Double closed-loop control method and system of hydrogen fuel cell grid-connected controller

Abstract

The invention provides a double-closed-loop control method and system for a hydrogen fuel cell grid-connected controller, and relates to the technical field of grid-connected control, and the method comprises the steps: setting an outer loop reference voltage, parameters of a PI controller and parameters of a quasi-PR controller, collecting a real-time output voltage and a real-time output current of an inverter, and carrying out the real-time control of the real-time output voltage and the real-time output current of the inverter; comparing the outer loop reference voltage with the real-time output voltage to obtain a voltage amplitude deviation, generating a control quantity by adopting a PI controller according to the voltage amplitude deviation, detecting the phase and the frequency of the power grid voltage in real time by a phase-locked loop, and generating a real-time inner loop reference current according to the control quantity, the phase and the frequency; the difference value between the real-time inner ring reference current and the real-time output current is calculated, a quasi-PR controller is adopted to generate a PWM modulation signal according to the difference value, the inverter is regulated and controlled according to the PWM modulation signal, and the change of a power grid can be tracked more accurately.

Description

Technical Field

[0001] The present application relates to the technical field of grid-connected control, and in particular to a dual-closed-loop control method and system for a hydrogen fuel cell grid-connected controller. Background Art

[0002] With the growing global demand for clean energy and growing concerns about the environmental pollution and energy depletion associated with traditional fossil fuels, hydrogen fuel cells, as a highly efficient and clean energy conversion device, have garnered widespread attention. Hydrogen fuel cells directly convert chemical energy into electrical energy through an electrochemical reaction between hydrogen and oxygen, with the sole product being water. These fuel cells offer advantages such as zero emissions, high energy conversion efficiency, and quiet operation, demonstrating significant potential for application in distributed power generation, transportation, and portable power supplies. In the distributed power generation sector, hydrogen fuel cells can be connected to the power grid, providing stable power support. This helps improve energy efficiency, reduce reliance on traditional grids, and promote energy mix optimization and sustainable development. When the DC power output of a hydrogen fuel cell needs to be connected to the grid, it must be converted to AC power via an inverter. This inverter must ensure that the AC output matches the grid's voltage, frequency, and phase parameters to ensure safe, stable, and efficient grid-connected operation. The quality of the grid-connected control technology directly impacts the performance, power quality, and grid stability of the hydrogen fuel cell power generation system.

[0003] Traditional single-loop control methods typically only control the inverter's output voltage or current, failing to simultaneously address the dynamic performance and steady-state accuracy of both voltage and current. While some existing dual-loop control methods have improved grid-connected control performance to a certain extent, they are inflexible when dealing with phase and frequency variations in grid voltage and cannot accurately track grid changes in real time. This can lead to desynchronization between grid-connected current and grid voltage, resulting in significant harmonics and reactive power, impacting power quality and grid stability. Summary of the Invention

[0004] In order to more accurately track changes in the power grid, the present application provides a dual closed-loop control method and system for a hydrogen fuel cell grid-connected controller.

[0005] In the first aspect, the present application provides a dual closed-loop control method for a hydrogen fuel cell grid-connected controller, which adopts the following technical solutions: A dual closed-loop control method for a hydrogen fuel cell grid-connected controller, wherein the direct current output of the hydrogen fuel cell is converted into alternating current via an inverter, comprises the following steps: Setting the outer loop reference voltage, parameters of the PI controller, and parameters of the quasi-PR controller, collecting the real-time output voltage and real-time output current of the inverter, comparing the outer loop reference voltage with the real-time output voltage to obtain a voltage amplitude deviation, using the PI controller to generate a control variable based on the voltage amplitude deviation, using a phase-locked loop to detect the phase and frequency of the grid voltage in real time, and generating a real-time inner loop reference current based on the control variable, phase, and frequency; The difference between the real-time inner loop reference current and the real-time output current is calculated, a quasi-PR controller is used to generate a PWM modulation signal according to the difference, and the inverter is regulated according to the PWM modulation signal.

[0006] The PI controller generates a control quantity based on the amplitude deviation between the outer loop reference voltage and the real-time output voltage of the inverter. The phase-locked loop monitors the phase and frequency of the grid voltage in real time, and quickly generates a real-time inner loop reference current based on the control quantity, phase and frequency information. The quasi-PR controller generates a PWM modulation signal based on the difference between the reference current and the actual output current, and uses the PWM modulation signal to regulate the inverter so that the output current quickly adapts to changes in the grid. The above scheme can quickly adjust the output of the inverter in the event of grid voltage fluctuations, frequency offsets, etc., reducing grid connection failures or power quality problems caused by grid changes. The present application uses a quasi-PR controller to achieve precise current tracking so that the current output by the inverter is in phase with the grid voltage, thereby improving the stability of the power factor. A stable power factor means that the reactive power loss in the grid is reduced, the active power transmission efficiency is improved, and the power quality is further improved.

[0007] Optionally, the method further includes: Acquire training data and training labels, wherein the training data includes the historical output current of the inverter, the historical inner loop reference current, and the difference between the historical output current and the historical inner loop reference current, and the training labels include parameters Kp, Kr and ; Construct a neural network model, train the neural network model using training data and training labels, and obtain a trained neural network model; The real-time inner loop reference current, the real-time output current, and the difference between the real-time inner loop reference current and the real-time output current are input into the trained neural network model to obtain the predicted parameters Kp, Kr and , using the predicted parameters Kp, Kr and The parameters of the quasi-PR controller are updated, and the quasi-PR controller with updated parameters is recorded as a new quasi-PR controller.

[0008] This application obtains the historical output current of the inverter, the historical inner loop reference current and their difference as training data, and collects the parameters Kp, Kr and As training labels, these actual operating data are used to build a neural network model and train it. The neural network can learn complex nonlinear relationships from historical data. Compared with the traditional parameter setting method based on experience and theoretical deduction, this solution can more accurately predict the parameters Kp, Kr and ,Since these parameters are obtained based on real-time data and the learning ability of the neural network, the present application can more accurately match the operating state of the current system, thereby improving the accuracy of the quasi-PR controller in current tracking, making the output current of the inverter more accurately follow the inner loop reference current, and improving the control accuracy of the entire grid-connected control system.

[0009] The traditional fixed parameter quasi-PR controller may have parameter mismatch problems under different working conditions, resulting in decreased system stability. However, the neural network can predict and update the parameters Kp, Kr and , which can keep the parameters of the quasi-PR controller at an optimal state, minimize system oscillations and overshoot caused by inappropriate parameters, and thus optimize system stability. Accurate parameter updates help the quasi-PR controller more effectively suppress harmonics and interference in the inverter output current.

[0010] Optionally, the method further includes: Constructing an electrochemical model of a hydrogen fuel cell, collecting historical and real-time environmental data, including temperature, humidity, and hydrogen pressure data, calculating the statistical distribution of the historical environmental data, and using a Monte Carlo simulation algorithm to generate multiple sets of environmental scenarios based on the statistical distribution; Calculate the output voltage deviation under each set of environmental scenarios using an electrochemical model, establish a mapping model between the output voltage deviation and environmental data, solve the coefficients of the mapping model using the least squares method, and obtain the voltage correction value under the real-time environmental data based on the mapping model and the coefficients of the mapping model; The sum of the voltage amplitude deviation and the voltage correction value is taken as the new voltage amplitude deviation.

[0011] The output voltage of hydrogen fuel cells will be significantly affected by environmental factors (temperature, humidity, pressure). By constructing an electrochemical model and collecting historical and real-time environmental data, this application can more comprehensively consider the effects of these environmental factors on the output voltage. Traditional methods are based only on electrical parameters for control, ignoring voltage fluctuations caused by environmental factors. This method takes environmental factors into consideration to make control more precise. Using the above scheme, this application can more accurately reflect the impact of the actual environment on the output voltage, thereby improving the accuracy of voltage control. For example, in a high temperature environment, the output voltage of a hydrogen fuel cell will drop. This method can calculate the corresponding voltage correction value, compensate for the voltage amplitude deviation, and make the output voltage closer to the set value.

[0012] Environmental conditions vary significantly across regions and seasons, and hydrogen fuel cell grid-connected systems need to be adaptable to these changes. This method, by establishing a mapping model between environmental scenarios and output voltage deviations, can quickly calculate voltage correction values ​​based on real-time environmental data, enabling the system to promptly adjust control strategies to varying environmental conditions. As environmental conditions change in real time, the voltage correction values ​​are dynamically updated. Using this corrected voltage amplitude deviation for control is equivalent to dynamically adjusting control parameters, enabling the system to respond to environmental changes in real time and improving its environmental adaptability and flexibility.

[0013] Optionally, the method further includes: Acquire environmental scenarios where the output voltage deviation is greater than a preset deviation threshold, and delete environmental scenarios where the output voltage deviation is not greater than the preset deviation threshold.

[0014] By adopting the above technical solution, the present application realizes the screening of environmental scenes and improves computing efficiency.

[0015] Optionally, the method further includes: respectively determining whether the real-time temperature data and / or the real-time humidity data are greater than a preset threshold, and if so, executing the step of adjusting the bandwidth; if not, performing no processing; Adjust bandwidth: Construct a bandwidth adjustment model for the phase-locked loop. The calculation model of the bandwidth adjustment model is as follows: ; in, is the bandwidth after the phase-locked loop is adjusted; is the standard bandwidth of the phase-locked loop, which is 50; is the temperature sensitivity coefficient; is the humidity sensitivity coefficient; T is the real-time temperature data; is the preset temperature threshold; H is the real-time humidity data; is the preset humidity threshold.

[0016] Real-time temperature and humidity data are important factors affecting the stability of hydrogen fuel cell grid-connected systems. When real-time temperature or humidity data exceeds a preset threshold, this application constructs a phase-locked loop (PLL) bandwidth adjustment model to dynamically adjust the bandwidth based on real-time environmental data. This allows the PLL to better adapt to environmental changes, minimizing potential loss of PLL lock or performance degradation due to environmental factors, thereby improving overall system stability.

[0017] Under standard conditions, a phase-locked loop (PLL) operates normally with a standard bandwidth. However, when the ambient temperature or humidity exceeds a certain range, the standard bandwidth may no longer be applicable. By adjusting the bandwidth, the PLL parameters can be aligned with the environmental conditions, preventing system oscillation or instability caused by parameter mismatch and ensuring stable operation of hydrogen fuel cells under various environmental conditions. The primary function of the PLL is to accurately detect the phase and frequency of the grid voltage and provide synchronization signals for the grid-connected inverter. By adjusting the bandwidth according to environmental conditions, the PLL can more accurately track changes in the grid's phase and frequency, ensuring that the current output by the grid-connected inverter is in phase and frequency with the grid voltage, thereby improving the quality of grid-connected power.

[0018] Optionally, the method further includes: Determine whether the real-time hydrogen pressure is less than a preset pressure threshold. If so, perform the step of adjusting the parameters; if not, do nothing. Adjusting parameters: reducing the current of a charge pump in a phase-locked loop, reducing the bandwidth of the phase-locked loop to a preset bandwidth threshold, and updating a frequency division ratio of the phase-locked loop based on the preset bandwidth threshold and a standard bandwidth of the phase-locked loop.

[0019] When the real-time hydrogen pressure falls below the preset pressure threshold, the output power and voltage of the hydrogen fuel cell decrease, affecting the input voltage and current of the grid-connected inverter. The phase-locked loop (PLL), a core component of grid-connected control, is responsible for accurately tracking the phase and frequency of the grid voltage. If its parameters are not adjusted promptly, the PLL may lose lock, preventing the grid-connected inverter from properly synchronizing with the grid, leading to system instability or even shutdown. Reducing the charge pump current and bandwidth, and updating the frequency division ratio, allows the PLL to better adapt to changes in input characteristics caused by hydrogen pressure fluctuations, effectively preventing the risk of loss of lock and ensuring stable system operation.

[0020] Optionally, after performing the step of adjusting the parameters, the method further includes: Adjust the proportional gain of the quasi-PR controller based on the real-time hydrogen pressure and the preset pressure threshold , the calculation model is as follows: ; in, is the updated proportional gain of the quasi-PR controller; is the pressure coefficient, which is 0.2; is the real-time hydrogen pressure; is the preset pressure threshold.

[0021] When the real-time hydrogen pressure is less than the preset pressure threshold, the electrochemical reaction in the hydrogen fuel cell is incomplete, and the battery's internal resistance increases. According to Ohm's law, when the battery's internal resistance increases, the output voltage drops significantly, resulting in voltage oscillation or instability. By adjusting the proportional gain of the quasi-PR controller based on the real-time hydrogen pressure and the preset pressure threshold, this application enables the grid-connected controller to better adapt to such changes and enhance the system's stability under abnormal hydrogen pressure conditions.

[0022] Optionally, before performing the step of adjusting the bandwidth, the method further includes: adding a nonlinear term to the mapping model to obtain a new mapping model.

[0023] Optionally, after performing the step of adjusting the bandwidth, the method further includes: Adjusting the resonant gain of the quasi-PR controller based on the bandwidth adjusted by the phase-locked loop , the calculation model is as follows: ; in, is the updated resonant gain of the quasi-PR controller.

[0024] In a second aspect, the present application provides a dual closed-loop control system for a hydrogen fuel cell grid-connected controller, which adopts the following technical solutions: A dual closed-loop control system for a hydrogen fuel cell grid-connected controller, comprising: a memory and a processor, The memory stores a computer-readable storage medium; When the processor processes the computer program stored on the computer-readable storage medium, the method described in the first aspect is implemented.

[0025] In summary, this application includes at least one of the following beneficial technical effects: 1. The PI controller generates a control variable based on the amplitude deviation between the outer loop reference voltage and the inverter's real-time output voltage. The phase-locked loop (PLL) monitors the grid voltage's phase and frequency in real time and rapidly generates a real-time inner loop reference current based on the control variable, phase, and frequency information. The quasi-PR controller generates a PWM modulation signal based on the difference between the reference current and the actual output current. This PWM modulation signal is then used to regulate the inverter, allowing the output current to quickly adapt to grid changes. This solution can rapidly adjust the inverter's output in the event of grid voltage fluctuations or frequency offsets, mitigating grid connection failures and power quality issues caused by grid variations.

[0026] 2. This application uses a quasi-PR controller to achieve precise current tracking so that the current output by the inverter is in phase with the grid voltage, thereby improving the stability of the power factor. A stable power factor means that the reactive power loss in the grid is reduced, the active power transmission efficiency is improved, and the power quality is further improved.

[0027] 3. This application obtains the historical output current of the inverter, the historical inner loop reference current and their difference as training data, and collects the parameters Kp, Kr and As training labels, these actual operating data are used to build a neural network model and train it. The neural network can learn complex nonlinear relationships from historical data. Compared with the traditional parameter setting method based on experience and theoretical deduction, this solution can more accurately predict the parameters Kp, Kr and ,Since these parameters are obtained based on real-time data and the learning ability of the neural network, the present application can more accurately match the operating state of the current system, thereby improving the accuracy of the quasi-PR controller in current tracking, making the output current of the inverter more accurately follow the inner loop reference current, and further improving the control accuracy of the entire grid-connected control system. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 is a flow chart of Example 1 of the present application; Figure 2 This is a flow chart of the process from S3 generating an environmental scenario to S5 updating a voltage amplitude deviation in Example 3 of the present application; Figure 3 This is a flowchart of adjusting bandwidth from S6 to proportional gain from S9 in Example 3 of the present application. DETAILED DESCRIPTION

[0029] The following combination Figures 1 to 3 This application is described in further detail.

[0030] The hydrogen fuel cell grid-connected controller combines inner and outer control loops to precisely regulate the hydrogen fuel cell's output power and ensure stable grid connection. The grid-connected inverter converts the hydrogen fuel cell's DC power into AC power with the same frequency and phase as the grid. The controller adjusts the inverter's switching frequency, duty cycle, and other parameters to control the output current and phase, thereby achieving precise regulation of output power.

[0031] Example 1: This example discloses a dual closed-loop control method for a hydrogen fuel cell grid-connected controller, referring to Figure 1The method includes: S1 calculating the real-time inner loop reference current and S2 regulation, first setting the parameters of the outer loop reference voltage, PI controller and quasi-PR controller, collecting the real-time output voltage and current of the inverter, comparing the outer loop reference voltage with the real-time output voltage to obtain the voltage amplitude deviation, and the PI controller generates the control quantity based on this; the phase-locked loop detects the grid voltage phase and frequency in real time, and generates the real-time inner loop reference current by combining the control quantity, phase and frequency; then calculates the difference between the real-time inner loop reference current and the real-time output current, and the quasi-PR controller generates a PWM modulation signal based on the difference to regulate the inverter. This embodiment includes the following steps: S1 calculates the real-time inner loop reference current, sets the outer loop reference voltage, the parameters of the PI controller, and the parameters of the quasi-PR controller.

[0032] The outer loop reference voltage is the target voltage for the entire grid-connected control system. It is set based on the grid's standard voltage level and grid-connected requirements. For example, in a single-phase 220V AC grid-connected system, the RMS value of the outer loop reference voltage is set to 220V. Setting the outer loop reference voltage ensures that the inverter output voltage matches the grid voltage, achieving stable grid-connected operation.

[0033] A PI controller consists of a proportional and an integral component. The proportional component responds quickly to voltage deviations, producing a control effect proportional to the deviation. Its parameter, Kp, determines the speed and magnitude of the response. The integral component eliminates static errors by adjusting the control variable based on the cumulative deviation over time. Its parameter, Ki, influences the strength of the integral effect. Increasing Kp speeds up the response, while increasing Ki allows for faster elimination of static errors.

[0034] The quasi-PR controller is an improvement on the traditional PR controller. It can provide a higher gain near a specific frequency, thereby more accurately tracking the sinusoidal signal. The parameters of the quasi-PR controller include proportional gain Kp, resonant gain Kr and cutoff frequency .

[0035] The real-time output voltage and real-time output current of the inverter are collected through various sensors, the set outer loop reference voltage is compared with the collected real-time output voltage of the inverter, the difference between the effective values ​​of the two is calculated, and the voltage amplitude deviation is obtained.

[0036] A PI controller is used to generate a control variable based on the voltage amplitude deviation. The proportional link of the PI controller will immediately respond to the deviation and produce a control effect proportional to the deviation; the integral link will accumulate the deviation and gradually eliminate the static error over time.

[0037] The phase-locked loop detects the phase and frequency of the grid voltage in real time. By processing and analyzing the grid voltage signal, the phase and frequency information of the grid voltage signal can be extracted.

[0038] According to the control quantity generated by the PI controller and the grid voltage phase and frequency detected by the phase-locked loop, a real-time inner loop reference current is generated. The calculation model is as follows: ; ; ; in, is the amplitude of the reference current; is the angular frequency of the grid voltage, which is detected by the phase-locked loop; is the phase of the grid voltage, which is detected by the phase-locked loop; is a coefficient, ranging from 0.1 to 10; is the initial current value; is the power of the grid-connected controller when it is no-load; U is the reference voltage of the outer loop.

[0039] S2 control calculates the difference between the real-time inner loop reference current and the real-time output current. The quasi-PR controller adjusts the output signal amplitude and phase based on this current difference, generating a corresponding PWM modulation signal. The duty cycle of the PWM modulation signal controls the on and off time of the inverter's switches, regulating the inverter based on the PWM modulation signal.

[0040] By adopting the above solution, the present application achieves accurate tracking of the inverter output current to the reference value, thereby ensuring that the system has both fast response and steady-state accuracy under dynamic conditions.

[0041] Example 2: This example differs from Example 1 in that the method further comprises: Acquire training data and training labels, wherein the training data includes the historical output current of the inverter, the historical inner loop reference current, and the difference between the historical output current and the historical inner loop reference current, and the training labels include the parameters Kp, Kr and Kp of the quasi-PR controller with the same timestamp as the training data. .

[0042] A neural network model is constructed, wherein the neural network model includes an input layer, a hidden layer, and an output layer.

[0043] The input of the input layer is: the historical output current of the inverter at the same timestamp, the historical inner loop reference current, and the difference between the historical output current and the historical inner loop reference current.

[0044] The hidden layer uses a 2- to 3-layer LSTM network, and each LSTM network layer contains 128 to 256 units to capture the dynamic characteristics of time series; each LSTM layer is followed by a fully connected layer (ReLU activation) to handle nonlinear relationships.

[0045] The output layer is set up with 3 independent neurons to predict Kp, Kr and .

[0046] The neural network model is trained using training data and training labels to obtain a trained neural network model.

[0047] The real-time inner loop reference current, the real-time output current, and the difference between the real-time inner loop reference current and the real-time output current are input into the trained neural network model to obtain the predicted parameters Kp, Kr and , the predicted parameters Kp, Kr and Write the quasi-PR controller, update the parameters of the quasi-PR controller set in S1 to calculate the real-time inner loop reference current, and record the quasi-PR controller with updated parameters as a new quasi-PR controller.

[0048] Through the above scheme, this embodiment can deeply integrate data-driven intelligent algorithms with traditional control theory, providing an efficient and robust quasi-PR controller parameter adaptation solution for hydrogen fuel cell grid-connected systems.

[0049] Example 3: Reference Figure 2 The difference between this embodiment and embodiment 1 is that the method further includes: S3 generates an environmental scenario and constructs an electrochemical model of a hydrogen fuel cell. The establishment of the electrochemical model is very mature and will not be described in detail in this embodiment.

[0050] Collect historical and real-time environmental data, including: Temperature data: fuel cell stack inlet / outlet temperature (range: -10°C~60°C) Humidity data: cathode / anode humidification humidity (RH: 20%~95%) Hydrogen pressure data: anode inlet pressure (0.5bar~3bar) The statistical distribution of historical environmental data was calculated, that is, the mean and standard deviation of each environmental data were calculated. Based on the mean and standard deviation of each environmental data, 10,000 groups of random samples were independently generated for each environmental parameter (temperature, humidity, hydrogen pressure) using the Monte Carlo simulation algorithm, and their statistical distribution was followed.

[0051] Acquire environmental scenarios where the output voltage deviation is greater than a preset deviation threshold, and delete environmental scenarios where the output voltage deviation is not greater than the preset deviation threshold.

[0052] S4 calculates the voltage correction value, inputs each set of scenario parameters into the electrochemical model, calculates the output voltage under each set of scenarios, and calculates the output voltage deviation under each remaining set of environmental scenarios through the electrochemical model and the real-time output voltage.

[0053] A mapping model of output voltage deviation and environmental data is established, and a nonlinear term is added to the mapping model to obtain a new mapping model. The calculation model of the new mapping model is as follows: ; ; ; in, is the nonlinear compensation coefficient; m is a constant ranging from 0 to 1; T is the temperature; is the preset temperature threshold; H is the humidity; is the preset humidity threshold; P is the hydrogen pressure; is the preset pressure threshold.

[0054] is a mapping model between output voltage deviation and environmental data. In this model, .

[0055] The coefficients of the new mapping model are solved by the least square method, and the voltage correction value under the real-time environmental data is obtained according to the new mapping model and the coefficients of the mapping model.

[0056] S5 updates the voltage amplitude deviation, and takes the sum of the voltage amplitude deviation and the voltage correction value as the new voltage amplitude deviation.

[0057] Reference Figure 3 In other embodiments, the method further comprises: It is determined whether the real-time temperature data and / or the real-time humidity data are greater than a preset threshold value. If so, step S6 is executed to adjust the bandwidth; if not, no processing is performed.

[0058] S6 adjusts the bandwidth and constructs a bandwidth adjustment model of the phase-locked loop. The calculation model of the bandwidth adjustment model is as follows: ; in, is the bandwidth after the phase-locked loop is adjusted; is the standard bandwidth of the phase-locked loop, which is 50; is the temperature sensitivity coefficient, which is 0.01; is the humidity sensitivity coefficient, which is 0.02; T is the real-time temperature data; is the preset temperature threshold; H is the real-time humidity data; is the preset humidity threshold.

[0059] S7 adjusts the resonant gain and updates the resonant gain of the quasi-PR controller based on the bandwidth adjusted by the phase-locked loop , the calculation model is as follows: ; in, is the resonant gain of the updated quasi-PR controller.

[0060] Determine whether the real-time hydrogen pressure is less than the preset pressure threshold. If so, execute S8 to adjust the parameters; if not, do nothing. S8 adjusts parameters, reduces the current of the charge pump in the phase-locked loop, reduces the bandwidth of the phase-locked loop to a preset bandwidth threshold, and updates the frequency division ratio of the phase-locked loop based on the preset bandwidth threshold and the standard bandwidth of the phase-locked loop.

[0061] ; in, is the frequency division ratio after the phase-locked loop is updated; N is the frequency division ratio before the phase-locked loop is updated; is the standard bandwidth of the phase-locked loop, which is 50; is the preset bandwidth threshold, which is 20 in this embodiment; in other embodiments, the preset bandwidth threshold can be set according to demand, and the value range is 20~30.

[0062] S9 adjusts the proportional gain, which is based on the real-time hydrogen pressure and the preset pressure threshold. , the calculation model is as follows: ; in, is the proportional gain of the updated quasi-PR controller; is the pressure coefficient, which is 0.2; is the real-time hydrogen pressure; is the preset pressure threshold.

[0063] Example 4: This example discloses a dual closed-loop control system for a hydrogen fuel cell grid-connected controller, the system comprising: a memory and a processor, The memory stores a computer-readable storage medium; When the processor processes the computer program stored on the computer-readable storage medium, it implements the dual closed-loop control method of the hydrogen fuel cell grid-connected controller.

[0064] The above are all preferred embodiments of the present application, and are not intended to limit the scope of protection of the present application. Therefore, any equivalent changes made based on the structure, shape, and principle of the present application should be included in the scope of protection of the present application.

Claims

1. A dual closed-loop control method for a hydrogen fuel cell grid-connected controller, wherein the DC power output by the hydrogen fuel cell is converted into AC power by an inverter, characterized in that: include: Setting the outer loop reference voltage, parameters of the PI controller, and parameters of the quasi-PR controller, collecting the real-time output voltage and real-time output current of the inverter, comparing the outer loop reference voltage with the real-time output voltage to obtain a voltage amplitude deviation, using the PI controller to generate a control variable based on the voltage amplitude deviation, using a phase-locked loop to detect the phase and frequency of the grid voltage in real time, and generating a real-time inner loop reference current based on the control variable, phase, and frequency; The difference between the real-time inner loop reference current and the real-time output current is calculated, a quasi-PR controller is used to generate a PWM modulation signal according to the difference, and the inverter is regulated according to the PWM modulation signal.

2. The dual closed-loop control method of the hydrogen fuel cell grid-connected controller according to claim 1, characterized in that: The method further comprises: Acquire training data and training labels, wherein the training data includes the historical output current of the inverter, the historical inner loop reference current, and the difference between the historical output current and the historical inner loop reference current, and the training labels include parameters Kp, Kr and ; Construct a neural network model, train the neural network model using training data and training labels, and obtain a trained neural network model; The real-time inner loop reference current, the real-time output current, and the difference between the real-time inner loop reference current and the real-time output current are input into the trained neural network model to obtain the predicted parameters Kp, Kr and , using the predicted parameters Kp, Kr and The parameters of the quasi-PR controller are updated, and the quasi-PR controller with updated parameters is recorded as a new quasi-PR controller.

3. The dual closed-loop control method of the hydrogen fuel cell grid-connected controller according to claim 1 or 2, characterized in that: The method further comprises: Constructing an electrochemical model of a hydrogen fuel cell, collecting historical and real-time environmental data, including temperature, humidity, and hydrogen pressure data, calculating the statistical distribution of the historical environmental data, and using a Monte Carlo simulation algorithm to generate multiple sets of environmental scenarios based on the statistical distribution; Calculate the output voltage deviation under each set of environmental scenarios using an electrochemical model, establish a mapping model between the output voltage deviation and environmental data, solve the coefficients of the mapping model using the least squares method, and obtain the voltage correction value under the real-time environmental data based on the mapping model and the coefficients of the mapping model; The sum of the voltage amplitude deviation and the voltage correction value is taken as the new voltage amplitude deviation.

4. The dual closed-loop control method of the hydrogen fuel cell grid-connected controller according to claim 3, characterized in that: The method further comprises: Acquire environmental scenarios where the output voltage deviation is greater than a preset deviation threshold, and delete environmental scenarios where the output voltage deviation is not greater than the preset deviation threshold.

5. The dual closed-loop control method of the hydrogen fuel cell grid-connected controller according to claim 3, characterized in that: The method further comprises: respectively determining whether the real-time temperature data and / or the real-time humidity data are greater than a preset threshold, and if so, executing the step of adjusting the bandwidth; if not, performing no processing; Adjust bandwidth: Construct a bandwidth adjustment model for the phase-locked loop. The calculation model of the bandwidth adjustment model is as follows: ； in, is the bandwidth after the phase-locked loop is adjusted; is the standard bandwidth of the phase-locked loop, which is 50; is the temperature sensitivity coefficient; is the humidity sensitivity coefficient; T is the real-time temperature data; is the preset temperature threshold; H is the real-time humidity data; is the preset humidity threshold.

6. The dual closed-loop control method of the hydrogen fuel cell grid-connected controller according to claim 3, characterized in that: The method further comprises: Determine whether the real-time hydrogen pressure is less than a preset pressure threshold. If so, perform the step of adjusting the parameters; if not, do nothing. Adjusting parameters: reducing the current of a charge pump in a phase-locked loop, reducing the bandwidth of the phase-locked loop to a preset bandwidth threshold, and updating a frequency division ratio of the phase-locked loop based on the preset bandwidth threshold and a standard bandwidth of the phase-locked loop.

7. The dual closed-loop control method of the hydrogen fuel cell grid-connected controller according to claim 6, characterized in that: After performing the step of adjusting the parameters, the method further includes: The proportional gain of the quasi-PR controller is adjusted based on the real-time hydrogen pressure and the preset pressure threshold. The calculation model is as follows: ； in, is the updated proportional gain of the quasi-PR controller; is the pressure coefficient, which is 0.2; is the real-time hydrogen pressure; is the preset pressure threshold.

8. The dual closed-loop control method of the hydrogen fuel cell grid-connected controller according to claim 5, characterized in that: Before performing the step of adjusting the bandwidth, the method further includes: adding a nonlinear term to the mapping model to obtain a new mapping model.

9. The dual closed-loop control method of the hydrogen fuel cell grid-connected controller according to claim 5, characterized in that: After performing the step of adjusting the bandwidth, the method further includes: Adjusting the resonant gain of the quasi-PR controller based on the bandwidth adjusted by the phase-locked loop , the calculation model is as follows: ； in, is the updated resonant gain of the quasi-PR controller.

10. A dual closed-loop control system for a hydrogen fuel cell grid-connected controller, characterized in that: include: memory and processor, The memory stores a computer-readable storage medium; When the processor processes the computer program stored on the computer-readable storage medium, the method according to any one of claims 1 to 9 is implemented.

Images