Microgrid control method and system

Abstract

The invention provides a micro-grid control method and system, and relates to the technical field of micro-grid control, and the method comprises the steps: obtaining the multi-source data of equipment, and carrying out the calculation according to the multi-source data, and obtaining a trigger function value; dynamically judging whether a trigger function value is greater than zero or not according to the multi-physical-quantity coupling trigger function so as to judge whether to trigger a control action or not; and when the trigger function value is greater than zero, triggering a control action, optimizing the external characteristic parameters and a load shedding instruction through an MPC algorithm, iteratively correcting the external characteristic parameters of the MPC algorithm according to a closed loop mechanism to generate a target control strategy, and cooperatively controlling the micro-grid based on the target control strategy. According to the invention, through deep fusion of dynamic event triggering and virtual external characteristic adaptation, efficient, stable and cooperative control of the micro-grid in a complex dynamic scene is realized.

Description

Technical Field

[0001] The present invention relates to the field of microgrid control technology, and in particular to a microgrid control method and system. Background Art

[0002] As the core carrier for absorbing high-penetration renewable energy, microgrids must achieve multi-source coordinated control and stability assurance in dynamic scenarios (such as sudden load changes, new energy fluctuations, and fault isolation). However, existing technologies in the field of multi-source coordinated control still have significant limitations, mainly manifested in the lack of adaptability due to fixed parameter design, prominent control response lag problems, lack of collaborative optimization of the design of external characteristic parameters, and the lack of multi-source coordinated control mechanisms. Specifically: At present, existing microgrid control methods generally rely on fixed parameter designs, such as droop control and virtual inertia control. These methods perform well under stable operating conditions, but are prone to power oscillation and frequency instability problems in complex dynamic scenarios. Secondly, existing control methods usually rely on periodic sampling or fixed threshold trigger mechanisms, which leads to delayed event response and makes it difficult to meet the rapid dynamic adjustment needs of microgrids. Thirdly, the design of external characteristic parameters (such as droop rate and virtual inertia) in existing technologies is often isolated from real-time operating conditions and fails to achieve multi-dimensional collaborative optimization. Furthermore, existing technologies mostly optimize single equipment or specific scenarios, and lack an overall design for multi-source collaboration in microgrids. Summary of the Invention Based on this, the purpose of the present invention is to provide a microgrid control method and system for dynamically adjusting the external characteristic parameters of each device to achieve coordinated optimization of multiple power sources, multiple loads and energy storage systems, thereby ensuring the stable operation of the microgrid.

[0003] One aspect of the present invention provides a microgrid control method, comprising: Acquire multi-source data of the equipment, the equipment including photovoltaic inverter, energy storage BMS, load controller and bus PMU, the multi-source data including photovoltaic output power, load power change rate, bus voltage deviation, frequency change rate, energy storage SOC and temperature change rate to obtain trigger function value, and calculate the trigger function value based on the multi-source data; Dynamically determine whether the trigger function value is greater than zero based on the multi-physical quantity coupling trigger function to determine whether to trigger the control action; When the trigger function value is greater than zero, the control action is triggered. The external characteristic parameters and the load shedding instruction are optimized by the MPC algorithm, and the external characteristic parameters of the MPC algorithm are iteratively corrected according to the closed-loop mechanism to generate a target control strategy so as to collaboratively control the microgrid based on the target control strategy. The external characteristic parameters include the frequency deviation weight factor, the harmonic suppression weight factor and the control parameters. The control parameters include the dynamic droop rate, the virtual impedance and the virtual inertia.

[0004] The above-mentioned microgrid control method achieves efficient, stable and coordinated control of the microgrid in complex dynamic scenarios by deeply integrating dynamic event triggering with virtual external characteristic adaptation. Specifically, the photovoltaic inverter and load controller are simultaneously involved in data acquisition and control execution to achieve integrated perception and control. The photovoltaic penetration rate is obtained based on the photovoltaic output power and load power, and the photovoltaic penetration rate is dynamically associated with the dynamic droop rate, virtual inertia, energy storage SOC, and temperature change rate for dynamic parameter adaptation. Furthermore, the frequency domain impedance is injected through the photovoltaic inverter to specifically solve the harmonic problem of new energy grid connection. By dynamically adjusting the external characteristic parameters of each device, the coordinated optimization of multiple power sources, multiple loads and energy storage systems is achieved to ensure the stable operation of the microgrid.

[0005] In addition, the microgrid control method according to the present invention may also have the following additional technical features: Furthermore, the expression of the multi-physics coupling trigger function is: ; Where, E Indicates the trigger function value, which is used to determine whether to trigger the control action. When E>0, the control action is triggered; Indicates the load power change rate; Indicates the photovoltaic power change rate; V Indicates voltage deviation; Indicates the temperature change rate of the energy storage battery; α 1, α 2, β , c are weight coefficients, representing the photovoltaic power weight factor, load power change rate weight factor, voltage deviation weight factor, and dynamic temperature weight factor respectively; Represents the temperature influence coefficient.

[0006] Furthermore, c The dynamic adjustment formula is: c new =c 0﹒ (1+ k SOC ﹒ (1-SOC / 100)+ k priority ﹒ P load ); Where, c new Indicates the trigger threshold of the temperature weight factor after dynamic adjustment; c 0 indicates the baseline trigger threshold; k SOC Indicates the energy storage charge state influence coefficient; energy storage SOC indicates the energy storage charge state;k priority Indicates the load priority influence coefficient; P load Indicates the total load power of the system at the current moment.

[0007] Furthermore, the calculation formula of the dynamic droop rate is: ; K droop Indicates the dynamic droop rate, which characterizes the slope of the power-frequency droop relationship; K 0 represents the base sag rate; k 1 represents the load factor influence coefficient; P load Indicates the total load power of the system at the current moment; P max Indicates the maximum allowable power of the system; k 2 represents the temperature change rate influence coefficient; k 3 represents the energy storage SOC impact coefficient; Indicates the temperature change rate of the energy storage battery.

[0008] Furthermore, the calculation formula of virtual inertia is: ; J virtual represents virtual inertia; J 0 represents the base virtual inertia; k 4 represents the frequency change rate influence coefficient; Indicates the rate of change of frequency; k 5 represents the photovoltaic penetration rate influence coefficient; P renew represents the photovoltaic penetration rate, and the calculation formula is: ; P PV Indicates photovoltaic output power; P demand The total load power of the system is summarized from the real-time data reported by the load controller.

[0009] Furthermore, the calculation formula of virtual impedance is: ; Where Z vir ( s ) represents the virtual impedance frequency domain model; K p represents the proportional gain; K d represents the differential gain; K irepresents the integral gain; f c represents the cutoff frequency; s represents a complex frequency variable, s = s + yes ,in: s is the real part, representing the attenuation coefficient; oh is the imaginary part, indicating the angular frequency; j represents the imaginary unit, j 2 = −1.

[0010] Furthermore, in the MPC algorithm, multi-objective optimization is achieved through the cost function, which is: ; Where, , represents the set of control input variables that need to be optimized, where K droop represents the dynamic droop rate, J virtual represents the virtual inertia, K p represents the proportional gain, K d represents the differential gain, K i represents the integral gain, L oad_shedding Indicates load switching; represents the frequency deviation weight coefficient; Indicates the THD weight coefficient; Power loss weight coefficient; l Represents the energy storage SOC penalty coefficient; represents the predicted value of frequency deviation at the next k moments; THD k Represents the predicted value of harmonic distortion rate at time k; k represents the discrete time index in the prediction time domain; SOC penalty represents the energy storage SOC out-of-bounds penalty term; N represents the prediction time domain; P loss Represents the comprehensive power loss of the system, including energy storage PCS loss, PV inverter loss and line transmission loss; J Indicates the comprehensive performance indicators of the system.

[0011] Furthermore, the system comprehensive power loss P loss In the calculation formula of energy storage PCS loss, the formula is: ; P loss,PCSIndicates the energy storage PCS loss; I PCS Indicates PCS current; R PCS Indicates PCS resistance; f SW Indicates the switching frequency; C SW Represents single switching loss; The calculation formula for photovoltaic inverter loss is: ; Where, Indicates the efficiency of the photovoltaic inverter; P PV Indicates photovoltaic output power; The formula for calculating line transmission loss is: ; Where, I line Indicates a line; R line Indicates the equivalent resistance of the line; The calculation formula for the system comprehensive power loss is: P loss = P loss,PCS + P loss,PV + P loss,line ; Where, P loss,PCS Indicates the energy storage PCS loss; P loss,PV Indicates the photovoltaic inverter loss; P loss,line Indicates line transmission loss.

[0012] Another aspect of the present invention provides a microgrid control system, the system comprising: An acquisition module is configured to acquire multi-source data of equipment, including a photovoltaic inverter, an energy storage BMS, a load controller, and a bus PMU. The multi-source data includes photovoltaic output power, load power change rate, bus voltage deviation, frequency change rate, energy storage SOC, and temperature change rate, and calculate a trigger function value based on the multi-source data. A judgment module, used to dynamically judge whether the trigger function value is greater than zero based on the multi-physical quantity coupling trigger function to determine whether to trigger a control action; The control module is used to trigger a control action when the trigger function value is greater than zero, optimize the external characteristic parameters and the load shedding instruction through the MPC algorithm, and iteratively correct the external characteristic parameters of the MPC algorithm according to the closed-loop mechanism to generate a target control strategy, so as to collaboratively control the microgrid based on the target control strategy. The external characteristic parameters include a frequency deviation weight factor, a harmonic suppression weight factor, and control parameters. The control parameters include a dynamic droop rate, a virtual impedance, and a virtual inertia.

[0013] Another aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which implements the microgrid control method as described above when the program is executed by a processor.

[0014] Another aspect of the present invention provides a data processing device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the microgrid control method as described above when executing the program. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flow chart of the microgrid control method of the present invention; The following specific embodiments will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION

[0016] To facilitate understanding of the present invention, the present invention will be described more fully below with reference to the accompanying drawings. The drawings illustrate several embodiments of the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and comprehensive understanding of the present invention.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one skilled in the art to which this invention pertains. The terms used in this specification of the present invention are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0018] During microgrid operation, complex dynamic scenarios such as load fluctuations, sudden changes in photovoltaic power generation, and energy storage system failures can severely impact system power balance and stability. To address these challenges, this paper proposes a microgrid control method and system. This system uses a sensing layer to collect real-time information about photovoltaic inverter output power, energy storage battery status, load power, and bus electrical quantities (voltage, frequency, and harmonics). The decision layer utilizes an MPC algorithm to dynamically determine whether control strategies need to be adjusted based on a multi-physics coupling trigger function. When the system detects events such as sudden load changes, sudden drops in photovoltaic power, or energy storage system failures, the MPC algorithm is triggered to optimize the power-frequency droop rate of the energy storage converter, the virtual inertia and frequency-domain impedance parameters of the photovoltaic inverter, and the load shedding instructions of the load controller. During the optimization process, the MPC algorithm uses a multi-objective cost function to balance system frequency stability, harmonic suppression, operating efficiency, and energy storage system health to generate an optimal control strategy. After the execution layer adjusts device parameters according to the strategy, it continuously monitors actual operating conditions (such as frequency deviation, harmonic content, and power balance error) through a closed-loop feedback mechanism. It dynamically corrects the external characteristic parameters of the MPC algorithm until the system stabilizes and meets preset indicators (frequency deviation less than 0.05Hz, total harmonic distortion less than 3%, and power balance error less than 5kW). This approach addresses the rigidity of traditional control parameters, delayed response, and insufficient multi-source coordination, significantly improving the stability and efficiency of microgrids in complex dynamic scenarios.

[0019] To facilitate understanding of the present invention, several embodiments of the present invention are provided below. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and comprehensive disclosure of the present invention.

[0020] Example 1 See also Figure 1 , which shows a microgrid control method in a first embodiment of the present invention, the method includes steps S101 to S104: S101. Acquire multi-source data of equipment, including photovoltaic inverters, energy storage BMS, load controllers, and bus PMUs. The multi-source data includes photovoltaic output power, load power change rate, bus voltage deviation, frequency change rate, energy storage SOC, and temperature change rate. Calculate the trigger function value based on the multi-source data.

[0021] In this embodiment, the system adopts a three-layer architecture consisting of a perception layer, a decision layer, and an execution layer to achieve multi-source data fusion and closed-loop control. Specifically, the perception layer collects electrical quantity, device status, and load priority data. Its core devices include photovoltaic inverters, energy storage battery management systems (BMSs), load controllers, and bus phase measurement units (PMUs). The decision layer performs dynamic event triggering, model predictive control, and command generation. Its core devices include edge controllers. In this embodiment, model predictive control uses the MPC (Model Predictive Control) algorithm. The execution layer adjusts device external characteristic parameters and executes load shedding or power limiting operation commands. Device external characteristic parameters include dynamic droop and virtual inertia. The core devices of the execution layer include energy storage power conversion systems (PCSs), photovoltaic inverters, and load controllers.

[0022] Furthermore, the energy storage SOC (State of Charge, SOC for short) of the energy storage BMS, temperature T and the temperature change rate of the energy storage battery are obtained. The main data collected by the bus PMU include three-phase voltage, three-phase current, frequency and total harmonic distortion (THD). Among them, the three-phase voltage is used to calculate the voltage deviation △V and subsequently extract the harmonic characteristics through the FFT algorithm; the three-phase current is used to calculate the power and harmonic content; the frequency is used to monitor the system frequency change rate. ; Total harmonic distortion is used to evaluate power quality. Get photovoltaic output power P PV , power change rate and DC side voltage V . Get the total load power of the system at the current moment P load , power change rate , priority label, and switch status.

[0023] After the decision layer adjusts the data, the adjusted data is output to the execution layer to control the equipment. Specifically, the virtual impedance Z of the photovoltaic inverter is adjusted. vir and power limit percentage; adjust the dynamic droop rate of the energy storage PCS K droop And inject virtual inertia J virtualThe load controller is instructed to cut the load. The device performs closed-loop control based on the parameters fed back by the execution layer, optimizing the device data until the optimal control strategy is achieved. During this closed-loop feedback process, data verification and iterative optimization are required. Specifically, data verification involves the decision layer comparing the command value with the actual value and calculating the deviation. Iterative optimization involves adjusting the MPC algorithm's weight coefficients based on the deviation.

[0024] S102 , dynamically determining whether a trigger function value is greater than zero based on the multi-physical quantity coupling trigger function to determine whether a control action is triggered.

[0025] When the trigger function value is greater than zero, the control action is triggered, and step S103 is executed; When the trigger function value is not greater than zero, no control action is triggered, and step S104 is executed.

[0026] Specifically, the expression of the multi-physics coupling trigger function is: ; Where, E Represents the trigger function value, dimensionless, used to determine whether to trigger the control action. When E>0, the control action is triggered; Indicates the load power change rate in kW / s, which indicates the power mutation reported by the load controller, such as large load switching; Indicates the photovoltaic power change rate, indicating a sudden drop in photovoltaic inverter output, such as cloud cover; V Indicates voltage deviation, unit v, representing the voltage amplitude U monitored by the bus PMU rms With rated value U n In this embodiment, U rms Calculated by the root mean square value of the three-phase voltage, U n =380v; Indicates the temperature change rate of the energy storage battery, in °C / s. It represents the converter junction temperature change rate reported by the energy storage BMS and is used to protect the device from overheating. α 1, α 2, β , c are weight coefficients, representing photovoltaic power weight factor, load power change rate weight factor, voltage deviation weight factor, and dynamic temperature weight factor. The typical values are α 1=0.8, α 2=0.7, β =0.5, c =0.2, Indicates the temperature influence coefficient, with a typical value of 0.15, which is used to amplify the effect of temperature changes on trigger sensitivity.

[0027] Further, cThe dynamic adjustment formula is: c new =c 0﹒ (1+ k SOC ﹒ (1-SOC / 100)+ k priority ﹒ P load ); Where, c new Represents the trigger threshold of the temperature weight factor after dynamic adjustment, dimensionless; c 0 represents the base trigger threshold, with a typical value of 1.2; k SOC Indicates the energy storage state of charge influence coefficient, with a typical value of 0.15. The trigger sensitivity is increased at low energy storage SOC. Energy storage SOC indicates the energy storage state of charge, in %, reported by the energy storage BMS, and ranges from 0% to 100%. k priority Indicates the load priority influence coefficient, with a typical value of 0.2. High-priority loads reduce the trigger threshold; P load Indicates the total load power of the system at the current moment. Load priority labels range from 1 to 5, with level 1 being the highest. This is reported by the load controller.

[0028] S103. Optimize external characteristic parameters and load shedding instructions through the MPC algorithm, and iteratively correct the external characteristic parameters of the MPC algorithm according to the closed-loop mechanism to generate a target control strategy, so as to collaboratively control the microgrid based on the target control strategy. The external characteristic parameters include a frequency deviation weight factor, a harmonic suppression weight factor, and control parameters. The control parameters include a dynamic droop rate, a virtual impedance, and a virtual inertia.

[0029] Furthermore, the calculation formula of the dynamic droop rate is: ; K droop Indicates the dynamic droop rate, in Hz / kW, which represents the slope of the power-frequency droop relationship; K 0 represents the base droop rate, with a typical value of 0.05 Hz / kW; k 1 represents the load factor influence coefficient, with a typical value of 0.15, which increases the droop rate at high loads; P load Indicates the total load power of the system at the current moment, reported by the load controller, in kW; P max Indicates the maximum allowable power of the system, in kW, determined by the system design; k2 represents the temperature change rate influence coefficient, with a typical value of 0.1. When the temperature rises too quickly, the droop rate is reduced to protect the equipment; k 3 represents the energy storage SOC influence coefficient, with a typical value of 0.1. When the energy storage SOC is low, the droop rate is reduced to reduce energy storage discharge; Indicates the temperature change rate of the energy storage battery, unit is ℃ / s, , where T k Indicates the current temperature, T K-1 Indicates the temperature at the previous moment, and △t indicates the sampling time interval.

[0030] Furthermore, the calculation formula of virtual inertia is: ; J virtual It represents the virtual inertia, unit is s, simulating the rotational inertia of the traditional generator; J 0 represents the baseline virtual inertia, and the typical value is 5s; k 4 represents the frequency change rate influence coefficient, with a typical value of 0.2. When the frequency changes drastically, the inertia support is enhanced; Indicates the frequency change rate in Hz / s, monitored by the bus PMU; k 5 represents the photovoltaic penetration coefficient, with a typical value of 0.15. At high penetration rates, inertia is increased to suppress frequency fluctuations. P renew It represents the photovoltaic penetration rate, which is calculated by the ratio of the photovoltaic inverter output power to the total load. It is dimensionless, where the photovoltaic penetration rate P renew The calculation formula is: ; P PV Indicates photovoltaic output power; P demand The total load power of the system is summarized from the real-time data reported by the load controller.

[0031] Furthermore, the calculation formula of virtual impedance is: ; Where Z vir ( s ) represents the virtual impedance frequency domain model, used to suppress broadband oscillation, unit Ω; K p Represents proportional gain, unit Ω, typical value is 0.4, controls the impedance amplitude in the low frequency band; K d It represents the differential gain, in Ω·s, with a typical value of 0.2, which suppresses high-frequency harmonics, such as noise near the switching frequency; Ki Indicates the integral gain, unit Ω / s, typical value is 0.1, which enhances the fundamental wave power distribution capability; f c Indicates the cutoff frequency in rad / s, which is set to half the switching frequency, for example, 2.5kHz for a 5kHz system; s represents a complex frequency variable, s = s + yes ,in: s is the real part, representing the attenuation coefficient; oh is the imaginary part, indicating the angular frequency; j represents the imaginary unit, j 2 = −1.

[0032] Furthermore, in the MPC algorithm, multi-objective optimization is achieved through the cost function, which is: ; Where, , represents the set of control input variables that need to be optimized, where K droop represents the dynamic droop rate, J virtual represents the virtual inertia, K p represents the proportional gain, K d represents the differential gain, K i represents the integral gain, L oad_shedding Indicates load switching, by adjusting the parameters in u, the cost function J Minimization to achieve multi-objective system optimization, such as frequency stability, harmonic suppression, and efficiency improvement; Indicates the frequency deviation weight coefficient, which prioritizes ensuring system frequency stability. It is dimensionless and has a typical value of 0.6. It increases dynamically according to the frequency drop rate. Indicates the THD weight coefficient, which suppresses harmonic distortion. It is dimensionless, with a typical value of 0.4 and is dynamically adjusted according to the harmonic content. Power loss weight coefficient, optimizes system operation efficiency, dimensionless, typical value is 0.3, dynamically adjusted according to energy storage SOC; l Represents the energy storage SOC penalty coefficient, which protects the health of energy storage. It is dimensionless, with a typical value of 0.2, and is dynamically adjusted according to the degree of energy storage SOC out of bounds. Represents the predicted value of frequency deviation at the next k moments, in Hz; THD k Represents the predicted value of harmonic distortion rate at time k, unit: %; k represents the discrete time index in the prediction time domain; SOC penalty represents the energy storage SOC out-of-bounds penalty term, dimensionless; N represents the prediction time domain (typical value is 5 sampling periods), dimensionless; P loss Represents the system's comprehensive power loss, in kW, including energy storage PCS loss, PV inverter loss, and line transmission loss; J Represents the comprehensive performance indicators of the system, and generates the optimal control strategy by weighted calculation of frequency deviation, harmonic distortion rate, power loss and energy storage SOC out-of-bounds risk at multiple moments in the future; J The smaller the value, the better the system stability, power quality and equipment health.

[0033] Furthermore, the system comprehensive power loss P loss In the calculation formula of energy storage PCS loss, the formula is: ; P loss,PCS Indicates the energy storage PCS loss; I PCS Indicates PCS current, unit A; R PCS Indicates PCS resistance, unit Ω; f SW Indicates the switching frequency, unit Hz; C SW Indicates single switching loss, unit is W / Hz; The calculation formula for photovoltaic inverter loss is: ; Where, Indicates the efficiency of photovoltaic inverter; P PV Indicates photovoltaic output power; The formula for calculating line transmission loss is: ; Where, I line Indicates line, unit A; R line Indicates the equivalent resistance of the line, unit Ω; The calculation formula for the system comprehensive power loss is: P loss = P loss,PCS + P loss,PV + P loss,line ; Where,P loss,PCS Indicates the energy storage PCS loss; P loss,PV Indicates the photovoltaic inverter loss; P loss,line Indicates line transmission loss.

[0034] S104: The system maintains the current state or normal operation mode and does not perform MPC algorithm optimization.

[0035] The MPC algorithm is a model predictive control algorithm that dynamically balances multiple control objectives through a cost function to generate a target control strategy. As a specific example, the core optimization logic of the cost function is as follows: (1) Frequency stability priority: The weight coefficient of frequency deviation is initially set to 0.6. When the system frequency drop rate exceeds the preset threshold (e.g., -0.5 Hz / s), the weight coefficient will dynamically increase to ensure that the frequency recovery speed is better than the traditional method.

[0036] (2) Harmonic distortion suppression: The weight coefficient of the total harmonic distortion (THD) is initially 0.4. When the harmonic content (such as the fifth harmonic) monitored by the bus exceeds 3%, the weight coefficient is automatically increased to give priority to suppressing harmonics.

[0037] (3) Operational efficiency optimization: The weight coefficient of power loss is initially set at 0.3 and is dynamically adjusted according to the energy storage state of charge (SOC). For example, when the energy storage SOC is lower than 20%, the weight is reduced to reduce energy storage device losses.

[0038] (4) Energy storage health protection: The penalty coefficient for energy storage SOC exceeding the limit is initially 0.2. When the energy storage SOC exceeds the limit (e.g., below 20% or above 90%), the coefficient increases dynamically to forcibly protect the energy storage equipment.

[0039] (5) Prediction and feedback logic: Prediction time domain: The decision layer predicts the system status for the next five sampling periods (e.g., a total time domain of 50ms), including frequency deviation, harmonic distortion rate, power loss, and energy storage SOC out-of-bounds degree.

[0040] (6) Discrete time index: Cycle-by-cycle rolling optimization of control parameters to ensure real-time response to high-frequency dynamic scenarios (e.g., second-level fluctuations in photovoltaic power).

[0041] (7) Closed-loop correction: After the execution layer returns the actual parameters (such as the energy storage output power error), the decision layer recalculates the deviation and dynamically adjusts the weight coefficient (such as increasing the frequency weight to improve the response priority) until the system meets the steady-state conditions (trigger function value ≤ 0, frequency deviation < 0.05 Hz, THD ≤ 3%, power balance error < 5 kW).

[0042] As a specific example, the application scenarios of the microgrid control method include parameter adjustment rules when the load changes, parameter adjustment rules when the photovoltaic power generation suddenly changes, and parameter adjustment rules when the energy storage system fails. The details are as follows: First, the parameter adjustment rules when the load changes. The corresponding scenarios include: when the load suddenly increases (such as: high-power equipment switching) or decreases (such as: non-critical load removal), the system may experience power imbalance, resulting in frequency drop or voltage deviation. The corresponding adjustment strategy is: (1) Energy storage PCS: Using MPC algorithm optimization, adjust the dynamic droop rate according to the load priority label K droop Release emergency power to maintain system frequency stability. If the load suddenly increases and the energy storage SOC is high, increase the energy storage discharge power; if the load suddenly decreases and the energy storage SOC is low, reduce the energy storage charging power.

[0043] (2) Photovoltaic inverter: Use MPC algorithm to optimize, according to the frequency change rate and photovoltaic power limiting operation or injection of virtual inertia J virtual , enhancing the system frequency support capability.

[0044] (3) Load controller: Using MPC algorithm optimization, it cuts off non-critical loads according to load priority labels (level 1-level 5) and power gap to ensure power supply to core equipment.

[0045] Secondly, the parameter adjustment rules for sudden changes in photovoltaic power generation include: the photovoltaic output drops suddenly due to cloud cover or sudden weather changes, for example, from 200kW to 50kW, which may cause system power shortage and rapid frequency drop. The corresponding adjustment strategy is: (1) Energy storage PCS: Using MPC algorithm optimization, according to the power change rate Quickly adjust the sag rate K droop , and release emergency power to supplement the power gap.

[0046] (2) Photovoltaic inverter: Use MPC algorithm optimization, dynamically adjust the output power limit, and inject virtual inertia J virtual , slowing down the frequency drop rate.

[0047] (3) Load controller: It uses MPC algorithm for optimization. If the power gap is large, some non-critical loads will be cut off according to priority to maintain power balance.

[0048] Furthermore, the parameter adjustment rules for energy storage system failures correspond to scenarios such as when the energy storage system fails to operate normally due to overheating, low energy storage SOC, or hardware failure, which may cause the system to lose important power regulation means. The adjustment strategy is: (1) Photovoltaic inverter: Using MPC algorithm optimization, dynamically adjust the photovoltaic power limit percentage and virtual inertia according to the remaining new energy output and load demand J virtual .

[0049] (2) Load controller: Using MPC algorithm optimization, according to the load priority label, some non-critical loads are cut off to reduce the system power demand.

[0050] (3) Backup power supply: Use the MPC algorithm to evaluate the power gap and start the backup power supply (such as diesel generator) to fill the gap.

[0051] The technical solution of this application is described below with a specific example: Scenario Background: The microgrid system's initial operating state is as follows: the PV inverter output power is 200 kW, and the DC side voltage is 600 V (the DC voltage before the inverter input). The load power suddenly increases from 100 kW to 250 kW. The busbar PMU monitors a voltage deviation of 6 V (rated voltage 380 V, current voltage 386 V), with a frequency drop rate of 0.5 Hz / s. The energy storage BMS reports an energy storage SOC of 80% (after Kalman filtering), and a temperature change rate of 1.5°C / s.

[0052] 1. Trigger function calculation The decision layer calculates the trigger function value based on the following parameters: load power change rate weight factor 0.7; voltage deviation weight factor 0.5; temperature impact weight factor 0.2; temperature impact coefficient 0.15. The calculation results are: E=0.7×150+0.5×6−0.2×(1+0.15×1.5)=107.755>0 The trigger control action is determined and the MPC algorithm is started to perform multi-objective optimization. The multi-objective optimization process of the MPC algorithm is as follows: (1) Frequency stability optimization: The goal is to control the frequency deviation within ±0.05 Hz. Parameter adjustment: The energy storage PCS dynamic droop rate is calculated to be 0.04975 Hz / kW, releasing 150 kW of emergency power to offset the sudden load increase; the PV inverter injects virtual inertia for 5.1 seconds (baseline 5 seconds + frequency change rate correction), slowing the frequency drop rate to 0.2 Hz / s (originally 0.5 Hz / s).

[0053] (2) Harmonic suppression optimization: The goal is to suppress the total harmonic distortion (THD) to below 3%. Parameter adjustment: The PV inverter virtual impedance model is adjusted (proportional gain 0.4Ω, differential gain 0.2Ω·s), and the 5th harmonic content is reduced from 5.8% to 2.1%.

[0054] (3) Operational efficiency optimization: The goal is to reduce the overall system power loss (including energy storage PCS, PV inverter, and line losses). Parameter adjustment: PV inverter power is limited to 160kW (originally 200kW) to reduce inverter switching losses; the energy storage PCS power loss weight coefficient is dynamically adjusted to 0.3 (originally 0.2) to optimize charging and discharging efficiency.

[0055] (4) Energy storage health protection: The goal is to maintain the energy storage SOC in the safe range of 20%-90%. Parameter adjustment: The energy storage SOC influence coefficient (0.1) is introduced into the dynamic droop rate formula to avoid over-discharge of energy storage (current energy storage SOC = 80%, energy storage SOC drops to 75% after power release).

[0056] 2. Execution layer control actions: (1) Energy storage PCS: Based on the optimized droop rate of 0.04975 Hz / kW, 150 kW of emergency power is released and the system frequency is restored to 50.02 Hz (deviation 0.02 Hz).

[0057] (2) Photovoltaic inverter: limit the power to 160kW, inject virtual inertia for 5.1s, adjust the virtual impedance parameters (0.4Ω + 0.2Ω·s), and reduce the 5th harmonic content from 5.8% to 2.1%.

[0058] (3) Load controller: By cutting off 50kW of non-critical loads at level 5, the system power supply and demand error was reduced from 50kW to 2kW (actual energy storage output 148kW vs. command 150kW).

[0059] (4) Feedback closed loop and dynamic correction: The decision-making layer continuously optimizes through the feedback closed loop mechanism. Deviation analysis: The actual output power of the energy storage system deviates by 2kW from the command, indicating that the load response priority needs to be increased.

[0060] Weight correction: Increase the load power change rate weight factor in the MPC algorithm (from 0.7 to 0.75) to enhance the subsequent control's responsiveness to sudden load changes; adjust the power loss weight coefficient (from 0.3 to 0.25) to balance efficiency and frequency stability.

[0061] In summary, the microgrid control method in the above-mentioned embodiment of the present invention realizes efficient, stable and coordinated control of the microgrid in complex dynamic scenarios by deeply integrating dynamic event triggering with virtual external characteristic adaptation; specifically, the photovoltaic inverter and the load controller are simultaneously involved in data acquisition and control execution to realize integrated perception and control; the photovoltaic penetration rate is obtained according to the photovoltaic output power and the load power, and the photovoltaic penetration rate is dynamically associated with the dynamic droop rate, virtual inertia, energy storage SOC, and temperature change rate for dynamic parameter adaptation; furthermore, the frequency domain impedance is injected by the photovoltaic inverter to solve the harmonic problem of new energy grid connection in a targeted manner; by dynamically adjusting the external characteristic parameters of each device, the coordinated optimization of multiple power sources, multiple loads and energy storage systems is realized to ensure the stable operation of the microgrid.

[0062] Example 2 A second embodiment of the present invention provides a microgrid control system, comprising: An acquisition module is configured to acquire multi-source data of equipment, including a photovoltaic inverter, an energy storage BMS, a load controller, and a bus PMU. The multi-source data includes photovoltaic output power, load power change rate, bus voltage deviation, frequency change rate, energy storage SOC, and temperature change rate, and calculate a trigger function value based on the multi-source data. A judgment module, used to dynamically judge whether the trigger function value is greater than zero based on the multi-physical quantity coupling trigger function to determine whether to trigger a control action; The control module is used to trigger a control action when the trigger function value is greater than zero, optimize the external characteristic parameters and the load shedding instruction through the MPC algorithm, and iteratively correct the external characteristic parameters of the MPC algorithm according to the closed-loop mechanism to generate a target control strategy, so as to collaboratively control the microgrid based on the target control strategy. The external characteristic parameters include a frequency deviation weight factor, a harmonic suppression weight factor, and control parameters. The control parameters include a dynamic droop rate, a virtual impedance, and a virtual inertia.

[0063] In addition, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the method in the above embodiment when the program is executed by a processor.

[0064] In addition, an embodiment of the present invention further provides a data processing device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method in the above embodiment when executing the program.

[0065] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0066] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting, or processing it in another suitable manner as necessary, and then storing it in a computer memory.

[0067] It should be understood that various components of the present invention may be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods may be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof may be used: a discrete logic circuit having logic gate circuits for implementing logic functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc.

[0068] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0069] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the claims and their equivalents.

Claims

1. A microgrid control method, characterized in that: include: Acquire multi-source data of equipment, including photovoltaic inverters, energy storage BMSs, load controllers, and bus PMUs. The multi-source data includes photovoltaic output power, load power change rate, bus voltage deviation, frequency change rate, energy storage SOC, and temperature change rate. Calculate a trigger function value based on the multi-source data. Dynamically determine whether the trigger function value is greater than zero based on the multi-physical quantity coupling trigger function to determine whether to trigger the control action; When the trigger function value is greater than zero, the control action is triggered, and the external characteristic parameters and the load shedding instruction are optimized by the MPC algorithm. The external characteristic parameters of the MPC algorithm are iteratively corrected according to the closed-loop mechanism to generate a target control strategy, so as to collaboratively control the microgrid based on the target control strategy. The external characteristic parameters include the frequency deviation weight factor, the harmonic suppression weight factor and the control parameters. The control parameters include the dynamic droop rate, the virtual impedance and the virtual inertia.

2. The microgrid control method according to claim 1, characterized in that: The expression of the multi-physics coupling trigger function is: ； Where, E Indicates the trigger function value, which is used to determine whether to trigger the control action. When E>0, the control action is triggered; Indicates the load power change rate; Indicates the photovoltaic power change rate; V Indicates voltage deviation; Indicates the temperature change rate of the energy storage battery; α 1, α 2, β , γ are weight coefficients, representing the photovoltaic power weight factor, load power change rate weight factor, voltage deviation weight factor, and dynamic temperature weight factor respectively; Represents the temperature influence coefficient.

3. The microgrid control method according to claim 2, characterized in that: γ The dynamic adjustment formula is: γ new =γ 0﹒(1+ k SOC ﹒(1-SOC / 100)+ k priority ﹒ P load ); Where, γ new Indicates the trigger threshold of the temperature weight factor after dynamic adjustment; γ 0 indicates the baseline trigger threshold; k SOC Indicates the energy storage charge state influence coefficient; energy storage SOC indicates the energy storage charge state; k priority Indicates the load priority influence coefficient; P load Indicates the total load power of the system at the current moment.

4. The microgrid control method according to claim 1, wherein: The calculation formula of dynamic droop rate is: ； K droop Indicates the dynamic droop rate, which characterizes the slope of the power-frequency droop relationship; K 0 represents the base sag rate; k 1 represents the load factor influence coefficient; P load Indicates the total load power of the system at the current moment; P max Indicates the maximum allowable power of the system; k 2 represents the temperature change rate influence coefficient; k 3 represents the energy storage SOC impact coefficient; Indicates the temperature change rate of the energy storage battery.

5. The microgrid control method according to claim 1, characterized in that: The calculation formula of virtual inertia is: ； J virtual represents virtual inertia; J 0 represents the base virtual inertia; k 4 represents the frequency change rate influence coefficient; Indicates the rate of change of frequency; k 5 represents the photovoltaic penetration rate influence coefficient; P renew represents the photovoltaic penetration rate, and the calculation formula is: ； P PV Indicates photovoltaic output power; P demand The total load power of the system is summarized from the real-time data reported by the load controller.

6. The microgrid control method according to claim 1, characterized in that: The calculation formula of virtual impedance is: ； Where Z vir ( s ) represents the virtual impedance frequency domain model; K p represents the proportional gain; K d represents the differential gain; K i represents the integral gain; f c represents the cutoff frequency; s represents a complex frequency variable, s = σ + jω ,in: σ is the real part, representing the attenuation coefficient; ω is the imaginary part, indicating the angular frequency; j represents the imaginary unit, j 2 = −1.

7. The microgrid control method according to claim 1, characterized in that: In the MPC algorithm, multi-objective optimization is achieved through the cost function, which is: ； Where, , represents the set of control input variables that need to be optimized, where K droop represents the dynamic droop rate, J virtual represents the virtual inertia, K p represents the proportional gain, K d represents the differential gain, K i represents the integral gain, L oad_shedding Indicates load switching; represents the frequency deviation weight coefficient; Indicates the THD weight coefficient; Power loss weight coefficient; λ Represents the energy storage SOC penalty coefficient; represents the predicted value of frequency deviation at the next k moments; THD k Represents the predicted value of harmonic distortion rate at time k; k represents the discrete time index in the prediction time domain; SOC penalty represents the energy storage SOC out-of-bounds penalty term; N represents the prediction time domain; P loss Represents the comprehensive power loss of the system, including energy storage PCS loss, PV inverter loss and line transmission loss; J Indicates the comprehensive performance indicators of the system.

8. The microgrid control method according to claim 7, characterized in that: The overall power loss in the system P loss In the calculation formula of energy storage PCS loss, the formula is: ； P loss,PCS Indicates the energy storage PCS loss; I PCS Indicates PCS current; R PCS Indicates PCS resistance; f SW Indicates the switching frequency; C SW Represents single switching loss; The calculation formula for photovoltaic inverter loss is: ； Where, Indicates the efficiency of photovoltaic inverter; P PV Indicates photovoltaic output power; The formula for calculating line transmission loss is: ； Where, I line Indicates a line; R line Indicates the equivalent resistance of the line; The calculation formula for the system comprehensive power loss is: P loss = P loss,PCS + P loss,PV + P loss,line ； Where, P loss,PCS Indicates the energy storage PCS loss; P loss,PV Indicates the photovoltaic inverter loss; P loss,line Indicates line transmission loss.

9. A microgrid control system, characterized in that: The system comprises: An acquisition module is configured to acquire multi-source data of equipment, including a photovoltaic inverter, an energy storage BMS, a load controller, and a bus PMU. The multi-source data includes photovoltaic output power, load power change rate, bus voltage deviation, frequency change rate, energy storage SOC, and temperature change rate, and calculate a trigger function value based on the multi-source data. A judgment module, used to dynamically judge whether the trigger function value is greater than zero based on the multi-physical quantity coupling trigger function to determine whether to trigger a control action; The control module is used to trigger a control action when the trigger function value is greater than zero, optimize the external characteristic parameters and the load shedding instruction through the MPC algorithm, and iteratively correct the external characteristic parameters of the MPC algorithm according to the closed-loop mechanism to generate a target control strategy, so as to collaboratively control the microgrid based on the target control strategy. The external characteristic parameters include a frequency deviation weight factor, a harmonic suppression weight factor, and control parameters. The control parameters include a dynamic droop rate, a virtual impedance, and a virtual inertia.

10. A data processing device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the microgrid control method according to any one of claims 1 to 8 is implemented.

Images