Micro-grid parallel operation and off operation switching method

Abstract

The invention discloses a micro-grid parallel-separation operation switching method, which relates to the technical field of micro-grids and comprises the following steps: S1, collecting real-time state data of a micro-grid, calculating system instability and generating a stability state label; s2, generating an ultra-short-term prediction result, and calculating an energy buffer potential well parameter set and a virtual inertia coefficient in combination with a stability label; s3, triggering a mode switching instruction based on the stability state label, executing the energy buffer potential well parameter set and the virtual inertia coefficient to control the output of the energy storage system, and completing grid-connected and off-grid mode switching; s4, the harmonic content of the power grid is collected, and a protection constant value and reactive power output are dynamically adjusted according to the instability degree of the system; and S5, constructing a system support framework by constructing a hierarchical communication network, deploying a redundancy mechanism and adopting a staged implementation strategy. And through real-time stability evaluation, virtual inertia adjustment and ultra-short-term prediction optimization, the stability, the electric energy quality and the adaptive capacity in the grid-connected and off-grid switching process of the micro-grid are improved, and high reliability and flexibility of the system are ensured.

Description

Technical Field

[0001] This invention relates to the field of microgrid technology, specifically to a method for switching between parallel and off-grid operation of a microgrid. Background Technology

[0002] With the rapid development of renewable energy, microgrids, as a flexible and adaptive power system, are gradually becoming an important component of the power grid. Microgrids can be self-sufficient, meeting load demands through local power generation, and can operate efficiently in both grid-connected and off-grid modes. However, maintaining system stability and power quality during grid-connected / off-grid switching remains a technical challenge.

[0003] Currently, the switching between grid connection and off-grid operation of microgrids mainly relies on the monitoring of electrical parameters (such as voltage and frequency) and traditional control algorithms. Although existing technologies can collect real-time data through monitoring systems and adjust the operating mode based on this data, these methods usually lack sufficient flexibility to cope with complex load and generation fluctuations.

[0004] Furthermore, power quality and protection coordination are also crucial during switching. Existing technologies rely on static setpoint control and simple power adjustment, but they may not be able to adequately guarantee power quality or prevent malfunctions of protection devices when faced with system instability or abnormal fluctuations.

[0005] In terms of communication and system support, although there are some basic redundancy and communication solutions, optimizing communication channels, improving redundancy design, and verifying control models remain challenges as the scale of microgrids expands.

[0006] Therefore, although existing technologies have solved the problem of switching microgrids to and from the grid to some extent, there is still room for optimization, especially in terms of flexible control, power quality assurance, system stability and redundancy mechanisms. Summary of the Invention

[0007] Based on the shortcomings of the prior art described above, the purpose of this invention is to provide a method for switching microgrids between parallel and off-grid operation to solve the aforementioned technical problems.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a microgrid parallel / offline operation switching method, comprising: S1: Collect real-time status data of the microgrid, calculate system instability, and generate system stability status labels; S2: Generate ultra-short-term prediction results based on real-time state data, and calculate the energy buffer potential well parameter set and dynamic virtual inertia coefficient by combining system stability state labels; S3: Based on the system stability status label, the mode switching command is triggered. By executing the energy buffer potential well parameter set and adjusting the dynamic virtual inertia coefficient, the output of the energy storage system is controlled to complete the grid connection and off-grid mode switching. S4: Collects the harmonic content of the power grid and dynamically adjusts the protection settings and reactive power output based on the system instability. S5: Construct the system's support framework by building a hierarchical communication network, deploying multiple redundancy mechanisms, and adopting a phased implementation strategy.

[0009] The present invention is further configured such that S1 includes: Real-time status data of the microgrid is collected using a synchronous phasor measurement unit and conventional sensors. The real-time status data includes voltage, current, frequency, and power. The collected real-time status data is filtered and verified to construct valid data; Based on the filtered effective data, key features including frequency change rate, voltage change rate, and power imbalance are extracted.

[0010] The present invention is further configured to calculate the system instability degree used to assess system stability based on key features using a weighted comprehensive evaluation model and machine learning algorithm; Based on the system instability, multiple threshold judgments are made to generate system stability status labels, which include: stable, warning, critical instability, and on the verge of collapse.

[0011] The present invention is further configured such that S2 includes: Based on historical load data and combined with real-time status data, a time series forecasting algorithm is used to predict the load power and renewable energy generation power within a preset time window, forming an ultra-short-term forecast result. The system stability status label output by S1 is matched with the predefined scenario pattern library to select the corresponding target scenario pattern. The scenario pattern library contains multiple target scenario patterns, and each target scenario pattern contains: pattern ID, stability label, and constraint conditions.

[0012] The present invention is further configured to use the ultra-short-term prediction results as the prediction model input of the model predictive control algorithm; The set of constraints corresponding to the target scenario mode is used as the optimization constraints for the model predictive control algorithm; With the goal of minimizing the frequency and voltage deviation during the microgrid grid-to-grid switching process, rolling optimization calculations of the model predictive control algorithm are performed to solve for the optimal power command sequence in the future time period. The optimal power command sequence is normalized to generate a parameter set for the energy buffer potential well. The parameter set includes: the expected energy gap, the power reference curve, and the effective duration of the power reference curve.

[0013] The present invention is further configured to calculate the difference between the total power generation and the total load power of the microgrid based on the ultra-short-term forecast results, and obtain the net active power of the system. By performing numerical differentiation on the net active power of the system, the changing acceleration is calculated, and based on a preset mapping relationship, the changing acceleration is converted into a dynamic virtual inertia coefficient.

[0014] The present invention is further configured such that S3 includes: Based on the system stability status label, trigger mode switching instructions are used to control the microgrid main circuit breaker to perform grid connection or off-grid operation; The mode switching command specifically includes: when the system stability status label is "critical instability" or "on the verge of collapse", and the electrical parameters of the grid connection point exceed the safe operation limit, a switching command from grid connection to off-grid is triggered; When the system stability status label is "stable" and the voltage difference, frequency difference, and phase difference between the microgrid and the main grid are all less than the preset synchronization threshold, the off-grid to grid switching command is triggered.

[0015] The present invention is further configured to switch the control mode of the main control energy storage converter to the virtual synchronous machine mode simultaneously with receiving the mode switching command; In virtual synchronization machine mode, the preset virtual inertia constant is adjusted in real time according to the dynamic virtual inertia coefficient; The system net active power output of the main control energy storage converter is controlled to track the power reference curve in the parameter set of the energy buffer potential well.

[0016] The present invention is further configured such that S4 includes: Real-time monitoring of power grid harmonic content using harmonic analysis equipment; When the system instability exceeds the preset first quality threshold and the harmonic content exceeds the preset harmonic limit, the adaptive filter is activated and the output of the reactive power compensation device is adjusted. Using system instability as an input parameter, the overcurrent protection setting is calculated and set in real time through a preset adjustment strategy. Based on the numerical range of system instability, the delay time of the protection action is adjusted. When the system instability reaches the preset second quality threshold, a shortened delay is used, and when the system instability is below the preset third quality threshold, the standard delay is restored.

[0017] The present invention is further configured such that S5 includes: Establish a hierarchical communication network and allocate communication channels with different priorities for commands and status monitoring data; Deploy redundancy backup mechanisms in control systems, communication networks, and power supplies; After verifying the core model through a digital simulation platform, a phased strategy was adopted to deploy the microgrid on-grid and off-grid switching system.

[0018] This invention provides a method for switching microgrids between grid and off-grid operation. The method comprises: S1: collecting real-time microgrid status data, calculating system instability, and generating a system stability status label; S2: generating ultra-short-term prediction results based on the real-time status data, and calculating the energy buffer potential well parameter set and dynamic virtual inertia coefficient based on the system stability status label; S3: triggering a mode switching command based on the system stability status label, controlling the energy storage system output by executing the energy buffer potential well parameter set and adjusting the dynamic virtual inertia coefficient to complete the grid-to-off mode switch; S4: collecting grid harmonic content, and dynamically adjusting protection settings and reactive power output based on system instability; S5: constructing a hierarchical communication network, deploying multiple redundancy mechanisms, and adopting a phased implementation strategy to build the system's support framework. The beneficial effects include: Improve system stability: By calculating system instability in real time and combining it with dynamic virtual inertia coefficient for switching control, the dynamic stability of the microgrid during grid connection and disconnection can be effectively guaranteed, avoiding system instability caused by frequency and voltage fluctuations, thereby improving the reliability and security of the microgrid.

[0019] Optimize power quality: By dynamically adjusting reactive power output and optimizing power quality in real time, harmonic interference can be significantly reduced and the power factor improved during the microgrid's on-grid and off-grid switching process, ensuring that the power quality in the grid reaches a high level and reducing the impact on sensitive loads.

[0020] Enhancing the system's flexibility and adaptability: By using an energy buffer potential well based on ultra-short-term prediction and a model predictive control algorithm, this invention can proactively address load fluctuations and power generation instability, enabling the microgrid to adapt more flexibly to different operating states and improving the system's regulation capability and emergency response speed under complex operating conditions.

[0021] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1The flowchart illustrates a microgrid parallel / offline operation switching method as an exemplary embodiment of the present invention. Detailed Implementation

[0023] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0024] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0025] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0026] Example: Microgrid parallel / offline operation switching methods, such as Figure 1 As shown, it includes: S1: Collect real-time status data of the microgrid, calculate system instability, and generate system stability status labels; S2: Generate ultra-short-term prediction results based on real-time state data, and calculate the energy buffer potential well parameter set and dynamic virtual inertia coefficient by combining system stability state labels; S3: Based on the system stability status label, the mode switching command is triggered. By executing the energy buffer potential well parameter set and adjusting the dynamic virtual inertia coefficient, the output of the energy storage system is controlled to complete the grid connection and off-grid mode switching. S4: Collects the harmonic content of the power grid and dynamically adjusts the protection settings and reactive power output based on the system instability. S5: Construct the system's support framework by building a hierarchical communication network, deploying multiple redundancy mechanisms, and adopting a phased implementation strategy.

[0027] The present invention is further configured such that S1 includes: Real-time status data of the microgrid is collected using a synchronous phasor measurement unit and conventional sensors. The real-time status data includes voltage, current, frequency, and power. The collected real-time status data is filtered and verified to construct valid data; Based on the filtered effective data, key features including frequency change rate, voltage change rate, and power imbalance are extracted. Specifically, a hierarchical sensing network consisting of a synchronized phasor measurement unit (PMU) and conventional sensors is deployed to collect real-time status data including voltage, current, frequency, and power. The PMU utilizes GPS clock synchronization technology to accurately capture grid dynamics at a rate exceeding 50 frames per second. Subsequently, the raw data is preprocessed to construct effective data: a Kalman filter algorithm is used to suppress noise by weighted averaging of system physical model predictions and measured values, and data validity verification is performed to remove outliers exceeding limits and correct timing errors. Based on this, the system extracts key dynamic features using numerical differentiation: the frequency change rate is obtained by calculating the quotient of the frequency difference between adjacent sampling points and the time difference. For example, if the frequency at time t0 is 49.80Hz and the frequency at time t1 is 49.82Hz, with a time interval of 0.020 seconds, the frequency change rate is calculated as (49.82Hz-49.80Hz) / 0.020s=1.0Hz / s, representing the frequency instability rate. Similarly, the voltage change rate is obtained by dividing the voltage difference between adjacent sampling points by the time interval, quantifying the voltage fluctuation intensity. The system also aggregates total power generation and total load power in real time and calculates their instantaneous difference to obtain a power imbalance measure that directly reflects the power balance. For example, the difference between the total power generation of 1050 kW and the total load power of 980 kW is +70kW, indicating a power imbalance that directly reflects the system's power balance. Finally, these three key features are combined to form a feature vector, providing a complete quantitative basis for system stability assessment.

[0028] The present invention is further configured to calculate the system instability degree used to assess system stability based on key features using a weighted comprehensive evaluation model and machine learning algorithm; The system performs multi-threshold judgments based on system instability to generate system stability state labels, including: stable, warning, critical instability, and near collapse. Specifically, after obtaining three key features—frequency change rate, voltage change rate, and power imbalance—the system calculates system instability using two methods: weighted comprehensive evaluation and machine learning. Either method can be chosen in actual deployment. First, in the weighted comprehensive evaluation model, the system assigns a weight coefficient to each key feature. The default weight configuration is 50% for frequency change rate, 30% for voltage change rate, and 20% for power imbalance. This weight allocation is based on the professional consensus that system stability is most sensitive to frequency dynamics. During calculation, the system first normalizes the measured values ​​of each key feature to eliminate the influence of dimensions, then multiplies the normalized feature values ​​by their corresponding weight coefficients, and finally sums the three weighted results to obtain a system instability value ranging from 0 to 1. Second, as an alternative, the system can use a lightweight gradient boosting tree machine learning algorithm for calculation. The model is pre-trained using a large amount of historical operating data and simulation cases to learn the complex mapping relationship between key features and system stability. In application, the system takes three key feature values ​​acquired in real time as input to the model, and the machine learning model directly outputs a corresponding estimated system instability. Subsequently, the system performs multi-threshold judgments based on the system instability to generate state labels. The system presets three stability thresholds, with default values ​​of 0.3, 0.5, and 0.7. The rationale for these settings is based on the statistical distribution of a large amount of simulation and measured data, which can divide the system's operating state into four significantly different intervals. The judgment logic is as follows: when the calculated system instability result is less than 0.3, a stable state label is generated; when the result is greater than or equal to 0.3 but less than 0.5, a warning state label is generated; when the result is greater than or equal to 0.5 but less than 0.7, a critical instability state label is generated; and when the result is greater than or equal to 0.7, a near-collapse state label is generated. These state labels are output in string form, providing an intuitive basis for subsequent control decisions regarding the stability level.

[0029] The present invention is further configured such that S2 includes: Based on historical load data and combined with real-time status data, a time series forecasting algorithm is used to predict the load power and renewable energy generation power within a preset time window, forming an ultra-short-term forecast result. The system stability state label output by S1 is matched with a predefined scenario pattern library to select the corresponding target scenario pattern. The scenario pattern library contains multiple target scenario patterns, and each target scenario pattern includes: a pattern ID, a stability label, and constraints. Specifically, after obtaining historical load data and real-time status data, the system generates ultra-short-term prediction results using a time series prediction algorithm. The autoregressive integral moving average model is used as the core prediction tool. This model can effectively capture the short-term variation patterns of load and renewable energy generation. The prediction time window is preset to 200 to 500 milliseconds. The reason for this default value is that this time scale matches the typical response cycle of the microgrid control system, providing necessary foresight for subsequent control decisions. The system independently predicts load power and renewable energy generation power, combining the prediction results to form a complete ultra-short-term prediction dataset. In the scenario pattern matching stage, the system establishes a predefined scenario pattern library, where each pattern contains three elements: a unique pattern identifier, a corresponding expected stability state label, and a set of optimized constraints for the model predictive control algorithm. The mode library contains four basic scenario modes by default: a mode for dealing with motor startup impact, a mode for dealing with sudden drops in photovoltaic power, a mode for dealing with sudden load increases, and a mode for dealing with sudden load decreases. The mode library is built based on a summary of common operational anomalies in microgrids, ensuring coverage of major operational risk types. The matching process involves the system precisely comparing the real-time obtained state labels with the pre-stored stability labels in the mode library. When the two are completely consistent, that mode is selected as the current target scenario mode. For example, if the system stability state label is critical instability, the specific scenario mode marked as critical instability in the mode library is automatically matched and activated. This matching mechanism ensures that the control strategy can adaptively and precisely switch according to the actual stable state of the system.

[0030] The present invention is further configured to use the ultra-short-term prediction results as the prediction model input of the model predictive control algorithm; The set of constraints corresponding to the target scenario mode is used as the optimization constraints for the model predictive control algorithm; With the goal of minimizing the frequency and voltage deviation during the microgrid grid-to-grid switching process, rolling optimization calculations of the model predictive control algorithm are performed to solve for the optimal power command sequence in the future time period. The optimal power command sequence is normalized to generate a parameter set for the energy buffer potential well. This parameter set includes the expected energy gap, the power reference curve, and the effective duration of the power reference curve. Specifically, firstly, the ultra-short-term forecast results, including the load power and renewable energy generation power over the next 300 milliseconds, are used as the input to the model predictive control algorithm. Simultaneously, the corresponding constraint set is extracted from the matched target scenario mode. For example, the corresponding constraint set is: maximum charging and discharging power of the energy storage converter ±200 kW, maximum power change rate 200 kW / s, and allowable voltage fluctuation range ±10%. This constraint set is used as the boundary conditions for the optimization calculation. Subsequently, the optimization objective is to minimize the sum of the squares of the frequency deviation and voltage deviation, where the frequency deviation and voltage deviation are assigned the same weight coefficient of 0.5 in the objective function. This default weight setting is to fully balance the importance of frequency stability and voltage stability during the switching process. The system performs rolling optimization calculations with a control cycle of 20 milliseconds. At each control moment, the algorithm, based on the current system state and the prediction model, solves for an optimal power command sequence covering a 300-millisecond time window. The optimal power command sequence consists of fifteen consecutive power values ​​with 2-millisecond intervals. The first power command is issued and executed immediately, and then a complete optimization calculation is performed again in the next control cycle based on the latest system state. The optimal power command sequence obtained from the rolling optimization is normalized to generate an energy buffer potential well parameter set: the expected energy gap is obtained by numerically integrating the optimal power command sequence within the 300-millisecond time window, representing the energy deficit that the system needs to supplement. For example, when the power sequence is mainly composed of positive values, the integral result is 15 kJ, which is the expected energy gap. The power command sequence is converted into a continuous power reference curve through linear interpolation. The entire optimization time domain of 300 milliseconds is directly defined as the effective duration of the power reference curve. The reason for setting this default duration is to keep it consistent with the prediction time domain of model predictive control, ensuring the integrity and consistency of control commands in the time dimension. The final result is a parameter set containing three elements: expected energy gap, power reference curve, and effective duration, providing a complete and standardized data foundation for subsequent execution control.

[0031] The present invention is further configured to calculate the difference between the total power generation and the total load power of the microgrid based on the ultra-short-term forecast results, and obtain the net active power of the system. The system's net active power is numerically differentiated to calculate the acceleration due to change. Based on a pre-defined mapping relationship, this acceleration is converted into a dynamic virtual inertia coefficient. Specifically, firstly, based on the total power generation and total load power data for the next 200 milliseconds from the ultra-short-term forecast results, the system's net active power is calculated in real time by subtracting the total power generation from the total load power at the same time point. Then, the system's net active power sequence is numerically differentiated to calculate the acceleration due to change: specifically, the central difference method is used, dividing the difference in net active power between two adjacent sampling points by the square of the time interval to obtain an acceleration measure characterizing the rate of power change, measured in kilowatts per square second. After obtaining the acceleration due to change, it is converted into a dynamic virtual inertia coefficient according to a pre-defined linear piecewise mapping relationship. The mapping relationship requires setting two key thresholds: when the absolute value of the change in acceleration is below 500 kW / s², it is judged as a minor disturbance, corresponding to an inertia coefficient of 1.0; when the absolute value of the change in acceleration exceeds 1500 kW / s², it is judged as a severe disturbance, corresponding to an inertia coefficient of 3.0; within this range, the mapping follows a linear relationship. The default values ​​for these two thresholds are based on statistical analysis of the disturbance intensity of typical microgrids. 500 kW / s² corresponds to the power change rate caused by conventional load switching, and 1500 kW / s² corresponds to the power change rate of severe faults such as distributed generation disconnection. The rationale for setting the inertia coefficient range is to ensure sufficient inertia support for the system through an upper limit of 3.0, while avoiding excessive inertia affecting the system's dynamic response performance through a lower limit of 1.0. A specific calculation example is as follows: when the measured change in acceleration is +800 kW / s², the dynamic virtual inertia coefficient calculated according to the linear mapping relationship is 2.0. The dynamic virtual inertia coefficient will be directly used to adjust the inertia constant of the virtual synchronizer. Specifically, a reference virtual inertia constant with a default value of 1000 g / m² is set, and the dynamic virtual inertia coefficient is used as the real-time adjustment multiplier of the reference inertia constant. It is directly applied to the reference value through scalar multiplication. The calculation formula is: Real-time virtual inertia constant = Reference virtual inertia constant * Dynamic virtual inertia coefficient. This achieves the adaptive control objective of enhancing system inertia when power changes drastically and reducing inertia when power is stable.

[0032] The present invention is further configured such that S3 includes: Based on the system stability status label, trigger mode switching instructions are used to control the microgrid main circuit breaker to perform grid connection or off-grid operation; The mode switching command specifically includes: when the system stability status label is "critical instability" or "on the verge of collapse", and the electrical parameters of the grid connection point exceed the safe operation limit, a switching command from grid connection to off-grid is triggered; When the system stability status label is "stable," and the voltage difference, frequency difference, and phase difference between the microgrid and the main grid are all less than the preset synchronization thresholds, a switching command from off-grid to grid-connected is triggered. Specifically, the system continuously monitors the system stability status label and real-time electrical parameters, and automatically triggers a mode switching command when specific conditions are met. Specifically, when the system stability status label is determined to be critically unstable or on the verge of collapse, and the grid-connected point voltage exceeds the rated value by ±10% for 20 milliseconds, or the frequency exceeds the rated value by ±0.5 Hz for 20 milliseconds, the system immediately triggers a switching command from grid-connected to off-grid. The default values ​​for safe operating limits of voltage and frequency are set according to the basic requirements of the power system safe operation regulations. The default value setting of 20 milliseconds is to effectively distinguish between transient disturbances and continuous faults, avoiding malfunctions. When the system stability status label remains stable, and the following three conditions are met simultaneously for 100 milliseconds, the system triggers a switching command from off-grid to grid-connected: the voltage amplitude difference between the microgrid side and the main grid voltage is less than 2%, the frequency difference is less than 0.1 Hz, and the phase difference is less than 2 degrees. The default values ​​for these synchronization thresholds are set with reference to power system synchronization and grid connection technical standards to ensure that the closing inrush current is within a safe range. The default value of 100 milliseconds for the duration is to ensure the stability of the synchronization state and prevent misjudgments during transient processes.

[0033] The present invention is further configured to switch the control mode of the main control energy storage converter to the virtual synchronous machine mode simultaneously with receiving the mode switching command; In virtual synchronization machine mode, the preset virtual inertia constant is adjusted in real time according to the dynamic virtual inertia coefficient; The system controls the net active power output of the main control energy storage converter to track the power reference curve in the parameter set of the energy buffer potential well. Specifically, within one control cycle after receiving the mode switching command, the main control energy storage converter switches the control mode from constant power mode to virtual synchronous machine mode. The default time is set to 2 milliseconds to ensure that the control mode switching speed matches the response capability of the power electronic devices. In virtual synchronous machine mode, the system adjusts the preset virtual inertia constant in real time according to the dynamic virtual inertia coefficient. Specifically, a reference virtual inertia constant with a default value of 1000 g / m² is set. The adjustment method uses scalar multiplication, with the dynamic virtual inertia coefficient as the real-time adjustment multiplier of the reference inertia constant. The calculation formula is: Real-time virtual inertia constant = Reference virtual inertia constant * Dynamic virtual inertia coefficient. Simultaneously, the system controls the net active power output of the main control energy storage converter to accurately track the power reference curve in the parameter set of the energy buffer potential well. The tracking control is implemented using a proportional-integral (PI) controller. The default value for the proportional gain is 5, and the default value for the integral gain is 100. These default values ​​are based on the tuning results of the dynamic characteristics of a typical energy storage converter. The specific execution process is as follows: In each control cycle, the system reads the power reference value corresponding to the current moment from the energy buffer potential well parameter set. The PPI controller then calculates the modulation signal of the converter, ultimately achieving closed-loop tracking of the power reference curve. The tracking error is controlled within 2% of the rated power. This default error range balances the requirements of control accuracy and system stability.

[0034] The present invention is further configured such that S4 includes: Real-time monitoring of power grid harmonic content using harmonic analysis equipment; When the system instability exceeds the preset first quality threshold and the harmonic content exceeds the preset harmonic limit, the adaptive filter is activated and the output of the reactive power compensation device is adjusted. Using system instability as an input parameter, the overcurrent protection setting is calculated and set in real time through a preset adjustment strategy. Based on the numerical range of system instability, the delay time of the protection action is adjusted. A shortened delay is used when the system instability reaches a preset second quality threshold, and the standard delay is restored when the system instability falls below a preset third quality threshold. Specifically, the system monitors the harmonic content of the power grid in real time using harmonic analysis equipment, covering characteristic harmonics from the 2nd to the 25th order. When the system instability exceeds a preset first quality threshold and the harmonic content exceeds a preset harmonic limit, the system immediately activates the adaptive filter and adjusts the output of the reactive power compensation device. The default value of the first quality threshold is 0.5, corresponding to the critical point where the system begins to show significant stability risks. The harmonic limit is set to 5% by default, conforming to the basic requirements of international power quality standards for total harmonic distortion (THD). The adaptive filter uses a least mean square algorithm to update the filter coefficients in real time, while the reactive power compensation device adjusts its output proportionally according to the current system reactive power deficit. The default adjustment ratio is 1.2, a setting that balances compensation speed and system stability. The system uses real-time collected system instability as input parameters and calculates and sets the overcurrent protection setting in real time through a preset linear adjustment strategy. The adjustment strategy is defined as follows: the overcurrent protection setting equals the baseline protection setting multiplied by 1 minus 0.3 multiplied by the system instability. The baseline protection setting is set to 1.5 times the rated current by default, ensuring sufficient overload capacity under normal operating conditions. The default value of 0.3 allows the protection setting to decrease smoothly as system instability increases; when system instability reaches its maximum value of 1, the protection setting drops to 70% of the baseline value. Based on the numerical range of system instability, the system dynamically adjusts the protection action delay time. When system instability reaches the preset second quality threshold of 0.6, the protection action delay automatically switches to a shortened delay of 50 milliseconds; when system instability falls below the preset third quality threshold of 0.3, the protection action delay reverts to the standard delay of 200 milliseconds. The default value of the second quality threshold of 0.6 is used to ensure that the corresponding system enters an emergency state requiring rapid fault isolation; the default value of the third quality threshold of 0.3 is used to characterize that the system has recovered to a stable state that can tolerate normal delays; the delay settings of 50 milliseconds and 200 milliseconds are based on engineering practices of the operating characteristics of the protection device and the fault tolerance of the system. All parameter updates and delay adjustments are performed with a cycle of 10 milliseconds. The 10-millisecond execution cycle setting can respond to changes in system status in a timely manner, while avoiding excessively frequent adjustments to the protection settings.

[0035] The present invention is further configured such that S5 includes: Establish a hierarchical communication network and allocate communication channels with different priorities for commands and status monitoring data; Deploy redundancy backup mechanisms in control systems, communication networks, and power supplies; After validating the core model through a digital simulation platform, a phased strategy is adopted to deploy the microgrid's on-grid / off-grid switching system. Specifically, a hierarchical communication network architecture is established by allocating differentiated communication channels for different types of transmission data. Real-time control commands and synchronization phasor data are allocated to high-speed communication channels using industrial Ethernet technology, with a default transmission latency requirement of less than 10 milliseconds. Status monitoring data is allocated to ordinary communication channels using fiber optic ring network technology, with a default transmission latency requirement of less than 100 milliseconds. At the control system level, a primary and backup controller hot redundancy mechanism is deployed. The backup controller maintains data synchronization with the primary controller, and the switching time is set to 50 milliseconds by default to ensure that the control system can recover quickly in the event of a fault without affecting continuous system operation. At the network level, a dual-ring network topology is constructed. When any communication link is interrupted, the system is configured to complete path switching within 20 milliseconds to ensure the continuous reliability of network communication. At the power supply level, an uninterruptible power supply (UPS) system is configured, with the backup power supply duration set to 30 minutes by default to ensure that the system has sufficient time to safely shut down or start backup power generation equipment. The system instability calculation model and energy buffer potential well control strategy were verified using a digital simulation platform. The simulation step size was set to 1 millisecond by default to ensure that the system could accurately simulate the rapid dynamic characteristics of power electronic devices. A phased deployment strategy was adopted for the entire system. The first phase deployed basic data acquisition and rule-based switching control, with a default implementation period of 30 days. The second phase deployed intelligent prediction and adaptive control functions, with a default implementation period of 15 days. These default periods were set to fully consider the system complexity and the actual needs of on-site debugging, ensuring the stability and reliability of each stage. After each stage was deployed, it underwent 720 hours of continuous trial operation. The default duration was set to ensure the stability and reliability of the system under different operating conditions.

[0036] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for switching between parallel and disconnected operation of a microgrid, characterized in that, include: S1: Collect real-time status data of the microgrid, calculate system instability, and generate system stability status labels; S2: Generate ultra-short-term prediction results based on real-time state data, and calculate the energy buffer potential well parameter set and dynamic virtual inertia coefficient by combining system stability state labels; S3: Based on the system stability status label, the mode switching command is triggered. By executing the energy buffer potential well parameter set and adjusting the dynamic virtual inertia coefficient, the output of the energy storage system is controlled to complete the grid connection and off-grid mode switching. S4: Collects the harmonic content of the power grid and dynamically adjusts the protection settings and reactive power output based on the system instability. S5: Construct the system's support framework by building a hierarchical communication network, deploying multiple redundancy mechanisms, and adopting a phased implementation strategy.

2. The microgrid parallel / offline operation switching method according to claim 1, characterized in that, S1 includes: Real-time status data of the microgrid is collected using a synchronous phasor measurement unit and conventional sensors. The real-time status data includes voltage, current, frequency, and power. The collected real-time status data is filtered and verified to construct valid data; Based on the filtered effective data, key features including frequency change rate, voltage change rate, and power imbalance are extracted.

3. The microgrid parallel / offline operation switching method according to claim 2, characterized in that, Based on key features, a weighted comprehensive evaluation model and machine learning algorithm are used to calculate the system instability degree used to assess system stability. Based on the system instability, multiple threshold judgments are made to generate system stability status labels, which include: stable, warning, critical instability, and on the verge of collapse.

4. The microgrid parallel / offline operation switching method according to claim 1, characterized in that, S2 includes: Based on historical load data and combined with real-time status data, a time series forecasting algorithm is used to predict the load power and renewable energy generation power within a preset time window, forming an ultra-short-term forecast result. The system stability status label output by S1 is matched with the predefined scenario pattern library to select the corresponding target scenario pattern. The scenario pattern library contains multiple target scenario patterns, and each target scenario pattern contains: pattern ID, stability label, and constraint conditions.

5. The microgrid parallel / offline operation switching method according to claim 4, characterized in that, The ultra-short-term forecast results are used as the prediction model input for the model predictive control algorithm; The set of constraints corresponding to the target scenario mode is used as the optimization constraints for the model predictive control algorithm; With the goal of minimizing the frequency and voltage deviation during the microgrid grid-connected / off-grid switching process, rolling optimization calculations of the model predictive control algorithm are performed to solve for the optimal power command sequence in the future time period. The optimal power command sequence is normalized to generate a parameter set for the energy buffer potential well. The parameter set includes: the expected energy gap, the power reference curve, and the effective duration of the power reference curve.

6. The microgrid parallel / offline operation switching method according to claim 5, characterized in that, Based on the ultra-short-term forecast results, the difference between the total power generation and the total load power of the microgrid is calculated to obtain the net active power of the system. By performing numerical differentiation on the net active power of the system, the changing acceleration is calculated, and based on a preset mapping relationship, the changing acceleration is converted into a dynamic virtual inertia coefficient.

7. The microgrid parallel / offline operation switching method according to claim 1, characterized in that, S3 includes: Based on the system stability status label, trigger mode switching instructions are used to control the microgrid main circuit breaker to perform grid connection or off-grid operation; The mode switching command specifically includes: when the system stability status label is "critical instability" or "on the verge of collapse" and the electrical parameters of the grid connection point exceed the safe operation limit, a switching command from grid connection to off-grid is triggered; When the system stability status label is "stable" and the voltage difference, frequency difference, and phase difference between the microgrid and the main grid are all less than the preset synchronization threshold, the off-grid to grid switching command is triggered.

8. The microgrid parallel / offline operation switching method according to claim 7, characterized in that, Upon receiving the mode switching command, the control mode of the main control energy storage converter is switched to virtual synchronous machine mode; In virtual synchronization machine mode, the preset virtual inertia constant is adjusted in real time according to the dynamic virtual inertia coefficient; The system net active power output of the main control energy storage converter is controlled to track the power reference curve in the parameter set of the energy buffer potential well.

9. The microgrid parallel / offline operation switching method according to claim 1, characterized in that, S4 includes: Real-time monitoring of power grid harmonic content using harmonic analysis equipment; When the system instability exceeds the preset first quality threshold and the harmonic content exceeds the preset harmonic limit, the adaptive filter is activated and the output of the reactive power compensation device is adjusted. Using system instability as an input parameter, the overcurrent protection setting is calculated and set in real time through a preset adjustment strategy. Based on the numerical range of system instability, the delay time of the protection action is adjusted. When the system instability reaches the preset second quality threshold, a shortened delay is used, and when the system instability is below the preset third quality threshold, the standard delay is restored.

10. The microgrid parallel / offline operation switching method according to claim 1, characterized in that, S5 includes: Establish a hierarchical communication network and allocate communication channels with different priorities for commands and status monitoring data; Deploy redundancy backup mechanisms in control systems, communication networks, and power supplies; After verifying the core model through a digital simulation platform, a phased strategy was adopted to deploy the microgrid on-grid and off-grid switching system.

Images