A method for constructing a microgrid based on an optimized linear active disturbance rejection controller

Abstract

The application discloses a micro-grid construction method based on an optimized linear active disturbance rejection controller. The active disturbance rejection controller converts input signals into state spaces and compares the actual output signals to form tracking error signals. A fuzzy logic parameter dynamic adjustment module amplifies the tracking error signals through proportional gain to form a first control signal. A first linear extended state observer adjusts the first control signal speed, position-based state and low-frequency disturbance output second control signal in real time according to the tracking error signal feedback item in the first control signal. A second linear extended state observer compensates for high-frequency disturbance or uncertain state in real time according to the tracking error signal feedback item in the first control signal to output a third control signal. A controlled adjustment module adjusts a control object according to the output second control signal and third control signal to drive the output signal of the control object to be consistent with the tracking signal. The application enhances the anti-interference ability of the micro-grid and significantly improves the robustness.

Description

Technical Field

[0001] This invention belongs to the field of microgrid control technology, specifically relating to a method for constructing a microgrid based on an optimized linear active disturbance rejection controller, which is particularly suitable for frequency stability control of AC microgrids containing distributed renewable energy. Background Technology

[0002] Microgrids are an effective means of realizing the local consumption and efficient grid connection of distributed renewable energy. As the core control unit of microgrids, grid-connected inverters, based on the control strategy of virtual synchronous generation control model, can endow microgrids with the inertia and damping characteristics of synchronous generators, improve the voltage and frequency support capabilities during grid faults, and become a key solution for large-scale renewable energy grid connection.

[0003] In the secondary frequency regulation control of the virtual synchronous power generation control model, traditional PI control bus and conventional linear active disturbance rejection controller are widely used. However, such control strategies have obvious technical defects: First, the overshoot suppression capability is insufficient, and the system frequency is prone to fluctuation when the load power fluctuates significantly or the frequency regulation command changes. Second, the response time is too long, and it is impossible to achieve rapid frequency stabilization. Third, the anti-interference capability is weak, and it is easily affected by disturbances. Moreover, simply increasing the observer bandwidth cannot effectively resist noise interference, but will reduce control stability instead.

[0004] Meanwhile, the performance of the virtual synchronous generation control model is significantly affected by key parameters such as inertia, damping, and voltage droop coefficient. The secondary frequency regulation scheme of the conventional linear active disturbance rejection controller combined with the virtual synchronous generation control model can only estimate a single disturbance, making it difficult to effectively compensate for cumulative disturbances. Moreover, the controller parameters rely heavily on engineering debugging experience and lack dynamic optimization methods. When the system undergoes nonlinear changes or the inherent parameters change, the original control parameters and inference rules are prone to failure, making it impossible to guarantee the stable control of the microgrid frequency and restricting the operational reliability of the grid-type microgrid under complex operating conditions. Summary of the Invention

[0005] This invention addresses the technical problems existing in the secondary frequency regulation control strategy of the virtual synchronous generation control model, such as poor overshoot suppression, slow response speed, weak anti-interference capability, and inability to dynamically optimize parameters. It provides a method for constructing a microgrid based on an optimized linear active disturbance rejection controller, which enables the microgrid to quickly stabilize its frequency under dynamic operating conditions such as sudden load increases and decreases, improves the system's anti-disturbance performance and tracking accuracy, and enhances the operational stability and robustness of the microgrid under complex operating conditions through dynamic adaptive adjustment of the active disturbance rejection controller parameters.

[0006] The present invention is implemented using the following technical solution: A method for constructing a microgrid based on an optimized linear active disturbance rejection controller (ADRC), wherein the microgrid includes a virtual synchronous generation control model and an ADRC, and the ADRC comprises a first linear extended state observer, a second linear extended state observer, a fuzzy logic parameter dynamic adjustment module, and a controlled adjustment module; including: The active disturbance rejection controller converts the input signal into a state space and compares it with the actual output signal to form a tracking error signal. The fuzzy logic parameter dynamic adjustment module amplifies the tracking error signal through proportional gain to form a first control signal; The first linear extended state observer adjusts the basic state of the speed and position of the first control signal and outputs a second control signal in real time based on the tracking error signal feedback term in the first control signal and low-frequency disturbances. The second linear extended state observer compensates for high-frequency disturbances or uncertainties in real time based on the tracking error signal feedback term in the first control signal and outputs a third control signal. The controlled adjustment module adjusts the controlled object to drive its output signal to be consistent with the tracking signal according to the output second control signal and the third control signal.

[0007] Furthermore, the controlled adjustment module is: In the formula: For system state variables; y Output variables for the system; This is represented as the output signal of the fuzzy logic parameter dynamic adjustment module; For total disturbance; The derivative of the state variable is given by, where, For the differential components of the output; The differential component of the sum of disturbances; To control the gain; matrix parameters: A The state variable gain matrix; B For control gain matrix; C For the output gain matrix, E This is the total disturbance correlation matrix.

[0008] Furthermore, the first linearly extended state observer is: In the formula: Output variables for the system y Observed values; for Observed values; , The gain of the first linearly extended state observer; This represents the observation error of the first linearly extended state observer; , , They are respectively , , The differential value of .

[0009] Furthermore, the second linearly extended state observer is: In the formula: It is a system output variable y Observed values; Used to observe the first linear expansion state in the observer The remaining disturbances that were not observed are used to supplement the initial disturbance observation results; and The gain of the second linearly extended state observer; This represents the observation error of the second linearly extended state observer; , , They are respectively , , The differential value of .

[0010] Furthermore, the first linearly extended state observer and the second linearly extended state observer use the same gain value, that is: In the formula: This represents the initial observer bandwidth for the linear active disturbance rejection controller.

[0011] Furthermore, the fuzzy logic parameter dynamic adjustment module is as follows: In the formula: The initial gain of the module is dynamically adjusted to control fuzzy logic parameters; This refers to the change in the control law gain. For proportional gain; This is the rated mechanical angular velocity.

[0012] Furthermore, the fuzzy logic parameter dynamic adjustment module dynamically optimizes the bandwidth of the linear active disturbance rejection controller, realizing the real-time adaptive adjustment process of the linear active disturbance rejection controller parameters: Microgrid frequency deviation e and the rate of change of frequency deviation Δ eAs input to the fuzzy logic parameter dynamic adjustment module, membership functions are designed for the two input quantities respectively, and the universe of discourse of each input quantity is divided into seven fuzzy levels to calculate the membership degree of each input quantity corresponding to different fuzzy levels. The continuous and precise physical quantity is transformed into a fuzzy set that conforms to the rules of fuzzy logic, thus completing the fuzzification preprocessing of the input signal. Based on the fuzzy set of the input after fuzzification, fuzzy reasoning is carried out according to the fuzzy control rule table, combined with control experience and system logic. By matching the input fuzzy quantity with the rule base entries, the fuzzy decision result corresponding to the output variable is derived, forming the fuzzy control instruction to be parsed. Based on the generated fuzzy inference results, an appropriate defuzzification algorithm is used to perform numerical conversion, remove fuzzy semantics and eliminate fuzziness, and restore the abstract fuzzy decision results to precise control values ​​that can be recognized by the computer, thus completing the reverse mapping from fuzzy quantities to precise control quantities. The defuzzified precise control value is transformed into dynamic parameters that can be executed by the linear active disturbance rejection controller (ADRC), and directly sent to the ADRC to complete the parameter update, thereby realizing real-time closed-loop adjustment of the ADRC parameters.

[0013] Beneficial effects 1. The present invention proposes a microgrid construction based on the collaborative construction of an optimized linear active disturbance rejection controller and a virtual synchronous generation control model. In particular, the design of a cascaded linear extended state observer (ILESO-2) and a fuzzy logic parameter dynamic adjustment module in the optimized linear active disturbance rejection controller work together to specifically solve the technical defects of traditional PI control and conventional linear active disturbance rejection controllers in the secondary frequency regulation application of virtual synchronous generation control models, such as poor overshoot suppression, long response time, weak anti-interference ability, and inability to dynamically adapt parameters. This significantly improves the frequency stability, anti-interference ability, and practical engineering adaptability of microgrid network construction control.

[0014] 2. This invention eliminates reliance on engineering debugging experience by optimizing the dynamic adaptive adjustment of linear active disturbance rejection controller parameters; this invention introduces a fuzzy logic parameter dynamic adjustment module to adjust the frequency deviation... e and the rate of change of frequency deviation Δ e Using this as input, the bandwidth variation of the linear active disturbance rejection controller is adjusted and optimized. A fuzzy control rule table is established by combining the experience of power system control engineering experts, thereby realizing real-time dynamic optimization of the bandwidth of the linear active disturbance rejection controller.

[0015] 3. This invention optimizes the linear active disturbance rejection controller so that it can automatically adjust parameters according to the nonlinear changes in the system's operating state and the changes in inherent parameters. This solves the problems of traditional controllers relying on manual adjustment of parameters and fixed parameters that cannot adapt to changes in operating conditions, thus improving the disturbance rejection, speed and practical engineering operability of the control strategy.

[0016] 4. This invention enhances the microgrid's support capability for the main grid by constructing a network control performance through a virtual synchronous generation control model. At the same time, based on the secondary frequency regulation framework of the virtual synchronous generation control model, this invention deeply integrates the virtual inertia and damping characteristics of the linear active disturbance rejection controller with those of the virtual synchronous generation control model, derives and optimizes the transfer function of the control loop of the virtual synchronous generation control model, and constructs a closed-loop control system.

[0017] 5. This invention effectively improves the damping and inertia characteristics of the grid-type inverter in the virtual synchronous power generation control model, enabling the microgrid to provide more stable voltage and frequency support to the main power grid during grid faults and grid connection state switching, thereby enhancing the main grid's anti-disturbance performance and providing stable control guarantees for the efficient grid connection and local consumption of large-scale distributed renewable energy.

[0018] 6. This invention is highly versatile and scalable, possessing significant engineering application value. In particular, by adjusting the virtual inertia, damping coefficient, ILESO-2 observer bandwidth, and fuzzy control rules of the virtual synchronous power generation control model, this invention can adapt to AC microgrid systems of different capacities and renewable energy integration ratios. Furthermore, this invention can be directly compiled and ported to industrial controller hardware platforms such as DSPs and PLCs, without requiring large-scale modifications to the original microgrid topology. It is suitable for the network construction and control of microgrids, microgrid clusters, and new energy power plants with distributed energy sources such as photovoltaics, wind power, and hydropower, and possesses broad engineering promotion and practical application value. Attached Figure Description

[0019] Figure 1 This is a signal block diagram of the virtual synchronous power generation control model in this invention; Figure 2 This is a structural diagram of the active disturbance rejection controller in this invention; Figure 3 This is a schematic diagram of the membership function change in this invention; where: (a) voltage deviation and (b) rate of change of voltage deviation; Figure 4 This is a schematic diagram of the output of the fuzzy logic parameter dynamic adjustment module of the present invention; Figure 5 This is a frequency response simulation diagram from Embodiment 1 of the present invention; Figure 6 This is a frequency response simulation diagram from Embodiment 2 of the present invention.

[0020] Marker explanation: 100. Virtual synchronous power generation control model; 101. Active disturbance rejection controller; 102. Fuzzy logic parameter dynamic adjustment module; 101a. First linear extended state observer; 101b. Second linear extended state observer; 104. Controlled regulation module; Detailed Implementation

[0021] The following is in conjunction with the appendix Figure 1 -Appendix Figure 6 The present invention will be described in detail below: like Figure 1 As shown, the present invention provides a method for constructing a microgrid based on an optimized linear active disturbance rejection controller. The microgrid includes a virtual synchronous generation control model 100 and an active disturbance rejection controller 101. The active disturbance rejection controller 101 consists of a first linear extended state observer 101a, a second linear extended state observer 101b, a fuzzy logic parameter dynamic adjustment module 102, and a controlled adjustment module 104. like Figure 2 As shown in the figure: 1 / s Represents integration. s Represents differentiation operation, taking the reference signal r With output signal y The difference is calculated by amplifying the signal after passing through the fuzzy logic parameter dynamic adjustment module 102, and then subtracting the amplified signal from the signal outputs of the first linear extended state observer 101a and the second linear extended state observer 101b. r The reference signal input to the active disturbance rejection controller 101 is compared with the output signal y of the controlled regulation module. After difference calculation, the signal is amplified by the fuzzy logic parameter dynamic adjustment module 102. Simultaneously, the output signal undergoes a series of operations, including differentiation and integration, by the first linear extended state observer and the second state observer. Finally, a difference calculation is performed with the result of fuzzy control to achieve real-time tracking of the output signal. This includes: the microgrid comparing the input and output signals to generate an error, which is then amplified by a proportional gain and used to drive the active disturbance rejection controller 101 to output a compensated control signal; this signal is used to control the controlled regulation module 104 to achieve signal tracking output, including: The active disturbance rejection controller 101 converts the input signal into a state space and compares it with the actual output signal to form a tracking error signal; The fuzzy logic parameter dynamic adjustment module 102 amplifies the tracking error signal through proportional gain to form a first control signal; The first linear expansion state observer 101a outputs a second control signal based on the tracking error signal feedback term in the first control signal, the basic state of the control signal speed and position, and low-frequency disturbances in real time. The second linear extended state observer 101b compensates for high-frequency disturbances or uncertainties in real time based on the tracking error signal feedback term in the first control signal and outputs a third control signal. The controlled adjustment module 104 adjusts the controlled object according to the compensated control signals output by the first linear expansion state observer and the second linear expansion state observer; the output signal is consistent with the tracking signal.

[0022] The microgrid is based on the secondary frequency regulation framework of the virtual synchronous generation control model 100. The optimized linear active disturbance rejection controller 101 is composed of cascaded linear extended state observers. The fuzzy logic parameter dynamic adjustment module 102 realizes dynamic adjustment of the linear active disturbance rejection controller parameters, constructing a secondary frequency regulation network control strategy for the virtual synchronous generation control model that combines fast tracking, strong anti-interference, and small overshoot. The cascaded linear extended state observers include a first linear extended state observer 101a and a second linear extended state observer 101b. The specific technical solution is as follows: The virtual synchronous power generation control model 100 includes a second-order mathematical model constructed using virtual moment of inertia and damping coefficients, which serves as the model basis for grid control. At the same time, a microgrid topology containing photovoltaic power generation units and energy storage units is built, with the energy storage system serving as a power support unit to reduce the fluctuation of microgrid power output.

[0023] The virtual synchronous generator control model simulates the operating characteristics of a synchronous generator, utilizing virtual moment of inertia and damping characteristics to achieve dynamic power regulation and mitigate system frequency fluctuations. Its core structure matches the requirements of microgrid network control. For example... Figure 1 As shown, the mathematical model of the virtual synchronous generation control model is: In the formula: J This is a virtual moment of inertia; For mechanical torque; Electromagnetic torque; This is the damping torque; Mechanical power; Electromagnetic power; D The damping coefficient; To output power angle; It is the mechanical angular velocity; This is the rated mechanical angular velocity.

[0024] The process of constructing a cascaded linearly extended state observer: Based on the mathematical model of the virtual synchronous power generation control model, the controlled regulation module 104 (Plant) is obtained according to the following formula: In the formula: For system state variables; y Output variables for the system; This is represented as the output signal of the fuzzy logic parameter dynamic adjustment module; For total disturbance; The derivative of the state variable is given by, where, For the differential components of the output; The differential component of the sum of disturbances; To control the gain; matrix parameters: A The state variable gain matrix; B For control gain matrix; C For the output gain matrix, E This is the total disturbance correlation matrix.

[0025] The active disturbance rejection controller of the present invention utilizes a first linear extended state observer 101a and a second linear extended state observer 101b. , State update changes, with observation error The core feedback signal of the observer is used to estimate the frequency state and total disturbance in real time, and to perform dynamic correction through error feedback. The first linear extended state observer 101a and the second linear extended state observer 101b in this invention utilize only measurable outputs. and the input of the fuzzy logic parameter dynamic adjustment module Construct an internal observer model that is similar in structure to the real system, and continuously compare the outputs. Updates, changes, and reality The error is corrected through feedback, thus improving the observer state. Gradually approaching the expected reality.

[0026] To further optimize the perturbation observation performance, the second linearly extended state observer in this invention is designed based on the same principle as the first linearly extended state observer. Specifically: Constructing the first linear extended state observer 101a based on the secondary frequency regulation state of the virtual synchronous generation control model: In the formula: Output variables for the system y Observed values; for Observed values; , The gain of the first linearly extended state observer; This represents the observation error of the first linearly extended state observer; , , They are respectively , , The differential value of .

[0027] The second linearly extended state observer 101b is: It is a system output variable yObserved values; Used to observe the first linear expansion state in the observer The remaining disturbances that were not observed are used to supplement the initial disturbance observation results; and The gain of the second linearly extended state observer; This represents the observation error of the second linearly extended state observer; , , They are respectively , , The differential value of .

[0028] To further optimize the linear active disturbance rejection controller structure, the first linear extended state observer 101a and the second linear extended state observer 101b are made to have the same gain value, that is: In the formula: This represents the initial observer bandwidth for the linear active disturbance rejection controller; The fuzzy logic parameter dynamic adjustment module 102 is: In the formula: The initial gain of the module is dynamically adjusted to control fuzzy logic parameters; This refers to the change in the control law gain. For proportional gain; This is the rated mechanical angular velocity.

[0029] To address the problem that traditional controller parameters rely on engineering debugging experience and cannot adapt to nonlinear changes in the system, this invention introduces a fuzzy logic parameter dynamic adjustment module 102 to dynamically optimize the bandwidth of the optimized linear active disturbance rejection controller 101. This transforms the practical experience of engineering experts into computer-executable control rules, enabling real-time adaptive adjustment of controller parameters. The specific implementation design is as follows: Microgrid frequency deviation e and the rate of change of frequency deviation Δ e As the input to the fuzzy logic parameter dynamic adjustment module, according to e and Δ e The bandwidth of the linear active disturbance rejection controller is dynamically adjusted and optimized based on real-time changes.

[0030] In fuzzy adaptive control strategies, frequency deviation (denoted as...) is used as the basis for... e ) and the rate of change of frequency deviation Δ e(As two input variables of the fuzzy controller, membership functions are designed for these two input quantities (as shown in Table 1), and the universe of discourse of each input quantity is divided into seven fuzzy levels (e.g., {negative large, negative medium, negative small, zero, positive small, positive medium, positive large}, abbreviated as {NB, NM, NS, ZO, PS, PM, PB}). This invention uses this membership function to achieve the control of continuous frequency deviations such as... Figure 3 (a) Rate of change of frequency deviation as shown Figure 3 (b) fuzzification maps precise continuous input quantities to fuzzy sets, laying the foundation for subsequent fuzzy inference.

[0031] Based on the engineering expert knowledge and practical experience of power system control, a fuzzy adjustment rule table is established (as shown in Table 1):

[0032] Therefore, the characteristic output surface of the fuzzy logic parameter dynamic adjustment module 102, such as Figure 4 As shown. Figure 4 The different frequency deviations are clearly defined. e And deviation change Δ e Under the combined conditions, the optimal adjustment strategy for controller bandwidth is determined. Fuzzy inference is performed using the rule table 1, and the inference results are defuzzified to output the actual controller bandwidth adjustment amount, thereby achieving dynamic parameter adaptation.

[0033] The fuzzy logic parameter dynamic adjustment module 102 achieves fully automated control based on fuzzy control algorithms. The overall operation process is strictly divided into four core steps that are progressive and connected in a closed loop. No manual intervention is required throughout the process. The computer autonomously calculates and optimizes the parameters iteratively. The specific execution process is as follows: (1) Input fuzzification Using frequency deviation and the rate of change of frequency deviation as precise continuous input quantities of the module, combined with preset membership functions and seven fuzzy levels (negative large NB, negative medium NM, negative small NS, zero ZO, positive small PS, positive medium PM, positive large PB), the membership degree of each input quantity corresponding to different fuzzy levels is calculated, transforming the continuous precise physical quantity into a fuzzy set that conforms to fuzzy logic rules, thus completing the fuzzification preprocessing of the input signal.

[0034] (2) Fuzzy rule reasoning (based on Table 1) Based on the fuzzy set of input after fuzzification, and strictly following the established fuzzy control rule table in Table 1, fuzzy inference operations are carried out by combining control experience and system logic. By matching the input fuzzy quantity with the rule base entries, the fuzzy decision result corresponding to the output variable is derived, forming the fuzzy control instruction to be parsed.

[0035] (3) Defuzzification of the reasoning results Based on the fuzzy inference results generated in the previous step, an appropriate defuzzification algorithm is used to perform numerical conversion, remove fuzzy semantics and eliminate fuzziness, and restore the abstract fuzzy decision results to precise control values ​​that can be recognized by the computer, thus completing the reverse mapping from fuzzy quantities to precise control quantities.

[0036] (4) Output of active disturbance rejection controller parameters The defuzzified precise control value is transformed into dynamic parameters that can be executed by the linear active disturbance rejection controller (ADRC), and directly sent to the ADRC to complete the parameter update, thereby realizing real-time closed-loop adjustment of the ADRC parameters.

[0037] Through the aforementioned standardized and automated process, computers replace manual labor to complete online parameter adjustments. During the real-time operation of the microgrid system, the key parameters of the linear active disturbance rejection controller are dynamically adjusted and optimized, enabling it to quickly adapt to complex operating conditions such as nonlinear fluctuations and inherent parameter drift in the microgrid, thereby significantly improving the robustness, anti-interference capability, and operational stability of the control system.

[0038] This invention, through the above-mentioned structural improvements and strategy integration, forms an integrated control scheme of "accurate disturbance estimation, tracking error optimization, and dynamic parameter adaptation". It solves the technical defects of traditional control strategies from three dimensions: observer structure, control law design, and parameter adjustment, and realizes the rapid stabilization of microgrid frequency under the grid control of virtual synchronous power generation control model.

[0039] The following are applications of this invention: Using MATLAB / Simulink as the core simulation and verification platform, the performance verification and parameter optimization of the control strategy are completed by reproducing typical dynamic operating conditions of microgrid load abrupt changes. Then, relying on an industrial-grade digital signal processor (DSP), hardware porting and engineering implementation are achieved. The overall implementation process follows the logic of simulation modeling - operating condition verification - parameter optimization - hardware porting - field commissioning. It can accurately adapt to AC microgrid network control scenarios including photovoltaic power generation and energy storage systems, and the implementation process is highly consistent with the simulation verification results. The MATLAB / Simulink simulation platform (2020b and above) is used, along with the SimPowerSystems power electronics simulation toolbox and the FuzzyLogic toolbox, which are used for microgrid topology modeling with a grid-type virtual synchronous generation control model, linear active disturbance rejection controller, and fuzzy logic parameter dynamic adjustment module development, respectively. The implementation steps and details are as follows: The core parameters of the simulation model are kept consistent with those of the physical test platform. The rated voltage of the virtual synchronous power generation control model is set to 380V and the rated frequency to 50Hz. Key parameters such as virtual moment of inertia and damping coefficient, as well as the bandwidth of the linear active disturbance rejection controller observer, are configured. The fuzzy logic parameter dynamic adjustment module is initially designed according to the classical design criteria of linear active disturbance rejection control. The parameters of the photovoltaic simulation source, energy storage battery pack and load device are configured according to the actual application scenario of the microgrid to meet the simulation requirements of dynamic operating conditions such as sudden load increase and decrease.

[0040] Based on the second-order electromechanical motion characteristics of a synchronous generator, a core mathematical model for a virtual synchronous generator control model is constructed, integrating virtual inertia, damping, and voltage droop control modules to simulate the operating characteristics of the synchronous generator. Simultaneously, the theoretical derivation of the linear active disturbance rejection control module controller is completed according to the technical solution of this invention, including the formula derivation and parameter initial setting of the controlled regulation module, the linear extended state observer (TLESO), ILESO-2, and the fuzzy logic parameter dynamic adjustment module, providing complete theoretical support for simulation modeling.

[0041] An AC microgrid simulation model was built in Simulink, comprising a photovoltaic power generation unit, an energy storage system, a virtual synchronous power generation control model inverter, a common bus, adjustable loads, and real-time measurement modules for frequency and voltage at a 10kHz sampling frequency. The energy storage system was connected to the DC side of the virtual synchronous power generation control model inverter as a power support unit. Simultaneously, comparative simulation models of conventional PI and conventional linear active disturbance rejection control were built. According to this invention, modular simulation modeling and integration of the controller for the conventional linear active disturbance rejection control strategy were completed. The state space of the virtual synchronous power generation control model and the cascaded linear extended state observer were built sequentially, along with a system based on… Figure 3 The membership function and the parameter dynamic adjustment module of the fuzzy control rule table in Table 1 are then deeply integrated with the secondary frequency regulation framework of the virtual synchronous power generation control model. The control loop transfer function is configured to realize signal communication and coordinated control, and a complete linear active disturbance rejection controller-virtual synchronous power generation control model secondary frequency regulation closed-loop simulation system is constructed.

[0042] Example 1 In the optimized linear active disturbance rejection controller-virtual synchronous power generation control model secondary frequency modulation closed-loop simulation system, this invention conducts dynamic response simulation tests on the system frequency to a sudden increase and decrease in load of 20KW. This verifies the performance of the optimized linear active disturbance rejection controller of this invention and compares its advantages and disadvantages with conventional PI control and traditional linear active disturbance rejection controllers. The specific simulation implementation process is as follows: System frequency response simulation under a sudden load increase of 20KW: Simulation initial conditions setting: Start the microgrid simulation model, set the photovoltaic simulation source to output power stably at rated power, the energy storage unit to be in normal power support state, and the initial load to be configured as 30KW. Let the simulated microgrid operate stably until the microgrid frequency and bus voltage reach steady state values ​​(frequency stabilizes at 50Hz, voltage stabilizes at 380V). Load surge triggering: After the simulation system reaches steady state, the load is suddenly increased from 30KW to 50KW through the adjustable load module of the simulation model to simulate the typical disturbance condition of sudden load increase in actual microgrid operation. Data Acquisition and Recording: The simulation data acquisition function is activated to record the microgrid frequency variation curves over time under three control strategies: optimized linear active disturbance rejection controller, PI bus, etc., generating a simulation diagram of the system frequency response to a sudden 20kW load increase. Figure 5 It also records core performance indicators such as frequency drop amplitude, overshoot, settling time, and steady-state recovery error. Preliminary analysis: By comparing the frequency response curves and performance indicators of the three control strategies, the ability of the optimized linear active disturbance rejection controller to suppress frequency fluctuations under sudden load increases is verified.

[0043] Simulation of system frequency response under sudden load reduction of 20KW: Simulation initial conditions setting: Restart the microgrid simulation model, the output status of the photovoltaic and energy storage systems is consistent with operating condition 1, the initial load is configured as 50KW, and the simulation system is allowed to run stably until the frequency and voltage reach a steady state; Load reduction trigger: After the system is in steady state, the load is suddenly reduced from 50KW to 30KW through the adjustable load module to simulate the typical disturbance condition of sudden load reduction in the actual operation of the microgrid. Data Acquisition and Recording: Real-time acquisition and recording of the dynamic frequency change curves of the system under three control strategies, generating a simulation diagram of the system frequency response to a sudden load reduction of 20KW. Figure 6 Simultaneously, it records core performance indicators such as frequency rise amplitude, overshoot, settling time, and steady-state error; Preliminary analysis: By comparing the frequency recovery characteristics of the three control strategies, the frequency stabilization capability of the conventional linear active disturbance rejection control module is verified under the condition of sudden load reduction.

[0044] Based on simulation data and frequency response curves under load surges and drops of 20kW, this invention, using an improved linear active disturbance rejection control (ADC) module control strategy, effectively suppresses system frequency fluctuations caused by load surges compared to conventional PI and traditional linear ADC modules. The frequency drop / rise amplitude is smaller with no significant overshoot, achieving overshoot-free frequency tracking. The system frequency can quickly converge to the rated value of 50Hz, with shorter adjustment time and minimal steady-state error, significantly improving the dynamic response performance of the virtual synchronous power generation control model's secondary frequency regulation. Furthermore, the control strategy exhibits stronger disturbance rejection capability, with no severe frequency oscillations under load surges, resulting in superior operational stability. Simultaneously, the fuzzy logic parameter adjustment module can dynamically adjust parameters in real time according to frequency deviation, further optimizing frequency response characteristics, fully verifying the rationality and superiority of the control strategy of this invention.

[0045] Based on the simulation results of a 20kW load surge and desurge, we conducted fine-tuning and optimization of the core parameters of the linear active disturbance rejection controller: First, we fine-tuned the observer bandwidth using the pole placement method combined with the frequency response curve analysis results. w 0. Bandwidth of the fuzzy logic parameter dynamic adjustment module k p The optimal range of parameter values ​​is determined with the objectives of minimizing system frequency overshoot, shortest settling time, and minimum steady-state error; this is then combined with the frequency deviation observed in the simulation. e and the rate of change of deviation Δ e The dynamic change pattern was analyzed, and individual rules in the fuzzy control rule table in Table 1 were fine-tuned to optimize the sensitivity and timeliness of parameter adjustment. Finally, the optimized parameters were substituted into the simulation system, and simulations of two types of load change conditions were repeatedly executed to verify the effect. If the optimal result was not achieved, fine-tuning and iterative simulation were continued until the optimized linear active disturbance rejection controller achieved fast system frequency, no overshoot, and high-precision stability under the two types of conditions. Finally, the optimal parameter set of the optimized linear active disturbance rejection controller was determined and saved, providing accurate and reliable parameter basis for subsequent hardware porting.

[0046] The optimized linear active disturbance rejection controller-virtual synchronous generation control model, after debugging, was used to construct a microgrid and applied to pilot projects of small distributed microgrids, such as industrial parks, photovoltaic power plant-supporting microgrids, and independent microgrids in remote areas. It was continuously operated for 3-6 months, with real-time monitoring of key indicators such as system frequency, voltage, and renewable energy integration rate to verify the long-term operational stability and reliability of the control strategy. For microgrid systems of different capacities and different renewable energy (PV / wind / hydro) integration ratios, only the basic parameters of the virtual synchronous generation control model (virtual inertia, damping coefficient, voltage droop coefficient), bandwidth, and the domain of discourse of the fuzzy logic module need to be adjusted; no modification to the control strategy is required. The core architecture allows for rapid adaptation and possesses excellent versatility. The linear active disturbance rejection controller of this invention can be directly extended to frequency control in complex hybrid systems of water-wind-solar-pumped storage-new energy storage, adapting to the high penetration of renewable energy and complex and variable load conditions in new power systems, providing technical support for the stable operation of multi-energy complementary systems. By combining the control strategy of this invention with microgrid cluster coordinated control technology, the linear active disturbance rejection control modules of multiple microgrids can be linked through a communication network, improving the overall frequency stability and disturbance rejection capability of regional microgrid clusters, and providing a guarantee for the efficient grid connection and local consumption of large-scale distributed renewable energy.

Claims

1. A method for constructing a microgrid based on an optimized linear active disturbance rejection controller, characterized in that: The method is based on a microgrid, which includes a virtual synchronous generation control model and an active disturbance rejection controller. The active disturbance rejection controller comprises a first linear extended state observer, a second linear extended state observer, a fuzzy logic parameter dynamic adjustment module, and a controlled adjustment module; including: The active disturbance rejection controller converts the input signal into a state space and compares it with the actual output signal to form a tracking error signal. The fuzzy logic parameter dynamic adjustment module amplifies the tracking error signal through proportional gain to form a first control signal; The first linear extended state observer adjusts the basic state of the speed and position of the first control signal and outputs a second control signal in real time based on the tracking error signal feedback term in the first control signal and low-frequency disturbances. The second linear extended state observer compensates for high-frequency disturbances or uncertainties in real time based on the tracking error signal feedback term in the first control signal and outputs a third control signal. The controlled adjustment module adjusts the controlled object to drive its output signal to be consistent with the tracking signal according to the output second control signal and the third control signal.

2. The method according to claim 1, characterized in that: The controlled adjustment module is: In the formula: For system state variables; y Output variables for the system; This is represented as the output signal of the fuzzy logic parameter dynamic adjustment module; For total disturbance; Differentiate the state variables; where, For the differential components of the output; The differential component of the sum of disturbances; To control the gain; matrix parameters: A The state variable gain matrix; B For control gain matrix; C This is the output gain matrix; E This is the total disturbance correlation matrix.

3. The method according to claim 2, characterized in that: The first linearly extended state observer is: In the formula: Output variables for the system y Observed values; for Observed values; , The gain of the first linearly extended state observer; This represents the observation error of the first linearly extended state observer; , , They are respectively , , The differential value of .

4. The method according to claim 3, characterized in that: The second linearly extended state observer is: In the formula: variables It is a system output variable y Observed values; Used to observe the first linear expansion state in the observer The remaining disturbances that were not observed are used to supplement the initial disturbance observation results; and The gain of the second linearly extended state observer; This represents the observation error of the second linearly extended state observer; , , They are respectively , , The differential value of .

5. The method according to claim 3 or 4, characterized in that; The first linearly extended state observer and the second linearly extended state observer use the same gain value, that is: In the formula: This represents the initial observer bandwidth for the linear active disturbance rejection controller.

6. The method according to claim 5, characterized in that; The fuzzy logic parameter dynamic adjustment module is: In the formula: The initial gain of the module is dynamically adjusted to control fuzzy logic parameters; This refers to the change in the control law gain. For proportional gain; This is the rated mechanical angular velocity.

7. The method according to claim 1, characterized in that; The fuzzy logic parameter dynamic adjustment module dynamically optimizes the bandwidth of the linear active disturbance rejection controller, realizing the real-time adaptive adjustment process of the linear active disturbance rejection controller parameters: Microgrid frequency deviation e and the rate of change of frequency deviation Δ e As input to the fuzzy logic parameter dynamic adjustment module, membership functions are designed for the two input quantities respectively, and the universe of discourse of each input quantity is divided into seven fuzzy levels to calculate the membership degree of each input quantity corresponding to different fuzzy levels. The continuous and precise physical quantity is transformed into a fuzzy set that conforms to the rules of fuzzy logic, thus completing the fuzzification preprocessing of the input signal. Based on the fuzzy set of the input after fuzzification, fuzzy reasoning operations are carried out in accordance with the fuzzy control rule table and combined with control experience and system logic. By matching the input fuzzy quantity with the rule base entries, the fuzzy decision result corresponding to the output variable is derived, and the fuzzy control instruction to be parsed is formed. Based on the generated fuzzy inference results, an appropriate defuzzification algorithm is used to perform numerical conversion, remove fuzzy semantics and eliminate fuzziness, and restore the abstract fuzzy decision results to precise control values ​​that can be recognized by the computer, thus completing the reverse mapping from fuzzy quantities to precise control quantities. The defuzzified precise control value is transformed into dynamic parameters that can be executed by the linear active disturbance rejection controller (ADRC), and directly sent to the ADRC to complete the parameter update, thereby realizing real-time closed-loop adjustment of the ADRC parameters.

Images