperature and ambient humidity are environmental data.

[0093] The data in the first dataset is input into the first intelligent agent, which outputs the reference current and reference voltage for the current operating condition, including:

[0094] The photovoltaic side current, photovoltaic side voltage, ambient temperature, and ambient humidity are input into the first intelligent agent, which determines the reference current and reference voltage based on a fuzzy logic control algorithm.

[0095] In this embodiment, the first intelligent agent first quantizes the input parameters. Next, based on a membership function, it determines the membership degree of the quantized values. Then, based on the membership degrees of each value and preset fuzzy rules, it performs fuzzy inference to obtain fuzzy conclusions. Finally, it converts the fuzzy conclusions into actual numerical values ​​to obtain the reference current and reference voltage.

[0096] The first agent can be a maximum power point tracking agent, used to output the current and voltage at the maximum power point, i.e., the reference current and reference voltage. Understandably, if the power switching device has low health, it may discard some power to protect it. Even if the second PWM duty cycle of the boost converter circuit is determined based on the output of the first agent, it may still fail to reach the maximum power point.

[0097] By obtaining reference current and reference voltage, a basis can be provided for controlling photovoltaic modules to reach their maximum power point.

[0098] In one embodiment, the first dataset includes grid-side voltage and grid-side current.

[0099] The data from the first dataset is input into the second agent, which outputs the harmonic compensation voltage components under the current operating condition, including:

[0100] The grid-side voltage and grid-side current are input into the second intelligent agent, which determines the harmonic compensation voltage component based on the trained first neural network model.

[0101] In this embodiment, the grid-side voltage is a three-phase voltage, and the grid-side current is a three-phase current. Grid-side voltage and grid-side current refer to the voltage and current input to the power grid.

[0102] For example, historical grid-side voltage and current are input into the first neural network model for training. The input historical grid-side voltage and current include labels such as harmonic amplitude, phase, and harmonic compensation voltage components. Training continues until a preset target is reached, at which point the first neural network model is trained. Then, the grid-side voltage and current from the first dataset are input into the trained first neural network model. The trained first neural network model extracts features based on the grid-side voltage and current from the first dataset, obtaining features such as harmonic amplitude and phase. Furthermore, the trained first neural network model outputs harmonic compensation voltage components based on the extracted harmonic amplitude and phase features. The structure of the first neural network model is not limited here; it can be an existing neural network model, as long as it can extract harmonic features based on historical data and output harmonic compensation voltage components based on these features.

[0103] By predicting the harmonic compensation voltage component using the first neural network model, the compensation lag problem caused by response delay in traditional methods can be effectively overcome. The harmonic compensation voltage component is superimposed on the voltage reference value output by the fundamental current loop, which can be used as the modulation wave of the inverter circuit to reduce grid-side harmonics and make the grid-connected current waveform close to a sine wave.

[0104] In one embodiment, the first dataset includes the on-state voltage drop, off-state overvoltage, and case temperature of all power switching devices in the photovoltaic inverter circuit.

[0105] In a photovoltaic inverter circuit, the on-state voltage drop and off-state overvoltage of all power switching devices are operating data, while the casing temperature is environmental data.

[0106] On-state voltage drop refers to the voltage drop between the main electrodes of a power switching device when it is in the on-state and a rated current flows through it. For example, in the case of an IGBT, the on-state voltage drop is the voltage drop between the collector and emitter of the IGBT. A larger on-state voltage drop results in greater losses and more severe heat generation for the same current. Turn-off overvoltage occurs when a power switching device is rapidly turned off from the on-state. Due to the parasitic inductance in the main circuit, the rapid change in current induces a voltage across the inductor, which is superimposed on the bus voltage, causing a voltage spike across the device exceeding the normal turn-off value. Case temperature refers to the temperature of the package housing of the power switching device.

[0107] The data from the first dataset is input into the third agent, which outputs the health status of the power switching devices in the photovoltaic inverter circuit, including:

[0108] The on-state voltage drop, off-state overvoltage, and casing temperature of all power switching devices in the photovoltaic inverter circuit are input into the third intelligent agent. Based on the trained second neural network model, the third intelligent agent outputs the initial health status of each power switching device. Based on the initial health status of each power switching device, the third intelligent agent determines the health status of the power switching devices in the photovoltaic inverter circuit.

[0109] For example, the on-state voltage drop, off-state overvoltage, and case temperature of each power switching device are normalized to form a time-series matrix. This time-series matrix is ​​then input into a trained second neural network model, which outputs the health status of each power switching device. The process of obtaining the trained second neural network model is similar to that of obtaining the trained first neural network model; it can also be supervised training using historical data, which may carry labels such as remaining service life and failure time. For example, a long short-term memory (LSTM) network model within a recurrent neural network is used to learn the time dependence and degradation trend of the input time-series matrix and extract deep degradation features. These deep degradation features are then input into a fully connected layer, and the health status of the power switching devices is output after passing through a sigmoid function. No specific restrictions are placed on the structure of the second neural network model.

[0110] As a core component of photovoltaic inverter circuits, power switching devices directly affect circuit safety. By using a second neural network model to predict the health of power switching devices, subtle degradation trends can be detected, improving the accuracy of health prediction and providing a foundation for subsequent control of photovoltaic inverter circuits.

[0111] In one embodiment, after obtaining an initial health status, the third agent determines the health status of the power switching devices in the photovoltaic inverter circuit based on the initial health status of each power switching device, including:

[0112] The minimum initial health value among all power switching devices is taken as the health value of the power switching devices in the photovoltaic inverter circuit.

[0113] In the embodiments of this application, both the boost converter circuit and the inverter circuit are equipped with power switching devices. It is understood that a lower health rating indicates a less healthy power switching device.

[0114] For example, if the initial health of the power switching devices in the boost converter circuit is less than that in the inverter circuit, then the initial health of the power switching devices in the boost converter circuit shall be used as the health of the power switching devices in the photovoltaic inverter circuit. If both the boost converter circuit and the inverter circuit contain multiple power switching devices, then the power switching device with the lowest initial health among all power switching devices shall be used as the health of the power switching devices in the photovoltaic inverter circuit.

[0115] Understandably, a severe decline in the health of any power switching device can trigger a chain reaction. Using the minimum value as the health status of the power switching devices in the photovoltaic inverter circuit is equivalent to identifying the weakest point in the photovoltaic inverter circuit. This can prevent a power switching device from operating as expected when it is close to failure, thereby avoiding catastrophic failures in advance.

[0116] In one embodiment, before determining the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit based on at least a portion of the first dataset, reference current, reference voltage, harmonic compensation voltage component, and health status, refer to Figure 4 The method further includes steps S202 and S204, wherein:

[0117] Step S202: Issue a warning signal when the health level is less than the preset threshold.

[0118] The warning signal can be either an audible or visual warning; no restrictions are placed on the warning signal.

[0119] Step S204: If the health level is greater than or equal to the preset threshold, proceed to the step of determining the first PWM duty cycle and the second PWM duty cycle.

[0120] Steps S202 and S204 are parallel steps; one of them is selected to be executed based on the health status condition.

[0121] If the health level is below the preset threshold, it indicates that the power switching devices are severely aged, and further use may cause circuit failure or even thermal runaway, requiring treatment of the power switching devices. If the health level is greater than or equal to the preset threshold, it indicates that the power switching devices can still be used, and the photovoltaic inverter circuit can be adjusted based on the acquired data.

[0122] In one embodiment, the first dataset includes the current power generation, the total harmonic distortion rate of the grid after harmonic compensation voltage component compensation, and the grid-side voltage.

[0123] Based on at least a portion of the first dataset, reference current, reference voltage, harmonic compensation voltage components, and health status, determine the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit, see [reference]. Figure 5 The process includes steps S302 to S310, wherein:

[0124] Step S302: Determine the reference PWM duty cycle of the boost converter circuit based on the reference voltage.

[0125] As mentioned earlier, the reference voltage is the voltage corresponding to the target maximum power point. The reference PWM duty cycle determined based on the reference voltage represents the duty cycle for controlling the output reference voltage of the photovoltaic module.

[0126] For example, based on a reference voltage, a reference PWM duty cycle for the boost converter circuit is generated through dual closed-loop voltage and current control. Specifically, the photovoltaic-side voltage in the first dataset is compared with the reference voltage. A reference current is output through a PI regulator. This reference current is then compared with the photovoltaic-side current in the first dataset. A modulation signal is output through the PI regulator, and this modulation signal is input to the PWM generator to obtain the reference PWM duty cycle. The reference PWM duty cycle is the duty cycle set to achieve the target maximum power point.

[0127] Step S304: Based on the power generation, total harmonic distortion rate (THD), grid-side voltage, and health status, determine the output weights of the first, second, and third intelligent agents. The sum of the output weights of the three agents is 1.

[0128] Power generation is the product of reference current and reference voltage. Total Harmonic Distortion (THD) is a core indicator for measuring power quality. THD quantifies the degree of distortion of the actual voltage or current waveform relative to an ideal sine wave; a lower THD value indicates higher power quality. For example, THD can be the grid voltage THD value, which can be calculated based on the harmonic compensation voltage components after fundamental wave compensation. Specifically, this calculation uses the effective values ​​of the fundamental and harmonic waves, which is existing technology and will not be elaborated upon here.

[0129] Understandably, power generation is a parameter related to the first intelligent agent, total harmonic distortion and grid-side voltage are parameters related to the second intelligent agent, and health is a parameter related to the third intelligent agent.

[0130] By determining the output weights of the first, second, and third intelligent agents based on power generation, total harmonic distortion (THD), grid-side voltage, and health status, the system can comprehensively consider power generation, power quality, and the health status of power switching devices, and output weights that are adapted to the current operating conditions. These weights can adjust the duty cycle of the first and second PWM circuits, making the control of the photovoltaic inverter circuit more suitable for the current operating conditions.

[0131] Step S306: Determine the deduction coefficient based on health status. Health status is positively correlated with the deduction coefficient; the deduction coefficient is greater than or equal to 0 and less than or equal to 1.

[0132] The derating factor is determined based on the health status and is used to adjust the duty cycle of the first PWM and the duty cycle of the second PWM.

[0133] For example, the deduction coefficient is positively correlated with health score; the lower the health score, the smaller the deduction coefficient.

[0134] For example, the deduction coefficient has a linear mapping relationship with health level. When health level is greater than or equal to 0.8, the deduction coefficient is 1. When health level is greater than or equal to 0.3 and less than 0.8, the deduction coefficient ranges from greater than or equal to 0.5 and less than 1. When health level is less than 0.3, the deduction coefficient is less than 0.5.

[0135] For example, the reduction coefficient = maximum reduction + (1 - maximum reduction) × health value raised to the power of B, where B is greater than 1, and the maximum reduction is greater than 0 and less than 1. By setting the reduction coefficient using an exponential function, the early reduction can be gradual, meaning that as the health value decreases, the change in the reduction coefficient becomes greater.

[0136] Step S308: Determine the first PWM duty cycle of the inverter circuit based on the output weight, reference current, derating factor, and harmonic compensation voltage component.

[0137] Understandably, if the first PWM duty cycle of the inverter circuit is determined directly based on the harmonic compensation voltage component, the health status of the power switching devices and the impact on the maximum power point may be ignored. By using the output weight, reference current, derating factor and harmonic compensation voltage component to jointly determine the first PWM duty cycle, multiple objectives such as the health of the power switching devices, power quality and photovoltaic module output power can be considered, and the first PWM duty cycle that is most suitable for the current operating conditions can be determined.

[0138] Step S310: Determine the second PWM duty cycle of the boost converter circuit based on the output weight, reference PWM duty cycle, and derating factor.

[0139] As mentioned earlier, by determining the boost converter circuit based on the output weight, reference PWM duty cycle, and derating factor, multiple objectives can be comprehensively considered to determine the second PWM duty cycle that best suits the current operating conditions.

[0140] In one embodiment, the first PWM duty cycle of the inverter circuit is determined based on the output weight, reference current, derating factor, and harmonic compensation voltage component. (See [reference]). Figure 6 The process includes steps S402 to S404, wherein:

[0141] Step S402: Determine the target current based on the reference current, derating factor, and output weight of the third agent.

[0142] Understandably, the reference current is determined based on the target maximum power point under the current operating conditions. When control is strictly implemented according to the target maximum power point, ignoring power quality and the health of the power switching devices, the target current serves as the reference current. If the health of the power switching devices is insufficient to support the photovoltaic inverter circuit operating at the target maximum power point, the reference current can be adjusted using a derating factor and the output weight of the third-party intelligent agent to reduce it to the target current, thereby reducing the load on the power switching devices.

[0143] As mentioned earlier, when control is performed entirely according to the target maximum power point, the target current serves as the reference current, meaning the output weight of the first agent is 1, and the output weights of the other agents are 0. In other words, the output weight of the first agent can be directly reflected from the relationship between the reference current and the target current. Therefore, the target current determined based on the reference current, the derating factor, and the output weight of the third agent can directly reflect the output weights of both the first and third agents.

[0144] Step S404: Determine the first PWM duty cycle based on the target current and harmonic compensation voltage components.

[0145] For example, the first PWM duty cycle is determined based on the target current and the voltage after harmonic compensation voltage component compensation. This can be done by directly outputting the first PWM duty cycle from a preset formula, or by predicting it through a pre-set model. There are no restrictions here, as long as the first PWM duty cycle can be determined based on the target current and the voltage after harmonic compensation voltage component compensation.

[0146] As mentioned earlier, the sum of the output weights of the three agents is 1. When the target current directly reflects the output weights of the first and third agents, it is equivalent to adjusting the reference current based on the output weights of the three agents to obtain the target current. The target current is the current best suited to the current operating condition. Based on the target current and harmonic compensation voltage components, the first PWM duty cycle is determined, which is the duty cycle of the inverter circuit best suited to the current operating condition.

[0147] In one embodiment, the target current is determined based on the reference current, the derating factor, and the output weights of the third agent, including:

[0148] The target current is calculated based on the following formula:

[0149]

[0150] in, Indicates the reference current. Indicates the target current. This represents the output weights of the third agent. This represents the reduction coefficient.

[0151] For example, if the reference current is 100A, the output weight of the third agent is 0.6, and the derating factor is 0.6, then the target current is 100A × (1 - 0.6 × 0.4) = 76A. If the output weight of the third agent is 0.1 and the derating factor is 0.9, then the target current is 100A × (1 - 0.1 × 0.1) = 99A.

[0152] Comparing the two examples reveals that when the output weight of the third agent is large, the difference between the target current and the reference current is significant. When the output weight of the third agent is small, the difference between the target current and the reference current is small. However, when the output weight of the first agent is large, the target current is closer to the reference current. Therefore, the target current reflects the output weights of the three agents, and it is the current of the inverter circuit that best suits the current operating condition, determined based on the output weights of the three agents.

[0153] In one embodiment, the second PWM duty cycle of the boost converter circuit is determined based on the output weight, the reference PWM duty cycle, and the derating factor, including:

[0154] Based on the output weights of the first agent, the second agent, the third agent, the reference pulse width modulation signal, and the derating factor, the second PWM duty cycle is determined using the following formula:

[0155]

[0156] in, Indicates the reference PWM duty cycle. Indicates the duty cycle of the second PWM. This represents the output weight of the first agent. This represents the output weights of the third agent. This represents the reduction coefficient. This represents the output weight of the second agent.

[0157] For example, with a reference PWM duty cycle of 0.85, the output weight of the first agent is 0.8, the output weight of the second agent is 0.1, the output weight of the third agent is 0.1, and the derating factor is 1, then the second PWM duty cycle is 0.8×0.85 + 0.1×0.85×1 + 0.1×0.85 = 0.85. With a reference PWM duty cycle of 0.8, the output weight of the first agent is 0.6, the output weight of the second agent is 0.2, the output weight of the third agent is 0.2, and the derating factor is 0.8, then the second PWM duty cycle is 0.6×0.8 + 0.2×0.8×0.8 + 0.2×0.8 = 0.77.

[0158] By adjusting the reference PWM duty cycle based on the output weights of the three agents, the actual situation of power quality and the health of power switching devices can be comprehensively considered to determine the second PWM duty cycle of the boost converter circuit that is suitable for the current operating conditions. This reduces the losses in the photovoltaic inverter circuit caused by increasing the second PWM duty cycle due to neglecting power quality and the health of power switching devices, thereby achieving multi-objective collaborative optimization.

[0159] In one embodiment, the output weights of the first, second, and third intelligent agents are determined based on power generation, total harmonic distortion (THD), grid-side voltage, and health status. (See [reference]) Figure 7 The process includes steps S502 to S506, wherein:

[0160] Step S502: Use the current power generation, total harmonic distortion rate, grid-side voltage, and health status as the state inputs for the reinforcement learning algorithm.

[0161] The process of outputting the agent's output weights through reinforcement learning algorithms is actually a Markov decision-making process. It combines the power generation, total harmonic distortion rate, grid-side voltage, and health status corresponding to each operating condition into a state vector to provide a basis for subsequent output weights.

[0162] Step S504: Set the reward function for the reinforcement learning algorithm based on power generation, total harmonic distortion rate connected to the grid, and health status.

[0163] For example, a reward function is set with the goals of maximizing power generation, minimizing total harmonic distortion (THD) of the grid, and maximizing health. When the goals are achieved, a reward is given, and when the goals are below a standard threshold, a penalty is imposed.

[0164] Step S506: Use the output weights of the first agent, the second agent, and the third agent as the output of the reinforcement learning algorithm.

[0165] Based on the state vector and reward function, the output weights of each agent can be obtained.

[0166] When power quality is good and power switching devices are healthy, the output weights of the second and third intelligent agents can be reduced, while the output weight of the first intelligent agent can be increased to improve power generation and approach the target maximum power point. When power switching devices are aging, the output weights of the first and second intelligent agents can be reduced, while the output weight of the third intelligent agent can be increased. This essentially involves appropriately sacrificing power generation and power quality to prioritize the protection of the power switching devices. If harmonics are severe and the power switching devices have a certain degree of aging, harmonics can be suppressed as much as possible while ensuring the power switching devices are not damaged. In this case, the output weight of the first intelligent agent can be sacrificed to the greatest extent possible, while the output weights of the second and third intelligent agents can be increased to protect power quality and the power switching devices.

[0167] In one implementation, operational and environmental data are first collected using sensors or sampling circuits to form the first dataset. The collected data can also be preprocessed before forming the first dataset. Next, data from the first dataset is input into a first, second, and third intelligent agent. Each agent processes a portion of the data in the first dataset and outputs reference current, reference voltage, harmonic compensation voltage components, and the health status of the power switching devices in the photovoltaic inverter circuit. If the health status is below a preset threshold, a warning signal is issued; if the health status is above the preset threshold, subsequent processing steps are initiated. Then, based on the partial data and health status of the first dataset, the output weights of the first, second, and third intelligent agents are determined. Finally, based on the output weights of each agent and the reference current, reference voltage, harmonic compensation voltage components, and the health status of the power switching devices in the photovoltaic inverter circuit, the first PWM duty cycle and the second PWM duty cycle are output.

[0168] The PWM duty cycle is the most crucial actuator variable in a photovoltaic inverter circuit. It controls the flow of energy from the input side to the output side, thereby accurately determining the magnitude, waveform, and transmitted power of the output current and voltage. By determining the PWM duty cycle, the grid-connected power quality, the load on the power switching devices, and the output power can be directly determined.

[0169] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0170] Based on the same inventive concept, this application also provides a control device for a photovoltaic inverter circuit to implement the control method for the photovoltaic inverter circuit described above. The solution provided by this device is similar to the implementation described in the above method. Therefore, the specific limitations in one or more embodiments of the control device for the photovoltaic inverter circuit provided below can be found in the limitations of the control method for the photovoltaic inverter circuit described above, and will not be repeated here.

[0171] In one exemplary embodiment, such as Figure 8 As shown, this application also provides a control device for a photovoltaic inverter circuit. The photovoltaic inverter circuit includes a boost converter circuit and an inverter circuit. The boost converter circuit is electrically connected to the photovoltaic array, and the inverter circuit is electrically connected to the boost converter circuit. The device includes a data acquisition module 202, a first intelligent agent 204, a second intelligent agent 206, a third intelligent agent 208, and a data processing module 210.

[0172] The data acquisition module 202 is used to acquire the operating data and environmental data of the photovoltaic inverter in real time as the first dataset.

[0173] The first intelligent agent 204 is used to output the reference current and reference voltage for the current operating condition based on the data collected in the first dataset. The reference current and reference voltage refer to the current and voltage corresponding to the target maximum power point under the current operating condition, respectively.

[0174] The second intelligent agent 206 is used to output the harmonic compensation voltage component of the current operating condition based on the data collected in the first dataset.

[0175] The third agent 208 is used to output the health status of the power switching devices in the photovoltaic inverter circuit based on the data collected in the first dataset.

[0176] The data processing module 210 is used to determine the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit based on at least a portion of the first dataset, reference current, reference voltage, harmonic compensation voltage component, and health status.

[0177] In one embodiment, see [reference] Figure 9 The device also includes an early warning module 212, which is used to issue an early warning signal when the health level is less than a preset threshold.

[0178] The data acquisition module 202 can be used to implement all the steps of acquiring the first dataset in the above embodiments. The first intelligent agent 204 can be used to implement all the steps of determining the reference current and reference voltage in the above embodiments. The second intelligent agent 206 can be used to implement all the steps of determining the harmonic compensation voltage component in the above embodiments. The third intelligent agent 208 can be used to implement all the steps of determining the health of the power switching devices in the photovoltaic inverter circuit in the above embodiments. The data processing module 210 can be used to implement all the steps of determining the first PWM duty cycle and the second PWM duty cycle in the above embodiments. The details of each step will not be elaborated here.

[0179] Each module in the control device of the aforementioned photovoltaic inverter circuit can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0180] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 10 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores control data for the photovoltaic inverter circuit. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a control method for the photovoltaic inverter circuit.

[0181] Those skilled in the art will understand that Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0182] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the embodiments described above.

[0183] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps of the above embodiments.

[0184] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the embodiments described above.

[0185] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0186] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0187] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A control method for a photovoltaic inverter circuit, characterized in that, The photovoltaic inverter circuit includes a boost converter circuit and an inverter circuit, wherein the boost converter circuit is electrically connected to the photovoltaic array, and the inverter circuit is electrically connected to the boost converter circuit; the method includes: Real-time acquisition of photovoltaic inverter operating data and environmental data is used as the primary dataset; The data in the first dataset is input into the first intelligent agent, which outputs the reference current and reference voltage for the current operating condition; the reference current and reference voltage refer to the current and voltage corresponding to the target maximum power point under the current operating condition, respectively. The data in the first dataset is input into the second intelligent agent, which outputs the harmonic compensation voltage component of the current operating condition. The data in the first dataset is input into the third intelligent agent, and the health status of the power switching devices in the photovoltaic inverter circuit is output. Based on at least a portion of the first dataset, the reference current, the reference voltage, the harmonic compensation voltage component, and the health status, the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit are determined.

2. The method according to claim 1, characterized in that, The first dataset includes the current power generation, the total harmonic distortion rate of the grid after harmonic compensation voltage component compensation, and the grid-side voltage; Determining the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit based on at least a portion of the first dataset, the reference current, the reference voltage, the harmonic compensation voltage component, and the health status includes: The reference PWM duty cycle of the boost converter circuit is determined based on the reference voltage. Based on the power generation, the total harmonic distortion rate of the grid, the grid-side voltage, and the health status, the output weights of the first agent, the second agent, and the third agent are determined; the sum of the output weights of the three agents is 1. The deduction coefficient is determined based on the health score, and the health score is positively correlated with the deduction coefficient; the deduction coefficient is greater than or equal to 0 and less than or equal to 1. The first PWM duty cycle of the inverter circuit is determined based on the output weight, the reference current, the derating factor, and the harmonic compensation voltage component. The second PWM duty cycle of the boost converter circuit is determined based on the output weight, the reference PWM duty cycle, and the derating factor.

3. The method according to claim 2, characterized in that, Determining the first PWM duty cycle of the inverter circuit based on the output weight, the reference current, the derating factor, and the harmonic compensation voltage component includes: The target current is determined based on the reference current, the derating factor, and the output weight of the third agent; The first PWM duty cycle is determined based on the target current and the harmonic compensation voltage component.

4. The method according to claim 3, characterized in that, Determining the target current based on the reference current, the derating factor, and the output weights of the third agent includes: The target current is calculated based on the following formula: in, Indicates the reference current. Indicates the target current. This represents the output weights of the third agent. This represents the reduction coefficient.

5. The method according to claim 2, characterized in that, The step of determining the second PWM duty cycle of the boost converter circuit based on the output weight, the reference PWM duty cycle, and the derating factor includes: Based on the output weights of the first agent, the second agent, the third agent, the reference pulse width modulation signal, and the derating factor, the second PWM duty cycle is determined using the following formula: in, Indicates the reference PWM duty cycle. Indicates the duty cycle of the second PWM. This represents the output weight of the first agent. This represents the output weights of the third agent. This represents the reduction coefficient. This represents the output weight of the second agent.

6. The method according to claim 2, characterized in that, The determination of the output weights of the first agent, the second agent, and the third agent based on the power generation, the total harmonic distortion rate connected to the grid, the grid-side voltage, and the health status includes: The power generation, total harmonic distortion rate, grid-side voltage, and health status under the current operating conditions are used as the state inputs for the reinforcement learning algorithm. The reward function of the reinforcement learning algorithm is set based on the power generation, the total harmonic distortion rate connected to the grid, and the health status. The output weights of the first agent, the second agent, and the third agent are used as the output of the reinforcement learning algorithm.

7. The method according to claim 1, characterized in that, The first dataset includes photovoltaic-side current, photovoltaic-side voltage, ambient temperature, and ambient humidity; The step of inputting data from the first dataset into the first intelligent agent and outputting the reference current and reference voltage for the current operating condition includes: The photovoltaic-side current, photovoltaic-side voltage, ambient temperature, and ambient humidity are input into a first intelligent agent, which determines the reference current and reference voltage based on a fuzzy logic control algorithm; and / or The first dataset includes grid-side voltage and grid-side current; The step of inputting data from the first dataset into the second intelligent agent and outputting the harmonic compensation voltage component of the current operating condition includes: The grid-side voltage and grid-side current are input into the second intelligent agent, which determines the harmonic compensation voltage component based on a trained first neural network model; and / or The first dataset includes the on-state voltage drop, off-state overvoltage, and casing temperature of all power switching devices in the photovoltaic inverter circuit; The step of inputting data from the first dataset into the third intelligent agent and outputting the health status of the power switching devices in the photovoltaic inverter circuit includes: The on-state voltage drop, off-state overvoltage, and casing temperature of all power switching devices in the photovoltaic inverter circuit are input to a third intelligent agent. Based on the trained second neural network model, the third intelligent agent outputs the initial health status of each power switching device. Based on the initial health status of each power switching device, the third intelligent agent determines the health status of the power switching devices in the photovoltaic inverter circuit.

8. The method according to claim 7, characterized in that, Having obtained the initial health status, the third agent determines the health status of the power switching devices in the photovoltaic inverter circuit based on the initial health status of each power switching device, including: The minimum initial health value among all power switching devices is taken as the health value of the power switching devices in the photovoltaic inverter circuit.

9. The method according to claim 1, characterized in that, Before determining the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit based on at least a portion of the first dataset, the reference current, the reference voltage, the harmonic compensation voltage component, and the health status, the method further includes: If the health status is lower than a preset threshold, an early warning signal will be issued; If the health status is greater than or equal to a preset threshold, proceed to the step of determining the first PWM duty cycle and the second PWM duty cycle.

10. A control device for a photovoltaic inverter circuit, characterized in that, The photovoltaic inverter circuit includes a boost converter circuit and an inverter circuit. The boost converter circuit is electrically connected to the photovoltaic array, and the inverter circuit is electrically connected to the boost converter circuit. The device includes: The data acquisition module is used to acquire real-time operating data and environmental data of the photovoltaic inverter as the first dataset; The first intelligent agent is used to output the reference current and reference voltage of the current operating condition based on the data in the first dataset; the reference current and the reference voltage refer to the current and voltage corresponding to the target maximum power point under the current operating condition, respectively. The second intelligent agent is used to output the harmonic compensation voltage component of the current operating condition based on the data in the first dataset. A third intelligent agent is used to output the health status of the power switching devices in the photovoltaic inverter circuit based on the data in the first dataset. The data processing module is used to determine the first PWM duty cycle of the inverter circuit and the second PWM duty cycle of the boost converter circuit based on at least a portion of the first dataset, the reference current, the reference voltage, the harmonic compensation voltage component, and the health status.

Images