Microgrid adaptive switching control method and system based on multi-source information fusion

Abstract

The invention relates to the technical field of micro-grid control, and discloses a micro-grid adaptive switching control method and system based on multi-source information fusion, and the method comprises the steps: obtaining the internal operation parameters of a micro-grid, and the state parameters of an external large power grid connected with the micro-grid; the internal operation parameters and the external large power grid state parameters are fused, and a real-time comprehensive operation state model of the micro-grid is constructed; based on the real-time comprehensive operation state model, adaptively determining a target operation mode of the micro-grid; and according to the target operation mode, generating and executing a smooth switching control instruction of the micro-grid. According to the method, the multi-source operation information of the large power grid inside and outside the microgrid is fused, the comprehensive operation state model is constructed in real time, the target operation mode is adaptively selected, and switching control is smoothly executed, so that efficient and stable switching of the microgrid between multiple modes such as grid connection and island can be realized, and the power supply reliability, the system stability and the electric energy quality are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of micro-grid control, in particular to a micro-grid adaptive switching control method and system based on multi-source information fusion. BACKGROUND

[0002] With the large-scale access of distributed power sources and energy storage devices, micro-grids have been widely concerned as an important form of realizing distributed energy consumption and improving power supply reliability. In actual operation, micro-grids can either run independently in island mode to meet local load power supply demand or run in parallel with the main grid to realize bidirectional energy flow and optimal scheduling. However, due to the complex internal structure of micro-grids and the strong dynamic nature of the operating environment, how to realize smooth and efficient switching between different operating modes becomes a key problem.

[0003] Existing micro-grid operating mode switching control methods usually make decisions based on a single parameter or limited monitoring signals, such as determining grid-connected or islanded state only according to voltage or frequency deviation. Such methods have obvious limitations when facing complex operating scenarios: first, they cannot comprehensively reflect the dynamic changes of distributed power output, energy storage device state and load demand within the micro-grid; second, they lack effective perception of external main grid voltage, frequency, power flow and fault information, resulting in inaccurate operating mode selection and even causing power quality degradation or power supply interruption during the switching process.

[0004] Therefore, there is an urgent need for a micro-grid adaptive switching control method that can integrate multi-source information, comprehensively model and analyze internal micro-grid operating parameters and external main grid state parameters, and then realize intelligent identification and adaptive switching of operating modes. By improving the comprehensiveness of information utilization and the intelligence level of control strategies, the stability and reliability of micro-grid switching between different operating modes can be effectively improved, meeting the needs of new power systems for safe, flexible and efficient operation. SUMMARY

[0005] In view of this, the present application proposes a micro-grid adaptive switching control method and system based on multi-source information fusion, aiming to solve the problem in the prior art that existing micro-grid operating mode switching methods rely on single or limited parameters, lack comprehensive perception and fusion of internal and external operating states, and result in inaccurate mode selection and unstable switching process.

[0006] The present application proposes a micro-grid adaptive switching control method based on multi-source information fusion, comprising: obtaining internal operation parameters of the micro-grid and state parameters of an external large grid connected with the micro-grid, wherein the internal operation parameters include distributed power output, load demand, energy storage device state of charge, and micro-grid internal node voltage and current; and the state parameters of the external large grid include large grid frequency, voltage, power flow, and fault signal; fusing the internal operation parameters and the state parameters of the external large grid to construct a real-time comprehensive operation state model of the micro-grid; based on the real-time comprehensive operation state model, adaptively determining a target operation mode of the micro-grid, wherein the target operation mode includes a grid-connected mode, an islanded mode, a grid recovery mode, or an internal reconstruction mode; generating and executing smooth switching control instructions of the micro-grid according to the target operation mode.

[0007] Further, when obtaining the internal operation parameters of the micro-grid and the state parameters of the external large grid connected with the micro-grid, the following steps are included: based on a distributed monitoring unit, real-time output data of various types of distributed power sources in the micro-grid are collected in real time, the real-time output data including direct current side power and alternating current side power of a photovoltaic array, and output power and impeller speed of a wind turbine generator set; instantaneous active power, reactive power, current effective value, and voltage effective value of each key load point of the micro-grid are obtained, and short-term load prediction based on historical trends and weather conditions is performed on the load data; the state of charge, battery pack voltage, charge and discharge current, battery health status, and cycle life parameters of the energy storage device are monitored in real time; voltage amplitude, frequency, phase angle of key busbars inside the micro-grid, and active and reactive power flow information of each tie line are collected; grid frequency, point of common coupling voltage deviation, large grid fault protection signal, and tie line power transmission state of the external large grid are obtained, and the fault protection signal includes overcurrent, undervoltage, or frequency out-of-limit alarm signal.

[0008] Further, when fusing the internal operation parameters of the micro-grid and the state parameters of the external large grid to construct the real-time comprehensive operation state model of the micro-grid, the following steps are included: standardized preprocessing is performed on the obtained multi-source operation information, including abnormal data elimination, missing data interpolation and filling, and multi-source data timestamp alignment, to construct a time sequence consistent micro-grid operation data set; feature parameters are extracted from the micro-grid operation data set, including system active power balance deviation, reactive power margin, voltage deviation rate, and frequency deviation rate; the feature parameters are input into a feature fusion network model based on deep learning to generate a high-dimensional comprehensive state feature vector; Based on the high-dimensional comprehensive state feature vector, combined with the preset micro-grid operation constraint condition and optimization target, the power quality state, power supply reliability level, economic operation efficiency and system transient stability index of the micro-grid are evaluated in real time through fuzzy logic reasoning or expert system rules; According to the power quality state, power supply reliability level, economic operation efficiency and system transient stability index of the micro-grid, a real-time comprehensive operation state model of the micro-grid is constructed.

[0009] Further, the obtained multi-source operation information is standardized and preprocessed, including abnormal data elimination, missing data interpolation and filling, and multi-source data timestamp alignment, to construct a time-consistent micro-grid operation data set, including: For the real-time collected distributed power output data, load demand data, energy storage device operation data and grid state data, a statistical-based outlier detection algorithm is implemented to identify and eliminate outlier data points that exceed the historical data range or physical limits; For data with missing time points, a cubic spline interpolation or an autoregressive model based on historical data patterns is used for accurate filling to ensure the continuity of the data set; Through a time synchronization protocol, asynchronous data from different monitoring devices is timestamp calibrated and aligned to eliminate time deviation caused by data transmission delay or asynchronous collection; All preprocessed original data is normalized to map feature values of different dimensions to a unified numerical interval, and is used to support the input of the deep learning model; A time series data set containing instantaneous power, voltage, frequency, current, and corresponding fluctuation rate and deviation rate is constructed, and the data set is structured as a multi-dimensional tensor.

[0010] Further, based on the real-time comprehensive operation state model, the target operation mode of the micro-grid is adaptively determined, including: According to the real-time comprehensive operation state model, the fault state indication of the external power grid, the power balance deviation of the micro-grid, the state of charge of the energy storage device, and the priority information of the key load are extracted; The fault state indication is compared with the preset fault threshold to determine the fault Boolean marker of the external grid; The power balance deviation is compared with the preset stable operation interval to determine the internal stability Boolean marker of the micro-grid; The state of charge of the energy storage device is compared with the preset minimum state of charge threshold to determine the energy storage availability Boolean marker; The priority information of the key load is compared with the preset power supply guarantee level to determine the load importance Boolean marker; Based on the configured multi-mode decision rule set, the Boolean markers are logically aggregated; According to the logical aggregation result, the target operation mode of the micro-grid is adaptively determined, wherein: If the external large power grid is in failure and the micro-grid is internally stable, the emergency island mode is determined; If the external large power grid is restored and the micro-grid is internally smooth, the restored grid-connected mode is determined; If the distributed power supply output in the micro-grid is excessive, the optimized grid-connected mode is determined; If the distributed power supply output in the micro-grid is insufficient and the energy storage availability is low, the priority grid-connected mode is determined.

[0011] Further, when adaptively determining the target operation mode of the micro-grid according to the logical aggregation result, the following is included: If the external power grid failure Boolean flag is true and the micro-grid internal stability Boolean flag is true, the micro-grid is determined to enter the emergency island mode; If the external power grid failure Boolean flag is false and the micro-grid internal stability Boolean flag is true, and the energy storage availability Boolean flag is true, the micro-grid is determined to enter the optimized grid-connected mode; If the external power grid failure Boolean flag is false and the micro-grid internal stability Boolean flag is true, but the energy storage availability Boolean flag is false, and the distributed power supply output in the micro-grid is insufficient to meet the load demand, the micro-grid is determined to enter the priority grid-connected mode; If the external power grid failure Boolean flag is false and the micro-grid internal stability Boolean flag is true, and the distributed power supply output in the micro-grid is excessive, the micro-grid is determined to enter the grid-fed mode; If the external power grid failure Boolean flag is false and the micro-grid internal stability Boolean flag is true, and the distributed power supply output in the micro-grid meets the load demand, but the energy storage availability Boolean flag is false, the micro-grid is determined to enter the independent operation mode.

[0012] Further, when configuring the multi-mode decision rule set, the following is included: Key operation index threshold values in different operation modes are obtained, including the grid voltage and frequency stability interval in the grid-connected mode, the minimum power balance margin in the island mode, the charge and discharge safety limit of the energy storage device, and the rated output range of the distributed power supply; According to the key operation index threshold values, operation condition descriptions of the micro-grid under various external power grid states, internal power generation capacity and load demand changes are established, and the condition is defined as a series of logical judgment rules; For each condition, a corresponding target operation mode and mode switching priority are preset; In combination with the evaluation results in the real-time comprehensive operation state model, a mode switching decision matrix based on a multi-objective optimization algorithm is constructed, which considers multiple factors such as power supply reliability, operation economy, power quality and system stability, and assigns weights to each factor; Based on the mode switching decision matrix, a dynamically adjusted fuzzy logic rule base or a strategy table based on reinforcement learning is established, which can adaptively adjust the triggering conditions and execution strategies of mode switching according to the subtle changes in real-time operation state.

[0013] Further, when generating and executing the smooth switching control instructions of the microgrid, the following steps are included: For the target operation mode, the operation parameters of the distributed power sources and energy storage devices within the microgrid are dynamically adjusted to meet the power balance requirements before mode switching; Before executing the mode switching instructions, the voltage, frequency and phase angle of the key nodes of the microgrid are monitored in real time to ensure that they are within the preset synchronization tolerance range; When it is determined to switch from the grid-connected mode to the islanded mode, a tripping instruction is sent to the grid-connected circuit breaker connected to the external large power grid, and the frequency and voltage droop control of the main control inverter within the microgrid is started to maintain the voltage and frequency stability within the microgrid; When it is determined to switch from the islanded mode to the grid-connected mode, accurate grid synchronization detection is performed, including voltage amplitude, frequency and phase angle matching detection; after the synchronization conditions are met, a closing instruction is sent to the grid-connected circuit breaker, and the droop control within the microgrid is gradually removed, and the control of the load and distributed power sources is smoothly transferred to the large power grid; During the entire switching process, the transient impact within the microgrid is monitored in real time, and the fast-responding energy storage device or reactive power compensation equipment is used for dynamic suppression to minimize voltage drop, frequency fluctuation and current impact.

[0014] Further, during the entire switching process, when the transient impact within the microgrid is monitored in real time and the fast-responding energy storage device or reactive power compensation equipment is used for dynamic suppression, the following steps are included: The voltage, current and frequency data of the microgrid public connection point and key load bus are collected in real time, and high-frequency sampling technology is used to capture transient transient change waveforms; The transient waveforms are analyzed by using Fast Fourier Transform (FFT) or wavelet analysis method to identify transient fault characteristics such as voltage drop, frequency oscillation or instantaneous power impact, and calculate their severity indexes; When the severity index of the transient fault characteristics exceeds the preset threshold, the fast-response mechanism is triggered; For voltage sag or overvoltage, the power converter of the energy storage device quickly adjusts the output voltage amplitude and phase angle, or the dynamic reactive power compensation device instantaneously injects or absorbs reactive power, to quickly restore voltage stability; For frequency oscillation or frequency limit, the energy storage device is controlled to charge and discharge quickly to provide inertia support or power balance, thereby effectively suppressing frequency fluctuations; at the same time, the output power of the controllable distributed power supply is coordinated and controlled to participate in primary and secondary frequency regulation; The dynamic suppression strategy includes an adaptive control algorithm based on model predictive control (MPC), which adjusts the compensation amount and compensation timing according to real-time transient response.

[0015] Compared with the prior art, the beneficial effects of the present application are that by fusing the internal operating parameters of the microgrid and the external grid state parameters, a real-time comprehensive operating state model is constructed, which can more comprehensively and accurately reflect the overall operating condition of the microgrid compared with the traditional method which relies only on a single monitoring signal. This multi-source information fusion mechanism avoids mode determination errors caused by incomplete information, thereby improving the accuracy of switching decisions. Secondly, based on the comprehensive operating state model, the target operating mode is adaptively determined, so that the microgrid can dynamically select the optimal mode according to different operating environments and working conditions, realizing the intelligence and flexibility of mode switching. For example, in the event of a grid fault or abnormal condition, the system can timely identify and switch to island mode to ensure continuous power supply for local loads; when the external grid is restored to stability, it can smoothly realize grid connection or reconstruction operation, thereby enhancing the continuity and reliability of power supply. Finally, by generating and executing smooth switching control instructions, the present application can effectively reduce voltage fluctuations, frequency deviations and power surges during the switching process, avoiding the decline in power quality and the risk of equipment operation caused by mode switching. Overall, this method not only improves the stability and safety of microgrid operation, but also enhances its adaptability and robustness in collaborative operation with the main grid, and has good popularization and application value.

[0016] On the other hand, the present application also provides a microgrid adaptive switching control system based on multi-source information fusion, comprising: The information acquisition module is configured to acquire the internal operating parameters of the microgrid and the state parameters of the external grid in real time, wherein the internal operating parameters include distributed power output, load demand, energy storage device state of charge, microgrid internal node voltage and current; the state parameters of the external grid include grid frequency, voltage, power flow and fault signals; The data evaluation module is electrically connected with the information acquisition module and is configured to receive multi-source operating information and perform data preprocessing on the information, including abnormal data rejection, missing data interpolation and time stamp alignment; the data evaluation module is further configured to fuse the preprocessed multi-source operating information to construct a real-time comprehensive operating state model of the microgrid; The mode decision module is electrically connected with the data evaluation module and is configured to adaptively determine a target operation mode of the micro-grid based on the real-time comprehensive operation state model, wherein the target operation mode includes a grid-connected mode, an islanded mode, a grid recovery mode or an internal reconfiguration mode. The switching control module is electrically connected with the mode decision module and is configured to generate and execute a smooth switching control instruction of the micro-grid according to the target operation mode, wherein the instruction includes grid-connected breaker opening and closing control, distributed power output adjustment, energy storage device charging and discharging strategy and voltage frequency droop control; the switching control module is further configured to monitor system transient state in real time and perform effective dynamic suppression during the switching process.

[0017] It can be understood that the micro-grid adaptive switching control method and system based on multi-source information fusion in each of the above embodiments of the present application have the same beneficial effects, which will not be repeated. BRIEF DESCRIPTION OF DRAWINGS

[0018] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of preferred embodiments, and are not meant to limit the present application. Furthermore, the same reference numerals are used throughout the several drawings to designate the same or similar parts. In the drawings: Figure 1 A flow chart of the micro-grid adaptive switching control method based on multi-source information fusion provided by the embodiments of the present application is shown in the figure; Figure 2 A flow chart of the micro-grid adaptive switching control method based on multi-source information fusion provided by the embodiments of the present application is shown in the figure; Figure 3 A functional block diagram of the micro-grid adaptive switching control system based on multi-source information fusion provided by the embodiments of the present application is shown in the figure. DETAILED DESCRIPTION

[0019] Exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.

[0020] As Figure 1 - Figure 2As shown, in some embodiments of the present application, the present embodiment provides a micro-grid adaptive switching control method based on multi-source information fusion, comprising: Step S100, obtaining internal operation parameters of the micro-grid and state parameters of the external large grid connected by the micro-grid.

[0021] Specifically, the internal operation parameters include distributed power output, load demand, energy storage device state of charge, and micro-grid internal node voltage and current; the state parameters of the external large grid include large grid frequency, voltage, power flow and fault signal.

[0022] Specifically, when obtaining the internal operation parameters of the micro-grid and the state parameters of the external large grid connected by the micro-grid, it includes: based on the distributed monitoring unit, real-time acquisition of real-time output data of various types of distributed power sources in the micro-grid, the real-time output data including DC side power and AC side power of photovoltaic array, and output power and impeller speed of wind turbine generator set; obtaining the instantaneous active power, reactive power, current effective value and voltage effective value of each key load point of the micro-grid, and performing short-term load prediction based on historical trends and weather conditions on the load data; real-time monitoring of the state of charge, battery pack voltage, charge and discharge current, battery health status and cycle life parameters of the energy storage device; collecting the key bus voltage amplitude, frequency, phase angle inside the micro-grid and the active and reactive power flow information of each tie line; obtaining the grid frequency, point of common coupling voltage deviation, large grid fault protection signal and tie line power transmission state of the external large grid, the fault protection signal including overcurrent, undervoltage or frequency out-of-limit alarm signal.

[0023] It can be understood that by constructing a multi-dimensional monitoring system covering the internal and external large power grid of the micro-grid, the system state can be fully perceived in different operation scenarios. Specifically, within the micro-grid, the distributed monitoring unit is used to collect real-time output of photovoltaic, wind power and other distributed power sources, including not only direct current side and alternating current side power, but also fan impeller speed and other dynamic operation characteristics, so as to accurately depict the fluctuation characteristics of new energy. At the same time, by collecting current, voltage, active and reactive power and other electrical parameters of the key load point, and combining historical trends and weather conditions to carry out short-term prediction, the load change trend can be identified in advance, thereby providing a forward-looking basis for operation mode decision. In addition, real-time monitoring of the state of charge, charging and discharging current, battery health and life of the energy storage system helps to accurately assess the adjustable capacity and standby capacity of the energy storage system, providing reliable support for micro-grid operation mode selection. Finally, at the grid interconnection level, by monitoring bus voltage, frequency, phase angle and power flow of each tie line, the global perception of the electrical topology and power distribution within the micro-grid is realized; at the same time, the frequency deviation of the external large power grid, the voltage fluctuation of the public connection point, the power transmission state of the tie line and the fault protection signal are obtained, so that the operation health status and risk information of the large power grid can be mastered in real time. Through the integration of the above internal and external parameters, the system can establish a high-precision, dynamically updated comprehensive operation state model, providing data basis and decision basis for subsequent mode identification and switching control.

[0024] In the specific embodiments in the present application, the above steps realize the scene in the following manner: the internal and external multi-source parameters are collected in real time by the distributed monitoring unit, and are used for comprehensive operation state modeling and analysis. The specific process is as follows: First, on the side of the distributed power source, the direct current side power of the photovoltaic array is monitored to be 480kW, the alternating current side power is 450kW, and the conversion efficiency is about 93.75%; at the same time, the output power of the wind turbine generator set is 300kW, and the impeller speed is 12.5r / s, indicating that the new energy output is in a high level operation state.

[0025] Secondly, on the load side, the real-time monitoring data of the key load point shows that the instantaneous active power is 620kW, the reactive power is 160kVar, the voltage effective value is 380V, and the current effective value is 950A. Combined with the historical trend and the sunny high temperature weather condition of the day, it is predicted that the load will increase by about 8% in the next period.

[0026] In terms of energy storage devices, the battery pack monitoring results are state of charge (SOC) 72%, total voltage of battery pack 720V, charging current 120A, health status index 0.95, and remaining ratio of cycle life 88%, indicating that the energy storage has certain adjustable and standby capacity.

[0027] At the level of power grid interconnection, the bus voltage amplitude of 381V, the frequency of 49.95Hz, the phase angle offset of 0.6°, the active power of each tie line of -50kW (i.e. sending power outwards), the reactive power of 20kVar are collected; at the same time, the external large power grid frequency of 49.94Hz, the common connection point voltage deviation of -1.5% and the received primary under-voltage alarm signal are monitored, prompting that the large power grid operation has slight abnormality, and the specific is shown in Table 1: Table 1 Micro-grid operation state data collection table

[0028] It can be seen that through the collection and fusion of the above data, the system can comprehensively judge that the micro-grid is in the state of grid-connected operation and slight voltage risk. Combined with the SOC of the energy storage and the under-voltage signal of the large power grid, the control strategy can be self-adaptively prepared to switch to the island mode to ensure the continuity and stability of local power supply.

[0029] The above scenario is only a preferred embodiment of the present application and is not used to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0030] Step S200, fuse the internal operation parameters and the external large power grid state parameters to construct a real-time comprehensive operation state model of the micro-grid.

[0031] Specifically, when fusing the micro-grid internal operation parameters and the external large power grid state parameters to construct the real-time comprehensive operation state model of the micro-grid, it includes: standardizing preprocessing the obtained multi-source operation information, including abnormal data elimination, missing data interpolation and filling and multi-source data timestamp alignment, to construct a time sequence consistent micro-grid operation data set; extracting feature parameters from the micro-grid operation data set, including system active power balance deviation, reactive power margin, voltage deviation rate and frequency deviation rate; inputting the feature parameters into a feature fusion network model based on deep learning to generate a high-dimensional comprehensive state feature vector; based on the high-dimensional comprehensive state feature vector, combining the preset micro-grid operation constraint conditions and optimization objectives, through fuzzy logic reasoning or expert system rules, real-time evaluating the power quality state, power supply reliability level, economic operation efficiency and system transient stability index of the micro-grid; according to the power quality state, power supply reliability level, economic operation efficiency and system transient stability index of the micro-grid, constructing the real-time comprehensive operation state model of the micro-grid.

[0032] Specifically, the acquired multi-source operation information is standardized and preprocessed, including abnormal data elimination, missing data interpolation and filling, and multi-source data timestamp alignment, to construct a time-consistent microgrid operation dataset. For the real-time collected distributed power output data, load demand data, energy storage device operation data and grid state data, a statistical-based outlier detection algorithm is implemented to identify and eliminate outlier data points that exceed the historical data range or physical limits. For data with missing time points, a cubic spline interpolation or an autoregressive model based on historical data patterns is used for accurate filling to ensure the continuity of the dataset. Through time synchronization protocol, asynchronous data from different monitoring devices are timestamp calibrated and aligned to eliminate time deviation caused by data transmission delay or asynchronous collection. All preprocessed original data are normalized to map feature values of different dimensions to a unified numerical interval, which is used to support the input of the deep learning model. A time series dataset containing instantaneous power, voltage, frequency, current, and corresponding volatility and deviation rates is constructed, and the dataset is structured as a multi-dimensional tensor.

[0033] It can be understood that by fusing the internal operation parameters of the microgrid with the external grid state parameters, a real-time, comprehensive and structured comprehensive operation state model is constructed to realize accurate evaluation and intelligent decision of the microgrid operation state. First, for the acquired multi-source operation information, the system is standardized and preprocessed, including abnormal data elimination, missing data interpolation and filling, and multi-source data timestamp alignment, to form a time-consistent, continuous and reliable microgrid operation dataset. This process ensures that heterogeneous data from different monitoring devices and different dimensions can be represented synchronously on the same time axis, providing basic data support for subsequent feature extraction and fusion. Second, feature parameters are extracted from the standardized operation dataset, including system active power balance deviation, reactive power margin, voltage deviation rate and frequency deviation rate, and other key indicators. These features can comprehensively reflect the operation characteristics of the microgrid in terms of energy supply and demand, voltage stability and frequency fluctuation. Subsequently, these features are input into a deep learning-based feature fusion network model to generate a high-dimensional comprehensive state feature vector, which fully captures the internal power flow characteristics, energy storage regulation capacity and the influence of external grid disturbances on the microgrid operation through nonlinear mapping, achieving high-precision modeling of complex operation states. Finally, based on the high-dimensional comprehensive state feature vector and in combination with the pre-set microgrid operation constraints and optimization objectives, the power quality state, power supply reliability level, economic operation efficiency and system transient stability of the microgrid are evaluated in real time through fuzzy logic reasoning or expert system rules. Through comprehensive analysis of these indicators, a real-time comprehensive operation state model that can dynamically reflect the overall operation state of the microgrid is constructed, providing decision-making basis for subsequent target operation mode identification, adaptive switching control and smooth switching strategy, realizing intelligent, stable and efficient operation of the microgrid.

[0034] In the specific embodiments in the present application, the above steps realize the scene in the following manner: the system collects internal and external multi-source parameters in real time, and constructs a comprehensive operation state model through standardized preprocessing and feature extraction.

[0035] First, the distributed power output, load demand, energy storage device and external grid state data are standardized preprocessed. Statistical outliers are detected to remove outliers, for example, abnormal points of photovoltaic power exceeding the historical maximum value are removed; missing time point data is filled by cubic spline interpolation or autoregressive model to ensure data continuity; data from different monitoring devices are time stamped by time synchronization protocol, and all features are normalized to map different dimensional features such as power, voltage, frequency and current to a unified interval, forming a multi-dimensional time series tensor data set to provide input for the deep learning model.

[0036] Subsequently, key feature parameters are extracted from the data set, including system active power balance deviation, reactive power margin, voltage deviation rate and frequency deviation rate, and input into the feature fusion network of deep learning to generate a high-dimensional comprehensive state feature vector. For example, the microgrid data collected at a certain time is shown in Table 2: Table 2 Microgrid multi-source operation data and feature parameter table

[0037] As can be seen, through the fusion and feature extraction of the above multi-source data, the system can evaluate the power quality, power supply reliability and transient stability of the microgrid in real time, and provide decision basis for subsequent target operation mode selection and adaptive switching. For example, when the power balance deviation and slight fluctuation of the external grid are detected, the system can adjust the energy storage output or optimize the load distribution in advance to maintain the stability and reliability of the microgrid operation.

[0038] The above scene is only a preferred embodiment of the present application and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0039] Step S300, based on the real-time comprehensive operation state model, adaptively determines the target operation mode of the microgrid.

[0040] Specifically, the target operation mode includes grid-connected mode, island mode, grid recovery mode or internal reconstruction mode.

[0041] Specifically, based on the real-time comprehensive operation state model, the target operation mode of the micro-grid is adaptively determined, including: according to the real-time comprehensive operation state model, extracting the fault state indication of the external large power grid, the power balance deviation of the micro-grid, the state of charge of the energy storage device, and the priority information of the key load; comparing the fault state indication with the preset fault threshold to determine the fault Boolean flag of the external power grid; comparing the power balance deviation with the preset stable operation interval to determine the internal stability Boolean flag of the micro-grid; comparing the state of charge of the energy storage device with the preset minimum state of charge threshold to determine the energy storage availability Boolean flag; comparing the priority information of the key load with the preset power supply guarantee level to determine the load importance Boolean flag; based on the configured multi-mode decision rule set, the logic aggregation is performed on each Boolean flag; Specifically, according to the logic aggregation result, the target operation mode of the micro-grid is adaptively determined, wherein: if the external large power grid is in fault and the micro-grid is internally stable, the emergency island mode is determined; if the external large power grid is restored and the micro-grid is internally smooth, the restored grid-connected mode is determined; if the distributed power output in the micro-grid is excessive, the optimized grid-connected mode is determined; if the distributed power output in the micro-grid is insufficient and the energy storage availability is low, the priority grid-connected mode is determined.

[0042] Specifically, according to the logic aggregation result, the target operation mode of the micro-grid is adaptively determined, including: if the fault Boolean flag of the external power grid is true and the internal stability Boolean flag of the micro-grid is true, the micro-grid enters the emergency island mode; if the fault Boolean flag of the external power grid is false and the internal stability Boolean flag of the micro-grid is true, and the energy storage availability Boolean flag is true, the micro-grid enters the optimized grid-connected mode; if the fault Boolean flag of the external power grid is false and the internal stability Boolean flag of the micro-grid is true, but the energy storage availability Boolean flag is false, and the distributed power output in the micro-grid is insufficient to meet the load demand, the micro-grid enters the priority grid-connected mode; if the fault Boolean flag of the external power grid is false and the internal stability Boolean flag of the micro-grid is true, and the distributed power output in the micro-grid is excessive, the micro-grid enters the grid-fed mode; if the fault Boolean flag of the external power grid is false and the internal stability Boolean flag of the micro-grid is true, and the distributed power output in the micro-grid meets the load demand, but the energy storage availability Boolean flag is false, the micro-grid enters the independent operation mode.

[0043] Specifically, when configuring the multi-mode decision rule set, the following steps are included: obtaining key operating index thresholds in different operating modes, including grid voltage and frequency stability intervals in grid-connected mode, minimum power balance margin in island mode, charge and discharge safety limits of energy storage devices, and rated output range of distributed power sources; according to the key operating index thresholds, establishing operating condition conditions describing the operation of the microgrid under various external grid states, internal power generation capacity and load demand changes, the condition conditions are defined as a series of logical judgment rules; for each operating condition, the corresponding target operating mode and mode switching priority are preset; combined with the evaluation results in the real-time comprehensive operating state model, a mode switching decision matrix based on a multi-objective optimization algorithm is constructed, the matrix considers multiple factors such as power supply reliability, operating economy, power quality and system stability, and each factor is assigned a weight; based on the mode switching decision matrix, a set of dynamically adjusted fuzzy logic rule base or strategy table based on reinforcement learning is established, the rule base or strategy table can adaptively adjust the triggering conditions and execution strategies of mode switching according to the subtle changes of real-time operating state.

[0044] It can be understood that based on the real-time comprehensive operation state model, the optimal operation mode of the microgrid under different external grid conditions and internal operation conditions is adaptively identified, and intelligent and dynamic operation decision is realized. Specifically, first, key indicators are extracted from the comprehensive state model, including external large grid fault state, microgrid internal power balance deviation, energy storage device state of charge, and key load priority information. By comparing with the preset threshold, these indicators are converted into Boolean markers for determining whether the external grid is faulty, the microgrid is stable, the energy storage is available, and the power supply guarantee level of the key load, thereby forming a logical input for mode determination. Subsequently, based on the configured multi-mode decision rule set, the Boolean markers are logically aggregated to form a comprehensive evaluation of the current working condition of the microgrid. The rule set covers key indicator thresholds under different operation modes, such as voltage and frequency stability intervals in grid-connected mode, minimum power balance margin in island mode, energy storage device charging and discharging limits, and distributed power rated output range. By comparing the real-time state with these thresholds, the target operation mode can be adaptively selected under different working conditions of the microgrid, including emergency island mode, recovery grid-connected mode, optimized grid-connected mode, priority grid-connected mode or independent operation mode. Finally, a mode switching decision matrix based on multi-objective optimization is constructed, and factors such as power supply reliability, economic operation, power quality and stability are considered in weight distribution and optimization. Combined with dynamically adjusted fuzzy logic rule base or reinforcement learning strategy table, the mode switching trigger condition and execution strategy can be adaptively adjusted according to the slight changes of real-time operation state, so as to ensure that the microgrid can run smoothly and flexibly respond to external disturbances and internal fluctuations under different working conditions, and realize safe, reliable and economic adaptive operation.

[0045] In the specific embodiments in the present application, the above steps realize the scene in the following manner: based on the real-time comprehensive operation state model, the target operation mode of the microgrid is adaptively determined to ensure that the microgrid can still run stably, safely and efficiently under the changing external grid conditions and internal load fluctuations. First, key operation indicators are collected and processed, including external large grid fault state, microgrid power balance deviation, energy storage device state of charge, and key load priority information. The external large grid fault state is determined by the grid frequency deviation, the public connection point voltage deviation and the fault protection signal; the microgrid power balance deviation is calculated by the internal distributed power output, the load demand and the energy storage device charging and discharging capacity; the energy storage device state of charge includes the battery SOC, the battery voltage and the charging and discharging current; and the key load priority is evaluated by the load type, the importance level and the power supply guarantee demand.

[0046] Secondly, by comparing these indicators with preset thresholds, such as the power balance deviation compared with the stable operation interval (±50 kW) and the energy storage state of charge compared with the minimum available threshold (40% SOC), Boolean flags are generated to indicate whether the external power grid is faulty, the microgrid is stable, the energy storage is available, and the critical load needs to be prioritized. Then, these Boolean flags are input into the configured multi-mode decision rule set, combined with the mode switching decision matrix and the dynamically adjusted fuzzy logic rule or reinforcement learning strategy, and logically aggregated to adaptively select the target operation mode. As shown in Table 3: Table 3 Microgrid mode decision example data

[0047] In addition, according to the logic aggregation rule, when the external power grid is normal (Boolean flag False), the microgrid is stable (True), and the energy storage is available (True), the optimized grid-connected mode is automatically selected to fully utilize the distributed power output in the microgrid, achieving economic operation and load power supply balance. At another time, if the external power grid is faulty (Boolean flag True) and the microgrid is still stable (True), the emergency islanding mode is immediately switched to ensure that the microgrid can independently bear the load, while preventing external fault conduction from causing power quality degradation or power supply interruption. If the energy storage is insufficient and the distributed power output is lower than the load demand, the priority grid-connected mode is selected to prioritize stable power supply through grid connection.

[0048] Finally, the mode switching decision matrix also combines multi-objective optimization indicators such as power supply reliability, power quality, economic operation, and transient stability, assigns weights to each factor, and realizes real-time adjustment through dynamic fuzzy logic rules or reinforcement learning strategies. For example, when power fluctuations occur in the microgrid, the energy storage output or load distribution can be adjusted in advance to ensure smooth switching and unaffected microgrid operation.

[0049] As can be seen, through the above example, it can be seen that according to the multi-source indicators and real-time operation state, the optimal operation mode is adaptively selected, and safe, reliable and efficient microgrid operation is achieved under various working conditions, providing strong support for the practical application of intelligent microgrids.

[0050] The above scenarios are only preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0051] Step S400, according to the target operation mode, generating and executing the smooth switching control instruction of the microgrid.

[0052] Specifically, when generating and executing the smooth switching control instruction of the micro-grid, the following steps are included: dynamically adjusting the operation parameters of the distributed power supply and energy storage device inside the micro-grid for the target operation mode to meet the power balance requirement before mode switching; monitoring the voltage, frequency and phase angle of the key nodes of the micro-grid in real time before executing the mode switching instruction to ensure that they are within the preset synchronization tolerance range; when it is determined to switch from the grid-connected mode to the island mode, sending a tripping instruction of the grid-connected circuit breaker connected with the external large power grid, and simultaneously starting the frequency and voltage droop control of the main control inverter inside the micro-grid to maintain the voltage and frequency stability inside the micro-grid; when it is determined to switch from the island mode to the grid-connected mode, performing accurate grid-connected synchronization detection, including matching detection of the voltage amplitude, frequency and phase angle; after the synchronization condition is met, sending a closing instruction of the grid-connected circuit breaker, gradually releasing the droop control inside the micro-grid, and smoothly transferring the control right of the load and distributed power supply to the large power grid; during the whole switching process, the transient impact inside the micro-grid is monitored in real time, and the fast-responding energy storage device or reactive power compensation equipment is used for dynamic suppression to minimize voltage drop, frequency fluctuation and current impact.

[0053] Specifically, when monitoring the transient impact inside the micro-grid in real time and dynamically suppressing it by using the fast-responding energy storage device or reactive power compensation equipment during the whole switching process, the following steps are included: collecting the voltage, current and frequency data of the micro-grid public connection point and key load bus in real time, and capturing the instantaneous transient change waveform by using high-frequency sampling technology; using fast Fourier transform (FFT) or wavelet analysis method to perform frequency spectrum analysis on the transient waveform, identifying the transient fault characteristics such as voltage drop, frequency oscillation or instantaneous power impact, and calculating the severity index thereof; when the severity index of the transient fault characteristics exceeds the preset threshold, triggering the fast response mechanism; for voltage drop or overvoltage, controlling the power converter of the energy storage device to quickly adjust the output voltage amplitude and phase angle, or through the dynamic reactive power compensation device to instantaneously inject or absorb reactive power, so as to quickly restore voltage stability; for frequency oscillation or frequency overrun, controlling the energy storage device to perform fast charging and discharging to provide inertia support or power balance, thereby effectively suppressing frequency fluctuation; at the same time, the output power of the controllable distributed power supply is coordinated and controlled to participate in primary and secondary frequency regulation; the dynamic suppression strategy includes an adaptive control algorithm based on model predictive control (MPC), which adjusts the compensation amount and compensation timing according to real-time transient response.

[0054] It can be understood that the smooth switching control instruction is generated by the real-time comprehensive operation state model of the micro-grid to realize safe, reliable and efficient switching between different operation modes. First, for the target operation mode, the operation parameters of the distributed power supply and energy storage device in the micro-grid are dynamically adjusted to meet the power balance requirement before switching, so as to ensure that the internal load and power generation capacity of the grid are matched during the switching process, and frequency or voltage fluctuation caused by power imbalance is avoided. Before switching, the voltage, frequency and phase angle of the key nodes of the micro-grid are monitored in real time, and it is judged whether they are within the preset synchronization tolerance range, which provides necessary synchronization guarantee for mode switching. When the micro-grid switches from the grid-connected mode to the island mode, the opening instruction is sent to the grid-connected circuit breaker, and the voltage and frequency droop control of the internal main control inverter of the micro-grid is started, so as to realize the autonomous adjustment and stable maintenance of the voltage and frequency of the micro-grid. When switching to the grid-connected mode, accurate grid synchronization detection is performed, including voltage amplitude, frequency and phase angle matching, to ensure that the grid-connected condition is met before sending the closing instruction, and gradually releasing the internal droop control of the micro-grid, and smoothly transferring the control right of the load and distributed power supply to the large grid, realizing seamless grid connection. In the switching process, the transient impact on the micro-grid is monitored in real time, and the dynamic suppression is performed through the fast-responding energy storage device and reactive power compensation equipment. Specifically, the voltage, current and frequency data of the public connection point and the key load bus are collected at high frequency, and the frequency spectrum analysis is performed by using fast Fourier transform (FFT) or wavelet analysis, the transient characteristics such as voltage drop, frequency oscillation and instantaneous power impact are identified, and the severity index is calculated. When the index exceeds the preset threshold, the fast response mechanism is triggered: when the voltage drops or overvoltage occurs, the energy storage power converter can quickly adjust the output voltage amplitude and phase angle, or the dynamic reactive power compensation device can inject / absorb reactive power to restore voltage stability; when the frequency oscillation or over-limit occurs, the energy storage device performs fast charging and discharging to provide inertia support or power balance, and at the same time, the controllable distributed power supply participates in primary / secondary frequency regulation. Finally, the adaptive control algorithm based on model predictive control (MPC) is adopted to optimize the dynamic suppression strategy in real time, and the compensation amount and compensation time are adjusted according to the transient response to minimize voltage drop, frequency fluctuation and current impact, so as to realize the smoothing and high reliability operation of the micro-grid switching process. Through the above technical principles, the present application can realize safe, efficient and adaptive switching of the micro-grid in multiple operation modes, and significantly improve the power supply reliability and power quality of the micro-grid.

[0055] In the specific embodiments in the present application, the above steps realize the scene in the following manner: in the scene of switching the micro-grid from the grid-connected mode to the island mode, for example, when the external large grid fails (overcurrent or frequency out of limit), first, the output power of the distributed power source (such as a photovoltaic array, a wind generator) and the energy storage device (such as a lithium battery) in the micro-grid is dynamically adjusted to ensure power balance before switching. Subsequently, the voltage, frequency and phase angle of the key node are monitored to ensure that each parameter is within the synchronization tolerance range, and then the grid-connected breaker tripping instruction is sent, and the voltage and frequency droop control of the master control inverter is started to realize the stable operation of the island mode.

[0056] In the scene of switching the micro-grid from the island mode back to the grid-connected mode, accurate grid-connected synchronization detection is performed, including voltage amplitude, frequency and phase angle matching. When each parameter meets the condition, the grid-connected breaker closing instruction is sent, and the internal droop control of the micro-grid is gradually removed to smoothly transfer the control right of the load and the distributed power source to the large grid, avoiding the occurrence of power quality decline or transient power impact in the switching process.

[0057] In the switching process, the transient impact is monitored in real time. For example, the voltage, current and frequency data of the key bus are collected in real time through high-frequency sampling, and the transient waveform is analyzed by using fast Fourier transform (FFT). When the voltage drop or frequency oscillation exceeds the set threshold, the fast response mechanism is triggered: the energy storage device provides inertia support or power balance by fast charging and discharging, the dynamic reactive power compensation device injects or absorbs reactive power, and the controllable distributed power source participates in frequency regulation to ensure that the voltage and frequency of the micro-grid quickly return to the stable range, as shown in Table 4: Table 4 Data table of micro-grid switching example

[0058] As can be seen, through the above example, it can be seen that the present application realizes the smooth switching of the micro-grid between the grid-connected mode and the island mode by using the dynamic adjustment of the distributed power source and the energy storage device, real-time transient monitoring and fast response mechanism, and effectively suppresses the voltage drop, frequency fluctuation and transient power impact, and ensures the power quality and system stability.

[0059] The above scene is only a preferred embodiment of the present application and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0060] In the above embodiment, by fusing the internal operating parameters of the micro-grid and the external grid state parameters, a real-time comprehensive operating state model is constructed. Compared with the traditional method which only relies on a single monitoring signal, the overall operating condition of the micro-grid can be more comprehensively and accurately reflected. This multi-source information fusion mechanism avoids mode determination errors caused by incomplete information, thereby improving the accuracy of switching decisions. Secondly, based on the comprehensive operating state model, the target operating mode is adaptively determined, so that the micro-grid can dynamically select the optimal mode according to different operating environments and working conditions, realizing the intelligence and flexibility of mode switching. For example, in the case of grid failure or abnormality, the system can timely identify and switch to island mode to ensure the continuous power supply of local load; when the external grid is restored to stability, it can smoothly realize grid connection or reconstruction operation, thereby enhancing the continuity and reliability of power supply. Finally, by generating and executing smooth switching control instructions, the application can effectively reduce voltage fluctuations, frequency deviations and power surges during the switching process, avoiding the decline of power quality and the risk of equipment operation caused by mode switching. Overall, this method not only improves the stability and safety of micro-grid operation, but also enhances the adaptability and robustness of its coordinated operation with the grid, and has good popularization and application value.

[0061] In another preferred mode based on the above embodiment, as shown in Figure 3 The present embodiment provides a micro-grid adaptive switching control system based on multi-source information fusion, comprising: An information acquisition module is configured to acquire the internal operating parameters of the micro-grid and the state parameters of the external grid in real time, wherein the internal operating parameters include distributed power output, load demand, energy storage device state of charge, micro-grid internal node voltage and current; the state parameters of the external grid include grid frequency, voltage, power flow and fault signal; A data evaluation module is electrically connected with the information acquisition module and is configured to receive multi-source operating information and pre-process the information, including abnormal data elimination, missing data interpolation and time stamp alignment; the data evaluation module is also configured to fuse the pre-processed multi-source operating information to construct a real-time comprehensive operating state model of the micro-grid; A mode decision module is electrically connected with the data evaluation module and is configured to adaptively determine the target operating mode of the micro-grid based on the real-time comprehensive operating state model, wherein the target operating mode includes grid-connected mode, island mode, restored grid-connected mode or internal reconstruction mode; The switching control module is electrically connected with the mode decision module, and is configured to generate and execute a smooth switching control instruction of the micro-grid according to the target operation mode, wherein the instruction comprises grid-connected breaker opening and closing control, distributed power output adjustment, energy storage device charging and discharging strategy, and voltage frequency droop control; the switching control module is further configured to monitor system transient in real time and perform effective dynamic suppression during the switching process.

[0062] It can be understood that the micro-grid adaptive switching control method and system based on multi-source information fusion in each of the above embodiments of the application have the same beneficial effects, which will not be described again.

[0063] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can adopt a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) containing computer usable program codes.

[0064] The present application is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams can be implemented by computer program instructions, and the combination of flows and / or blocks in the flowcharts and / or block diagrams. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device that implements the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks Figure 1 The functions specified in one or more flows and / or blocks

[0065] These computer program instructions can also be stored in a computer readable memory capable of guiding the computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer readable memory produce a product including instruction devices, which implement the functions specified in the flowcharts and / or block diagrams. Figure 1 The functions specified in one or more flows and / or blocks Figure 1 The functions specified in one or more flows and / or blocks

[0066] These computer program instructions can also be loaded into a computer or other programmable data processing device, so that a series of operation steps are performed on the computer or other programmable device to produce a computer-implemented process, so that the instructions executed on the computer or other programmable device provide a product for implementing the functions specified in the flowcharts and / or block diagrams. Figure 1a process or processes and / or blocks Figure 1 steps of the functions specified in the block or blocks.

[0067] It should be noted that the above-mentioned embodiments are only used to illustrate the technical solutions of the present application, but not limit the technical solutions of the present application. Although the present application has been described in detail with reference to the above-mentioned embodiments, it should be understood by those skilled in the art that the specific embodiments of the present application can be modified or equivalent replaced without departing from the spirit and scope of the present application, and any modification or equivalent replacement should be covered in the protection scope of the claims of the present application.

Claims

1. A microgrid adaptive switching control method based on multi-source information fusion, characterized in that, The method comprises the following steps: acquiring internal operation parameters of the micro-grid and state parameters of an external large power grid connected to the micro-grid, wherein the internal operation parameters include distributed power output, load demand, energy storage device state of charge, and micro-grid internal node voltage and current; the state parameters of the external large power grid include large power grid frequency, voltage, power flow, and fault signal; fusing the internal operation parameters and the state parameters of the external large power grid to construct a real-time comprehensive operation state model of the micro-grid; based on the real-time comprehensive operation state model, adaptively determining a target operation mode of the micro-grid, wherein the target operation mode includes a grid-connected mode, an islanded mode, a grid recovery mode, or an internal reconstruction mode; generating and executing smooth switching control instructions of the micro-grid according to the target operation mode.

2. The microgrid adaptive switching control method based on multi-source information fusion according to claim 1, characterized in that, When acquiring the internal operation parameters of the micro-grid and the state parameters of the external large power grid connected to the micro-grid, the method comprises the following steps: based on a distributed monitoring unit, real-time output data of various types of distributed power sources in the micro-grid are collected in real time, the real-time output data including direct current side power and alternating current side power of a photovoltaic array, and output power and impeller speed of a wind turbine generator set; instantaneous active power, reactive power, current effective value, and voltage effective value of each key load point in the micro-grid are acquired, and short-term load prediction based on historical trends and weather conditions is performed on the load data; the state of charge, battery pack voltage, charging and discharging current, battery health status, and cycle life parameters of the energy storage device are monitored in real time; voltage amplitude, frequency, phase angle of key busbars inside the micro-grid, and active and reactive power flow information of each tie line are collected; the power grid frequency, point of common coupling voltage deviation, large power grid fault protection signal, and tie line power transmission state of the external large power grid are acquired, and the fault protection signal includes overcurrent, undervoltage, or frequency out-of-limit alarm signal.

3. The microgrid adaptive switching control method based on multi-source information fusion according to claim 1, characterized in that, When fusing the internal operation parameters of the micro-grid and the state parameters of the external large power grid to construct the real-time comprehensive operation state model of the micro-grid, the method comprises the following steps: standardized preprocessing of the acquired multi-source operation information is performed, including abnormal data elimination, missing data interpolation and filling, and multi-source data timestamp alignment, to construct a time-consistent micro-grid operation data set; feature parameters are extracted from the micro-grid operation data set, including system active power balance deviation, reactive power margin, voltage deviation rate, and frequency deviation rate; the feature parameters are input into a feature fusion network model based on deep learning to generate a high-dimensional comprehensive state feature vector; based on the high-dimensional comprehensive state feature vector, in combination with preset micro-grid operation constraint conditions and optimization objectives, through fuzzy logic reasoning or expert system rules, the real-time evaluation of the micro-grid is performed in terms of power quality state, power supply reliability level, economic operation efficiency, and system transient stability index; the real-time comprehensive operation state model of the micro-grid is constructed according to the power quality state, power supply reliability level, economic operation efficiency, and system transient stability index of the micro-grid.

4. The microgrid adaptive switching control method based on multi-source information fusion according to claim 3, characterized in that, When the standardized preprocessing of the acquired multi-source operation information is performed, including abnormal data elimination, missing data interpolation and filling, and multi-source data timestamp alignment, to construct a time-consistent micro-grid operation data set, the method comprises the following steps: For the real-time collected distributed power output data, load demand data, energy storage device operation data and power grid state data, a statistical-based outlier detection algorithm is implemented to identify and eliminate outlier data points that exceed the range of historical data or physical limits; For data with missing time points, cubic spline interpolation or autoregressive model based on historical data pattern is used for accurate filling to ensure the continuity of the data set; Through the time synchronization protocol, the asynchronous data from different monitoring devices are timestamp calibrated and aligned to eliminate the time deviation caused by data transmission delay or asynchronous collection; All pre-processed raw data are normalized to map the feature values of different dimensions to a unified numerical interval, and are used as input for the deep learning model; a time series data set containing instantaneous power, voltage, frequency, current and corresponding volatility rate, deviation rate is constructed, and the data set is structured as a multi-dimensional tensor.

5. The microgrid adaptive switching control method based on multi-source information fusion according to claim 3, characterized in that, Based on the real-time comprehensive operation state model, the target operation mode of the microgrid is adaptively determined, including: According to the real-time comprehensive operation state model, the fault state indication of the external large power grid, the power balance deviation of the microgrid, the state of charge of the energy storage device and the priority information of the key load are extracted; The fault state indication is compared with the preset fault threshold to determine the fault Boolean marker of the external power grid; The power balance deviation is compared with the preset stable operation interval to determine the internal stability Boolean marker of the microgrid; The state of charge of the energy storage device is compared with the preset minimum state of charge threshold to determine the energy storage availability Boolean marker; The priority information of the key load is compared with the preset power supply guarantee level to determine the load importance Boolean marker; Based on the configured multi-mode decision rule set, the Boolean markers are logically aggregated; According to the logical aggregation result, the target operation mode of the microgrid is adaptively determined, wherein: If the external large power grid is in fault and the microgrid is internally stable, the emergency island mode is determined; If the external large power grid is restored and the microgrid is internally smooth, the restored grid-connected mode is determined; If the distributed power output in the microgrid is excessive, the optimized grid-connected mode is determined; If the distributed power output in the microgrid is insufficient and the energy storage availability is low, the priority grid-connected mode is determined.

6. The microgrid adaptive switching control method based on multi-source information fusion according to claim 5, characterized in that, According to the logical aggregation result, the target operation mode of the microgrid is adaptively determined, including: If the fault Boolean marker of the external power grid is true and the internal stability Boolean marker of the microgrid is true, the microgrid enters the emergency island mode; If the fault Boolean marker of the external power grid is false and the internal stability Boolean marker of the microgrid is true, and the energy storage availability Boolean marker is true, the microgrid enters the optimized grid-connected mode; If the fault Boolean marker of the external power grid is false and the internal stability Boolean marker of the microgrid is true, but the energy storage availability Boolean marker is false, and the distributed power output in the microgrid is insufficient to meet the load demand, the microgrid enters the priority grid-connected mode; If the fault Boolean marker of the external power grid is false and the internal stability Boolean marker of the microgrid is true, and the distributed power output in the microgrid is excessive, the microgrid enters the grid feeding mode; If the external grid failure Boolean flag is false and the microgrid internal stability Boolean flag is true, and the microgrid internal distributed power output meets the load demand, but the energy storage availability Boolean flag is false, it is determined that the microgrid enters the independent operation mode.

7. The microgrid adaptive switching control method based on multi-source information fusion according to claim 5, characterized in that, When configuring the multi-mode decision rule set, the following steps are included: Obtaining the threshold values of key operation indicators in different operation modes, including the grid voltage and frequency stability interval in the grid-connected mode, the minimum power balance margin in the island mode, the charge and discharge safety limit of the energy storage device, and the rated output range of the distributed power supply; According to the threshold values of the key operation indicators, the operation condition of the microgrid under various external grid states, internal power generation capacity and load demand changes is established, and the condition is defined as a series of logical judgment rules; For each condition, the corresponding target operation mode and mode switching priority are preset; Combined with the evaluation results in the real-time comprehensive operation state model, a mode switching decision matrix based on a multi-objective optimization algorithm is constructed, which considers multiple factors such as power supply reliability, operation economy, power quality and system stability, and assigns weights to each factor; Based on the mode switching decision matrix, a set of dynamically adjusted fuzzy logic rule base or strategy table based on reinforcement learning is established, which can adaptively adjust the triggering conditions and execution strategies of mode switching according to the subtle changes of real-time operation state.

8. The microgrid adaptive switching control method based on multi-source information fusion according to claim 1, characterized in that, When generating and executing the smooth switching control instructions of the microgrid, the following steps are included: For the target operation mode, dynamically adjust the operation parameters of the distributed power supply and the energy storage device in the microgrid to meet the power balance requirement before mode switching; Before executing the mode switching instructions, real-time monitor the voltage, frequency and phase angle of the key nodes of the microgrid to ensure that they are within the preset synchronization tolerance range; When it is determined to switch from the grid-connected mode to the island mode, send the opening instruction of the grid-connected circuit breaker connected with the external large grid, and start the frequency and voltage droop control of the internal master inverter of the microgrid to maintain the voltage and frequency stability of the microgrid; When it is determined to switch from the island mode to the grid-connected mode, perform accurate grid synchronization detection, including voltage amplitude, frequency and phase angle matching detection; After meeting the synchronization conditions, send the closing instruction of the grid-connected circuit breaker, gradually release the droop control of the microgrid, and smoothly transfer the control of the load and distributed power supply to the large grid; During the whole switching process, real-time monitor the transient impact in the microgrid, and dynamically suppress it by using fast-responding energy storage devices or reactive power compensation equipment to minimize voltage drop, frequency fluctuation and current impact.

9. The microgrid adaptive switching control method based on multi-source information fusion according to claim 8, characterized in that, When real-time monitoring the transient impact in the microgrid and dynamically suppressing it by using fast-responding energy storage devices or reactive power compensation equipment during the whole switching process, the following steps are included: Real-time collect the voltage, current and frequency data of the microgrid public connection point and key load bus, and use high-frequency sampling technology to capture the instantaneous transient change waveform; Use fast Fourier transform (FFT) or wavelet analysis method to perform frequency spectrum analysis on the transient waveform, identify the transient fault characteristics such as voltage drop, frequency oscillation or instantaneous power impact, and calculate the severity index; When the severity index of the transient fault feature exceeds a preset threshold, a fast response mechanism is triggered; For voltage sag or overvoltage, the power converter of the energy storage device is controlled to quickly adjust the output voltage amplitude and phase angle, or the dynamic reactive power compensation device is used to instantaneously inject or absorb reactive power, so as to quickly restore voltage stability; For frequency oscillation or frequency limit, the energy storage device is controlled to perform fast charging and discharging to provide inertia support or power balance, thereby effectively suppressing frequency fluctuations; at the same time, the output power of the controllable distributed power supply is coordinated and controlled to participate in primary and secondary frequency regulation; The dynamic suppression strategy includes an adaptive control algorithm based on model predictive control (MPC), which adjusts the compensation amount and compensation timing according to real-time transient response.

10. A microgrid adaptive switching control system based on multi-source information fusion, suitable for the microgrid adaptive switching control method based on multi-source information fusion according to any one of claims 1-9, characterized in that, Comprise: An information acquisition module configured to acquire internal operating parameters of the microgrid and state parameters of an external large power grid in real time, wherein the internal operating parameters include distributed power output, load demand, energy storage device state of charge, microgrid internal node voltage and current; the state parameters of the external large power grid include large power grid frequency, voltage, power flow and fault signal; A data evaluation module electrically connected with the information acquisition module, configured to receive multi-source operating information and pre-process the information, including abnormal data rejection, missing data interpolation and time stamp alignment; the data evaluation module is also configured to fuse the pre-processed multi-source operating information to construct a real-time comprehensive operating state model of the microgrid; A mode decision module electrically connected with the data evaluation module, configured to adaptively determine the target operating mode of the microgrid based on the real-time comprehensive operating state model, wherein the target operating mode includes grid-connected mode, islanded mode, restoration grid-connected mode or internal reconfiguration mode; A switching control module electrically connected with the mode decision module, configured to generate and execute smooth switching control instructions of the microgrid according to the target operating mode, wherein the instructions include grid-connected circuit breaker opening and closing control, distributed power output adjustment, energy storage device charging and discharging strategy and voltage frequency droop control; the switching control module is also configured to monitor the system transient in real time and perform effective dynamic suppression during the switching process.

Images