Multimodal flexible control system for charging and swapping microgrids that integrates photovoltaic power consumption priority

Abstract

This invention relates to the field of charging and swapping microgrid technology, specifically to a multimodal flexible control system for charging and swapping microgrids that integrates photovoltaic (PV) consumption priority. It includes: a real-time microgrid status sensing unit; a PV consumption priority dynamic determination unit; a multimodal control unit for charging and swapping microgrids; and a flexible scheduling unit for charging and swapping resources. This invention, through a scenario adaptation module, accurately distinguishes three microgrid operation scenarios—PV energy surplus, peak charging and swapping load, and grid interaction constraints—based on multi-physical quantity coupling criteria. Simultaneously, the multimodal control unit can switch to three control modes: PV priority consumption, charging and swapping load guarantee, and grid interaction stability. This solves the problems of insufficient load adaptability and lack of dynamic response in multiple scenarios, and can specifically guarantee the rated operating power demand during peak charging and swapping load periods, reducing PV power waste and curtailment.

Description

Technical Field

[0001] This invention relates to the field of charging and swapping microgrid technology, and more specifically, to a multimodal flexible control system for charging and swapping microgrids that integrates photovoltaic power consumption priority. Background Technology

[0002] With the deep integration of photovoltaic new energy and electric vehicle charging and swapping services, charging and swapping microgrids have become the core carrier connecting clean energy and transportation electricity. They need to simultaneously achieve three core objectives: maximizing the absorption of photovoltaic output to reduce curtailment rates, ensuring stable power supply to charging and swapping loads (which exhibit significant peak-period characteristics), and controlling power fluctuations in the microgrid-main grid interconnection lines to avoid grid impact. However, photovoltaic output is highly random due to the influence of sunlight and temperature, and charging and swapping loads face the risk of supply-demand imbalance. Traditional microgrid control schemes often focus on a single objective (such as optimizing only photovoltaic absorption or load power supply), making it difficult to dynamically balance the contradictions among the three, resulting in insufficient flexibility and economy in microgrid operation. Therefore, flexible control technologies adapted to this scenario are urgently needed.

[0003] Existing technologies include research on photovoltaic microgrid control. For example, Chinese invention patent CN202310158626.4 discloses "A Smart Photovoltaic Microgrid Structure and Operation Control Method," which discloses a microgrid structure including photovoltaic modules, energy storage batteries, DC buses, and controllers. It reduces energy storage usage through dynamic access of group loads and optimization of the DC bus, and relies on hardware structure design to improve the efficiency of local photovoltaic consumption and power transmission. Another example is Chinese invention patent CN202411151746.2, which discloses "A Coordinated Control Strategy for Wind-Solar-Storage DC Microgrid." It proposes to analyze two operating modes: grid-connected free / dispatch, establish a multi-system control model for photovoltaic, wind power, and energy storage, and adjust power output by switching between MPPT (Maximum Power Point Tracking) and LPTC (Limited Power Point Tracking) modes to achieve energy coordination within the microgrid.

[0004] While the two existing technologies mentioned above provide ideas for photovoltaic microgrid control, they are not well adapted to the actual needs of charging and swapping microgrids and have obvious technical defects: First, they lack adaptability to charging and swapping loads and dynamic response capabilities in multiple scenarios. CN202310158626.4 relies solely on hardware structure optimization of "DC bus + group load" to achieve photovoltaic absorption, without designing a dedicated power supply guarantee mechanism for the time-limited peak characteristics of charging and swapping loads, and lacks dynamic adaptation capabilities for multiple operating scenarios such as "photovoltaic surplus, charging and swapping peak, and grid constraints"—when charging and swapping loads increase sharply or photovoltaic output changes abruptly, insufficient power supply or photovoltaic curtailment problems are likely to occur. Second, they lack prediction error correction and flexible scheduling mechanisms for charging and swapping resources. Although CN202411151746.2 establishes a multi-system control model, power regulation relies on fixed MPPT / LPTC mode switching and does not consider the impact of photovoltaic output prediction errors on power supply. The impact on control accuracy is significant. Firstly, the lack of a confidence correction mechanism for sliding window error estimation and the absence of weight smoothing constraints from first-order inertial filtering make priority determination susceptible to prediction bias. Furthermore, since the controlled object is a general microgrid load, it does not address power output adjustment or connection timing optimization for charging and swapping equipment, failing to meet the flexible scheduling requirements of charging and swapping loads. Secondly, it lacks control loop and grid interaction stability control capabilities. Neither CN202310158626.4 nor CN202411151746.2 constructs a complete control loop of "priority determination - mode switching - execution feedback." CN202310158626.4 lacks dynamic adjustment capabilities for control commands, and CN202411151746.2 does not mention control strategies for main grid tie-line power fluctuations, failing to address stable operation requirements under grid interaction constraints and easily leading to excessive tie-line power fluctuations. Therefore, we propose a multi-modal flexible control system for charging and swapping microgrids that integrates photovoltaic consumption priorities. Summary of the Invention

[0005] The purpose of this invention is to provide a multimodal flexible control system for charging and swapping microgrids that integrates photovoltaic consumption priority, in order to solve the problems mentioned in the background art, such as lack of charging and swapping load adaptability and multi-scenario dynamic response capability, lack of prediction error correction and flexible scheduling mechanism for charging and swapping resources, and lack of control closed loop and grid interaction stability control capability.

[0006] To address the aforementioned technical problems, the present invention aims to provide a multimodal flexible control system for charging and swapping microgrids that integrates photovoltaic power consumption priorities, comprising: The microgrid status real-time sensing unit collects in real time the voltage and current of photovoltaic modules, the load current of charging and swapping equipment, the voltage and current of energy storage batteries, and the power signal of the microgrid-main grid interconnection line in the charging and swapping microgrid. It performs noise reduction and outlier removal preprocessing on the collected signals and transmits the preprocessed data to the photovoltaic consumption priority dynamic determination unit. The photovoltaic (PV) grid integration priority dynamic determination unit receives preprocessed data and performs short-term trend predictions on PV output, remaining energy storage capacity, and charging / swapping load. It generates a prediction confidence factor through sliding window error estimation, applies smooth convergence constraints to weight adjustment by combining first-order inertial filtering and state timer, dynamically distinguishes microgrid operation scenarios and adapts the weight ratio of core indicators, generates a PV grid integration priority result with confidence correction, and transmits it to the charging / swapping microgrid multi-mode control unit. The charging and swapping microgrid multi-mode control unit receives the photovoltaic consumption priority result from the photovoltaic consumption priority dynamic determination unit, switches to three control modes corresponding to photovoltaic priority consumption, charging and swapping load guarantee, and grid interaction stability, and outputs control commands to the charging and swapping resource flexible scheduling unit. The flexible scheduling unit for charging and swapping resources receives control commands from the multi-mode control unit of the charging and swapping microgrid, adjusts the power output of the charging and swapping equipment, the control parameters of the energy storage battery charging and discharging circuit, and the load access sequence, so as to realize flexible control of the charging and swapping microgrid with photovoltaic consumption priority orientation.

[0007] As a further improvement to this technical solution, the real-time microgrid status sensing unit includes a distributed acquisition module, a signal preprocessing module, and a data transmission module, wherein: The distributed acquisition module consists of a voltage sensor, a current sensor, and a power sensor. The voltage sensor is used to acquire the terminal voltage of the photovoltaic module and the terminal voltage of the energy storage battery. The current sensor is used to acquire the load current of the charging and swapping equipment and the charging and discharging current of the energy storage battery. The power sensor is used to acquire the instantaneous power of the interconnection line between the microgrid and the main grid. The signal preprocessing module uses mean filtering for noise reduction, averaging multiple consecutive sets of signals from the same source to reduce signal fluctuations caused by electromagnetic interference. The outlier removal process of the signal preprocessing module is achieved by determining whether the amplitude of the acquired signal exceeds the normal operating range of the corresponding physical quantity. The outlier determination is based on the rated parameters and physical characteristics of the equipment. The data transmission module uses industrial Ethernet to transmit the pre-processed digital signal in a standardized format to the photovoltaic consumption priority dynamic determination unit in real time.

[0008] As a further improvement to this technical solution, the photovoltaic grid integration priority dynamic determination unit includes a trend prediction module, a confidence correction module, a scenario adaptation module, a weight constraint module, and a priority calculation module, wherein: The trend prediction module is used to receive preprocessed data output by the microgrid status real-time sensing unit, perform short-term trend prediction of photovoltaic output, remaining energy storage capacity and charging / swapping load and output the predicted value. The confidence correction module is used to receive the predicted value from the trend prediction module and the measured value from the preprocessed data, and generate a confidence factor through error calculation to provide a basis for correcting the priority results; The scenario adaptation module distinguishes microgrid operation scenarios based on the dynamic correlation between trend prediction values ​​and measured values, and dynamically allocates the weights of core indicators such as photovoltaic output fluctuation characteristics, remaining energy storage capacity, and charging and swapping load demand intensity according to scenario characteristics. The weight constraint module achieves weight smoothing through first-order inertial filtering and monitors the weight convergence state through a state timer to avoid sudden weight changes affecting the stability of the judgment. The priority calculation module is used to integrate the confidence factor, the weights of the core indicators after smooth convergence, and the corresponding indicator representation values ​​to calculate and generate the corrected photovoltaic consumption priority results.

[0009] As a further improvement to this technical solution, the trend prediction module performs short-term trend prediction and outputs the predicted value, including the following steps: S20.1 Receive the preprocessed data output by the microgrid status real-time sensing unit, extract the photovoltaic module voltage and current, the charging and swapping equipment load current, the energy storage battery terminal voltage and current signals, and establish a multi-physical quantity cross-correlation model: respectively construct the physical correlation model between photovoltaic power output and light intensity, module temperature, the dynamic mapping relationship between energy storage remaining capacity and charging and discharging current and internal resistance, and the coupling model between charging and swapping load and equipment operating status. S20.2 Select a sliding data window of a preset length and extract the measured values ​​of the core physical quantities within the window. (Corresponding to the measured values ​​of photovoltaic module voltage and current, energy storage battery terminal voltage and current, and charging / swapping equipment load current) and related physical quantities. (Corresponding to light intensity, component temperature, and equipment operating status parameters); S20.3 Adaptively determine dynamic coefficients based on the coupling strength of each physical quantity The result is obtained through sliding window fusion calculation. Time prediction value ,in Corresponding to the predicted photovoltaic power output values Predicted remaining energy storage capacity Or the predicted value of charging and swapping load ;in satisfy ; S20.4, will , , Output to the confidence correction module.

[0010] As a further improvement to this technical solution, the scenario adaptation module receives the photovoltaic power output prediction value output by the trend prediction module. Predicted remaining energy storage capacity Forecast values ​​of charging and swapping load and the power of the tie line in the preprocessed data And retrieve the rated power of the charging and swapping equipment. Transmission limits of tie lines Microgrid operation scenarios are divided based on multi-physical quantity coupling criteria, specifically including: when and When the battery is not fully charged, it is considered a scenario with surplus photovoltaic energy. when and The ratio reaches the peak load threshold set based on the equipment operating characteristics, and Less than When the preset ratio is reached, it is determined to be a peak charging / swapping load scenario; when and The ratio reaches the power grid constraint threshold set based on tie-line transmission characteristics, and Greater than When the preset ratio is reached, it is determined to be a power grid interaction constraint scenario; The scenario adaptation module dynamically allocates the weights of core indicators based on the energy flow characteristics of each scenario. ,in Weights for photovoltaic power output fluctuation characteristics, Weighting of remaining energy storage capacity Weighting the demand intensity of charging and swapping loads, and satisfying In scenarios where there is a surplus of photovoltaic energy, the focus is on photovoltaic energy consumption. To achieve the highest proportion, during peak charging and battery swapping load scenarios, the focus is on load assurance. With the highest proportion, energy storage regulation is the core in grid interaction constraint scenarios. To achieve the highest percentage, the dynamic weight will be used in the final calculation. , , Output to the weight constraint module.

[0011] As a further improvement to this technical solution, the confidence correction module generates the confidence factor by including the following steps: S21.1 Receive the predicted values ​​of each physical quantity output by the trend prediction module. Synchronously retrieve the measured values ​​of the corresponding physical quantities output by the real-time state sensing unit of the microgrid. Select a sliding error calculation window of a preset length; where, For indexing historical moments within the window, and ; S21.2 Calculate the mean absolute error (MAE) between the predicted and measured values ​​within the window using error statistics methods; S21.3 Obtain the rated operating range of the corresponding physical quantity Calculate the reasonable operating range span of the corresponding physical quantity. And adaptively adjust the correction coefficient based on the fluctuation characteristics of physical quantities. (The greater the fluctuation range of the physical quantity,) The closer to 1, the smaller the fluctuation range. (The closer to 0.5) S21.4. Convert MAE into a confidence factor in the 0-1 interval using a dynamic normalization method. ,Will Output to the priority calculation module.

[0012] As a further improvement to this technical solution, the weight constraint module implements weight smoothing and convergence determination through the following steps: S23.1, Receive the output of the scene adaptation module Calculate weights in real time retrieval Time-based historical weighting Calculate the change in weights ; S23.2, Set the basic filter coefficients based on the microgrid response characteristics. and the maximum allowable change in weight (Determined based on the response speed of microgrid devices), the result is obtained through correlation calculation between the weight change and the maximum allowable change. Time-adaptive filter coefficients ; S23.3, based on , and Perform weighted smoothing calculation to obtain Final weight at time ; S23.4 Start the state timer and calculate the weight change rate. When the rate of change is lower than the convergence threshold set based on the circuit response characteristics within a preset number of consecutive data periods, the weight is determined to have reached a convergence state, and the current weight is locked. The result is then output to the priority calculation module; if convergence is not achieved, steps S23.1 to S23.3 are repeated.

[0013] As a further improvement to this technical solution, the priority calculation module generates photovoltaic power consumption priority results through the following steps: S24.1 Receive the confidence factor output by the confidence correction module. The converged core indicator weights output by the weight constraint module , , The original physical quantities of photovoltaic power output fluctuations, remaining energy storage capacity, and charging / swapping load demand intensity output by the microgrid status real-time sensing unit. S24.2. Normalize the above original physical quantities based on their respective physical extreme values ​​to obtain the photovoltaic power output fluctuation characterization values. Energy storage remaining capacity characterization value Characteristic value of charging and swapping load demand intensity ,and All are in the 0-1 range; S24.3 Calculate the predicted photovoltaic output value With the current discharge power of energy storage The sum of the two, and the predicted charging and swapping load. Perform a difference comparison; if Set the energy balance correction coefficient ;otherwise, The value is adaptively adjusted according to the energy gap; S24.4. Photovoltaic grid integration priority results are obtained through deep fusion calculation of multi-dimensional factors. The result is then transmitted to the multi-mode control unit of the charging and swapping microgrid.

[0014] As a further improvement to this technical solution, the charging and swapping microgrid multimodal control unit includes a mode determination module, a multimodal control module, and an instruction output module, wherein: The mode determination module receives the photovoltaic consumption priority result Priority output by the photovoltaic consumption priority dynamic determination unit. Based on the physical balance relationship between photovoltaic consumption demand, charging and swapping load demand and grid interaction constraints reflected by the photovoltaic consumption priority result Priority, it determines the corresponding control mode triggering conditions. The multimodal control module switches to the corresponding control mode according to the triggering conditions, specifically including: When Priority primarily reflects the demand for photovoltaic (PV) grid integration, the system switches to PV priority integration control mode. By adjusting the charging and discharging power of energy storage and optimizing the access sequence of charging and swapping equipment, the system maximizes the integration of PV output. When Priority primarily reflects the charging and swapping load demand, the system switches to the charging and swapping load guarantee control mode. This mode stabilizes the energy storage discharge output and dynamically adjusts the tie-line power supplement to meet the rated operating power requirements of the charging and swapping equipment. When Priority primarily reflects grid interaction constraints, the system switches to grid interaction stability control mode. By suppressing photovoltaic output fluctuations and adjusting energy storage to smooth power, the power fluctuations of the tie line are controlled within the transmission limit. The instruction output module standardizes and processes the power adjustment instructions and equipment operation status adjustment instructions under each control mode, and then outputs them to the flexible scheduling unit for charging and swapping resources.

[0015] As a further improvement to this technical solution, the flexible scheduling unit for charging and swapping resources includes an instruction parsing module, a resource collaborative scheduling module, and an execution feedback module, wherein: The instruction parsing module receives standardized control instructions output by the multi-mode control unit of the charging and swapping microgrid, and splits them into power adjustment instructions and equipment operation status adjustment instructions according to the instruction type. It also parses the target parameter range and response time limit corresponding to the instruction. The resource collaborative scheduling module is based on the priority guidance of photovoltaic consumption and combines the real-time operation status of the microgrid to dynamically adjust the output power of charging and swapping equipment, the control parameters of the energy storage battery charging and discharging circuit (including converter modulation parameters and charging and discharging current constraint thresholds), and optimize the access sequence of charging and swapping loads. The execution feedback module feeds back the real-time operating parameters (including actual power output and circuit status) of the charging and swapping equipment and energy storage battery to the multi-mode control unit of the charging and swapping microgrid, forming a control closed loop.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention, through a scenario adaptation module, accurately distinguishes three microgrid operation scenarios—photovoltaic energy surplus, peak charging / swapping load, and grid interaction constraints—based on a combination of multiple physical quantities, including photovoltaic output prediction, energy storage remaining capacity prediction, charging / swapping load prediction, and the ratio of tie line power to transmission limit. Simultaneously, the charging / swapping microgrid multi-mode control unit can switch to three control modes: photovoltaic priority consumption, charging / swapping load guarantee, and grid interaction stability. This solves the problems of insufficient charging / swapping load adaptability and lack of dynamic response across multiple scenarios, and can specifically guarantee the rated operating power demand during peak charging / swapping load periods, reducing photovoltaic output waste and curtailment. 2. This invention uses a confidence correction module to calculate the average absolute error between predicted and measured values ​​of photovoltaic output, remaining energy storage capacity, and charging / swapping load using a sliding window error estimation. It also adaptively adjusts correction coefficients based on the fluctuation characteristics of physical quantities to generate a confidence factor in the 0-1 range. Simultaneously, a weight constraint module uses a first-order inertial filter to calculate adaptive filtering coefficients for weight smoothing, and a state timer monitors the weight convergence state to prevent abrupt changes. This, combined with a flexible scheduling unit for charging / swapping resources, adjusts the power output of charging / swapping equipment and the control parameters of the energy storage battery charging / discharging circuit (including converter modulation parameters and charging / discharging current constraint thresholds). This solves the problems of prediction errors affecting control accuracy and insufficient flexibility in charging / swapping resource scheduling, improving the accuracy of photovoltaic consumption priority determination and the flexibility of charging / swapping resource scheduling. 3. This invention, through an execution feedback module, feeds back real-time operating parameters such as the actual power output of the charging and swapping equipment and the status of the energy storage battery circuit to the multi-mode control unit of the charging and swapping microgrid, forming a complete control closed loop. Furthermore, under grid interaction constraints, the multi-mode control unit of the charging and swapping microgrid controls the power fluctuations of the tie line within the transmission limit range by suppressing photovoltaic power output fluctuations and adjusting the energy storage smoothing power. This solves the problems of missing control closed loops and insufficient grid interaction stability control, ensuring the stability of the power of the tie line between the microgrid and the main grid, and avoiding the impact of photovoltaic power output fluctuations on the main grid. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the system framework of the present invention; The meanings of the labels in the diagram are as follows: 1. Microgrid status real-time sensing unit; 10. Distributed acquisition module; 11. Signal preprocessing module; 12. Data transmission module; 2. Photovoltaic grid integration priority dynamic determination unit; 20. Trend prediction module; 21. Confidence correction module; 22. Scenario adaptation module; 23. Weight constraint module; 24. Priority calculation module; 3. Multimodal control unit for charging and swapping microgrid; 30. Mode determination module; 31. Multimodal control module; 32. Command output module; 4. Flexible scheduling unit for charging and swapping resources; 40. Instruction parsing module; 41. Resource collaborative scheduling module; 42. Execution feedback module. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] like Figure 1 As shown, this embodiment provides a multi-modal flexible control system for charging and swapping microgrids that integrates photovoltaic power consumption priorities, including: The real-time sensing unit 1 of the microgrid status collects the voltage and current of photovoltaic modules, the load current of charging and swapping equipment, the voltage and current of energy storage batteries, and the power signal of the interconnection line between the microgrid and the main grid in real time. It performs noise reduction and outlier removal preprocessing on the collected signals and transmits the preprocessed data to the dynamic determination unit 2 of photovoltaic consumption priority. The real-time microgrid status sensing unit 1, as the core of data acquisition and preprocessing in this system, undertakes the real-time sensing of multiple physical quantities within the charging and swapping microgrid, providing accurate and reliable basic data support for subsequent photovoltaic consumption priority determination and multi-modal control. Through the distributed deployment of multiple types of sensors, intelligent signal preprocessing, and high-speed data transmission, it achieves panoramic acquisition and standardized output of status information from photovoltaic modules, charging and swapping equipment, energy storage batteries, and main grid interconnection lines.

[0020] In this embodiment, the microgrid status real-time sensing unit 1 includes a distributed acquisition module 10, a signal preprocessing module 11, and a data transmission module 12, wherein: The distributed acquisition module 10 consists of a voltage sensor, a current sensor, and a power sensor. The voltage sensor is used to acquire the terminal voltage of the photovoltaic module and the terminal voltage of the energy storage battery. The current sensor is used to acquire the load current of the charging and swapping equipment and the charging and discharging current of the energy storage battery. The power sensor is used to acquire the instantaneous power of the interconnection line between the microgrid and the main grid. Specifically, the distributed acquisition module 10 adopts a "point-to-surface" sensor deployment strategy. Targeting the physical topology of the charging and swapping microgrid, voltage sensors are configured at the photovoltaic array combiner box to collect the terminal voltage of each string of photovoltaic modules in real time, monitoring the power generation uniformity of the photovoltaic array. Current sensors are configured on the DC input side of the charging and swapping equipment and in the charging and discharging circuit of the energy storage battery to accurately capture the dynamic changes in the charging and swapping load current and the energy storage charging and discharging current. Power sensors are configured at the connection line between the microgrid and the main grid to obtain the bidirectional power flow status in real time. The sampling frequency of all sensors is uniformly set to 100Hz to ensure that the high-frequency dynamic characteristics of physical quantities are fully captured.

[0021] The signal preprocessing module 11 uses mean filtering to reduce noise by averaging multiple consecutive sets of signals from the same source, thereby reducing signal fluctuations caused by electromagnetic interference. The outlier removal process of the signal preprocessing module 11 is achieved by judging whether the amplitude of the acquired signal exceeds the normal operating range of the corresponding physical quantity. The outlier determination is based on the rated parameters and physical characteristics of the equipment. Specifically, the signal preprocessing module 11 uses a two-step process of "noise reduction - outlier removal" to purify the acquired signal, providing high-quality input for subsequent data applications, as detailed below: Mean filtering noise reduction: To address high-frequency signal fluctuations caused by electromagnetic interference, a sliding window mean filtering method is used to average multiple consecutive sets of signals from the same source, effectively reducing signal glitches caused by electromagnetic radiation and sensor noise, and making signal fluctuations conform to the actual trend of physical quantity changes. Outlier Removal: Anomaly detection threshold ranges are constructed based on equipment rated parameters and physical characteristics. For example, the normal voltage range for photovoltaic modules is set to 80%~120% of their rated open-circuit voltage; the abnormal load current range for charging / swapping equipment is ±20% of its rated current; the abnormal voltage range for energy storage batteries is based on their SOC (State of Charge)-voltage curve. When the SOC is in the 20%~80% range, a voltage deviation of ±5% from the corresponding voltage value within that range is considered abnormal. For signal values ​​determined to be abnormal, the average of the previous three valid signals is used to replace them, ensuring data continuity.

[0022] The data transmission module 12 uses industrial Ethernet to transmit the pre-processed digital signal in a standardized format to the photovoltaic consumption priority dynamic determination unit 2 in real time.

[0023] Specifically, the data transmission module 12 constructs a high-speed data transmission channel based on industrial Ethernet, encapsulating the preprocessed digital signal into standardized data frames (containing fields such as timestamp, physical quantity type, value, and status identifier). The transmission rate is set to 100Mbps, and the transmission delay is ≤50ms. The data frame format follows the communication protocol specifications of this system, with the physical quantity type using coded identifiers to ensure that the photovoltaic consumption priority dynamic determination unit 2 can quickly parse and extract valid information. Simultaneously, the data transmission module 12 has a data caching function; when network congestion occurs momentarily, it can cache up to 100 sets of data frames to avoid data loss and ensure the continuity of analysis in subsequent units.

[0024] The photovoltaic consumption priority dynamic determination unit 2 receives preprocessed data and performs short-term trend predictions on photovoltaic output, remaining energy storage capacity, and charging / swapping load. It generates a prediction confidence factor through sliding window error estimation, applies smooth convergence constraints to the weight adjustment by combining first-order inertial filtering and state timer, dynamically distinguishes microgrid operation scenarios and adapts the weight ratio of core indicators, generates a photovoltaic consumption priority result with confidence correction, and transmits it to the charging / swapping microgrid multi-mode control unit 3. The photovoltaic consumption priority dynamic determination unit 2 receives standardized data transmitted by the microgrid status real-time sensing unit 1. Through the entire process of trend prediction, confidence correction, weight constraint, scenario adaptation and priority calculation, it realizes accurate identification of the charging and swapping microgrid operation scenario and dynamic determination of photovoltaic consumption priority, providing a decision basis for subsequent multimodal control.

[0025] To ensure the real-time and stable operation of the trend prediction module 20, confidence correction module 21, weight constraint module 23, scenario adaptation module 22, and priority calculation module 24 of the photovoltaic power consumption priority dynamic determination unit 2, and to achieve accurate prediction and dynamic priority determination of photovoltaic output, energy storage capacity, and load, this embodiment also provides the following hardware configuration example: Core processor: It can adopt a quad-core industrial-grade processor with ARM Cortex-A72 architecture and a main frequency of 1.8GHz. It is used to process algorithms such as dual physical quantity coupling calculation of trend prediction module 20, sliding window error statistics of confidence correction module 21, and multi-dimensional factor fusion of priority calculation module 24 in parallel, to ensure microsecond-level response of instructions of each module. Storage unit: Configurable with 4GB DDR4 memory and 128GB industrial-grade solid-state drive. The memory is used to cache real-time data streams such as photovoltaic power output, energy storage SOC, and load demand, providing data support for the trend prediction module 20 and the scenario adaptation module 22. The solid-state drive is used to store historical prediction model parameters, scenario judgment rule base, weight allocation strategy table, etc., to ensure historical data backtracking and strategy invocation for the weight constraint module 23 and the priority calculation module 24. Communication interfaces: It can be equipped with two gigabit Ethernet ports, four RS485 serial ports, and a CAN bus interface. The Ethernet is used for high-speed data interaction with the microgrid status real-time sensing unit 1 and the charging and swapping microgrid multimodal control unit 3 to transmit the judgment results of the priority calculation module 24; the RS485 and CAN bus are used for local device communication with the confidence correction module 21 and the scene adaptation module 22 to ensure reliable transmission of commands and data. Hardware acceleration module: It can integrate an FPGA coprocessor to accelerate the dynamic coefficient, adaptive calculation, and multi-dimensional factor fusion algorithm of the priority calculation module 24 of the trend prediction module 20, improve the execution efficiency of complex calculations, and meet the real-time determination requirements of photovoltaic consumption priority during peak charging periods. Input / output module: It can be configured with 16 analog inputs (16-bit resolution) and 8 digital input / output interfaces. The analog interfaces are used to collect raw physical quantities such as photovoltaic voltage and current, and energy storage terminal voltage and current, to provide data input for the trend prediction module 20. The digital interfaces are used to trigger control signals such as scene switching and weight convergence determination, so as to realize the hardware-level status monitoring and control of the scene adaptation module 22 and the weight constraint module 23.

[0026] In this embodiment, the photovoltaic grid integration priority dynamic determination unit 2 includes a trend prediction module 20, a confidence correction module 21, a scenario adaptation module 22, a weight constraint module 23, and a priority calculation module 24, wherein: In this embodiment, the trend prediction module 20 is used to receive preprocessed data output by the microgrid status real-time sensing unit 1, perform short-term trend prediction on photovoltaic output, remaining energy storage capacity, and charging / swapping load, and output predicted values; the trend prediction module 20 performs short-term trend prediction and outputs predicted values, including the following steps: S20.1 Receive the preprocessed data output by the microgrid status real-time sensing unit 1, extract the photovoltaic module voltage and current, the charging and swapping equipment load current, the energy storage battery terminal voltage and current signals, and establish a multi-physical quantity cross-correlation model: respectively construct the physical correlation model between photovoltaic power output and light intensity, module temperature, the dynamic mapping relationship between energy storage remaining capacity and charging and discharging current and internal resistance, and the coupling model between charging and swapping load and equipment operating status; S20.2 Select a sliding data window of a preset length and extract the measured values ​​of the core physical quantities within the window. (Corresponding to the measured values ​​of photovoltaic module voltage and current, energy storage battery terminal voltage and current, and charging / swapping equipment load current) and related physical quantities. (Corresponding to light intensity, component temperature, and equipment operating status parameters); S20.3 Adaptively determine dynamic coefficients based on the coupling strength of each physical quantity The result is obtained through sliding window fusion calculation. Time prediction value ,in Corresponding to the predicted photovoltaic power output values Predicted remaining energy storage capacity Or the predicted value of charging and swapping load ;in Full ; S20.4, will , , Output to confidence correction module 21.

[0027] Specifically, the trend prediction module 20 constructs a multi-physical quantity cross-correlation model based on preprocessed data to achieve short-term trend prediction of photovoltaic output, remaining energy storage capacity, and charging / swapping load, as detailed below: First, the preprocessed data output by the real-time microgrid status sensing unit 1 is received, and core physical quantities, including photovoltaic module voltage and current, are extracted. Load current of charging and swapping equipment Energy storage battery terminal voltage and current Simultaneously acquire the associated physical quantity, light intensity. Component temperature Operating status parameters of charging and swapping equipment ; Subsequently, a multi-physical quantity cross-correlation model was constructed: for photovoltaic power output prediction, a model was established. and The coupling model—through Calculate real-time photovoltaic power, combined with The trend of radiation intensity change The impact of factors on module conversion efficiency is investigated, establishing a dynamic correlation between photovoltaic power output and environmental factors; for predicting remaining energy storage capacity, a system is established. The dynamic mapping relationship between (terminal voltage, charging / discharging current) and energy storage internal resistance—through Historical charge and discharge data are used to fit the correlation curve between charge and discharge current and internal resistance changes, enabling dynamic estimation of remaining capacity; for charging and swapping load prediction, a system is established... and A coupled model of (equipment operating status and vehicle battery swapping sequence) – based on Historical fluctuation characteristics, combined with It reflects the start-up and shutdown status of equipment, the peak arrival patterns of vehicles, and captures load change trends; Next, the preset length of the sliding data window was set to 15 minutes (matching the regular scheduling cycle of the microgrid), and the sampling frequency was set to 1 minute / group. The measured values ​​of the core physical quantities within the window were then extracted. ( ) and measured values ​​of related physical quantities ( );right and Perform time-weighted aggregation separately, giving more recent data higher weight (weight coefficient). satisfy And the closer the time, the greater the weight, for example ), to obtain scalar ;in This represents the time-weighted aggregate value of the core physical quantities within the sliding data window; This represents the time-weighted aggregate value of associated physical quantities within the sliding data window; The dynamic coefficients are adaptively determined based on the coupling strength of each physical quantity. The coupling strength is determined by the Pearson correlation coefficient. Quantification, the formula is: ; A larger value indicates a stronger coupling between the core physical quantity and related physical quantities; simultaneously, the dynamic coefficient... Based on coupling strength and prediction error Adaptive adjustment, the formula is: ; ; in , For learning rate (e.g.) (Set according to the microgrid's operational stability requirements); photovoltaic output prediction is taken as... (This is the coefficient convergence value under typical operating conditions. Based on the strong coupling between photovoltaic module voltage and current and light and temperature, the core physical quantities have a slightly higher weight.) The predicted remaining energy storage capacity is taken from... (This is the coefficient convergence value under typical operating conditions. Based on the strong coupling between the voltage, current, and internal resistance of the energy storage terminal, the core physical quantities have a higher weight.) The charging / swapping load prediction is taken as follows: (This is the coefficient convergence value under typical operating conditions, with weighted balance based on the moderate coupling between charging / swapping load current, equipment status, and vehicle timing.) This is achieved through the formula... ,calculate Time prediction value ; Finally, , , The output is synchronously sent to the confidence correction module 21.

[0028] In the prediction of photovoltaic output, energy storage capacity, and load in photovoltaic charging and swapping microgrids, traditional technologies either rely on modeling a single physical quantity (such as using only photovoltaic voltage and current to predict output) or use fixed coefficients to integrate multiple quantities. This makes it difficult to adapt to the dynamic needs of "accurately matching charging load fluctuations in the short term and supporting energy storage planning trends in the medium and long term." During peak charging periods, relying on a single quantity for prediction can lead to accuracy gaps due to insufficient coupling, while relying on fixed coefficients during energy storage planning can result in biased predictions due to changes in the scenario. The trend prediction module 20 in this embodiment adopts the method of "dual physical quantity coupling (core physical quantity + related physical quantity) + dynamic coefficient self-adaptation (based on coupling strength adjustment)" to allow the two physical quantities to focus on prediction needs in different dimensions. Furthermore, the dynamic coefficient acts like an "intelligent adjuster," automatically adjusting the ratio of the fusion coefficients of the two physical quantities according to the coupling strength of the physical quantities. Its core function is to increase the weight of core physical quantities during the midday charging peak, ensuring the accuracy of real-time power supply prediction for the charging pile; and to adapt and increase the weight of related physical quantities during nighttime energy storage planning, providing an all-day trend reference for the energy storage system. At the same time, the independent modeling and dynamic fusion of the two physical quantities also reduces the risk of prediction collapse caused by sudden environmental changes (such as rapid cloud cover) of a single physical quantity, improving the robustness of the prediction system and ensuring that the prediction of photovoltaic-based charging and swapping microgrids has accurate data support at different times.

[0029] In this embodiment, the confidence correction module 21 receives the predicted value from the trend prediction module 20 and the measured value from the preprocessed data, and generates a confidence factor through error calculation to provide a basis for correcting the priority results. The confidence correction module 21 generates the confidence factor through the following steps: S21.1 Receive the predicted values ​​of each physical quantity output by the trend prediction module 20 Synchronously retrieve the measured values ​​of the corresponding physical quantities output by the real-time sensing unit 1 of the microgrid status. Select a sliding error calculation window of a preset length; where, For indexing historical moments within the window, and ; S21.2 Calculate the mean absolute error (MAE) between the predicted and measured values ​​within the window using error statistics methods; S21.3 Obtain the rated operating range of the corresponding physical quantity Calculate the reasonable operating range span of the corresponding physical quantity. And adaptively adjust the correction coefficient based on the fluctuation characteristics of physical quantities. (The greater the fluctuation range of the physical quantity,) The closer to 1, the smaller the fluctuation range. (The closer to 0.5) S21.4. Convert MAE into a confidence factor in the 0-1 interval using a dynamic normalization method. ,Will Output to priority calculation module 24.

[0030] Specifically, the confidence correction module 21 generates a confidence factor for the predicted value through sliding window error estimation and dynamic normalization, providing a basis for correction in priority calculation, as follows: First, receive the predicted values ​​of each physical quantity output by the trend prediction module 20. Synchronously retrieve the measured values ​​of the corresponding physical quantities output by the real-time sensing unit 1 of the microgrid status. A sliding error calculation window with a preset length of 5 scheduling cycles (75 minutes) is selected; among which, For indexing historical moments within the window, and ; Then, within the window, the mean absolute error (MAE) between the predicted and measured values ​​is calculated using error statistics methods, using the following formula: ; in For the first Predicted value at any time For the first Real-time measured values, respectively for , , Calculate their respective MAE; Next, obtain the rated operating range of the corresponding physical quantity. Calculate the reasonable operating range span Adaptive adjustment of correction coefficients based on the fluctuation characteristics of physical quantities (The greater the fluctuation range of the physical quantity,) The closer to 1, the smaller the fluctuation range. The closer to 0.5, the larger the fluctuation in photovoltaic output. (This value is based on the characteristic that photovoltaic output fluctuates greatly due to environmental factors (sunlight, temperature); the remaining capacity of energy storage fluctuates less, so it is taken as...) (This value is chosen based on the characteristic that the remaining energy storage capacity fluctuates little under the constraints of the dispatch strategy); the fluctuation range of charging and swapping load is moderate, so a value of [value missing] is chosen. (This value is based on the characteristic that the charging and swapping load fluctuates moderately due to the randomness of vehicle arrivals.) Finally, the confidence factor is calculated using the following formula. ( ): ; and the corresponding physical quantities Output to priority calculation module 24.

[0031] In the prediction confidence correction of photovoltaic charging and swapping microgrids, traditional technologies either use fixed coefficient correction (such as uniformly using a coefficient of 0.8) or ignore the fluctuation characteristics of the physical quantities themselves, making it difficult to adapt to the differentiated needs of "large fluctuations in photovoltaic output and small fluctuations in energy storage capacity"—using fixed coefficients in photovoltaic output prediction will lead to confidence deviation due to large fluctuations, while in energy storage capacity prediction, it will lead to overcorrection due to small fluctuations. The confidence correction module 21 in this embodiment adopts "sliding window error statistics + adaptive correction coefficient based on physical quantity fluctuation characteristics". This approach allows error calculation and coefficient adjustment to focus on the characteristics of different physical quantities, and further utilizes adaptive coefficients. Like a "smart calibrator," it adjusts based on the amplitude of fluctuations in physical quantities (the greater the fluctuation...). The closer to 1, the smaller the value (closer to 0.5), the automatic adjustment of the correction intensity. Its core function is to enable accurate prediction of photovoltaic power output. Improved adaptation ensures accurate correction of prediction confidence during peak charging periods; energy storage capacity prediction... The reduced adaptability provides a reasonable confidence reference for nighttime energy storage planning. At the same time, the combination of sliding window and adaptive coefficient also reduces the risk of confidence distortion caused by changes in the scenario (such as alternating sunny and rainy weather) due to a single correction logic, improves the adaptability of confidence correction, and provides reliable confidence support for the calculation of photovoltaic consumption priority.

[0032] In this embodiment, the scenario adaptation module 22 distinguishes microgrid operation scenarios based on the dynamic correlation between trend prediction values ​​and measured values, and dynamically allocates the weights of core indicators such as photovoltaic output fluctuation characteristics, remaining energy storage capacity, and charging / swapping load demand intensity according to scenario characteristics; the scenario adaptation module 22 receives the photovoltaic output prediction values ​​output by the trend prediction module 20. Predicted remaining energy storage capacity Forecast values ​​of charging and swapping load and the power of the tie line in the preprocessed data And retrieve the rated power of the charging and swapping equipment. Transmission limits of tie lines Microgrid operation scenarios are divided based on multi-physical quantity coupling criteria, specifically including: when and When the battery is not fully charged, it is considered a scenario with surplus photovoltaic energy. when and The ratio reaches the peak load threshold set based on the equipment operating characteristics, and Less than When the preset ratio is reached, it is determined to be a peak charging / swapping load scenario; when and The ratio reaches the power grid constraint threshold set based on tie-line transmission characteristics, and Greater than When the preset ratio is reached, it is determined to be a power grid interaction constraint scenario; The scene adaptation module 22 dynamically allocates the weights of core indicators based on the energy flow characteristics of each scene. ,in Weights for photovoltaic power output fluctuation characteristics, Weighting of remaining energy storage capacity Weighting the demand intensity of charging and swapping loads, and satisfying In scenarios where there is a surplus of photovoltaic energy, the focus is on photovoltaic energy consumption. To achieve the highest proportion, during peak charging and battery swapping load scenarios, the focus is on load assurance. With the highest proportion, energy storage regulation is the core in grid interaction constraint scenarios. To achieve the highest percentage, the dynamic weight will be used in the final calculation. , , Output to weight constraint module 23.

[0033] Specifically, the scene adaptation module 22 is based on dynamic scene segmentation using multi-physical quantity coupling criteria, plus scene-driven core indicator weight allocation, breaking through the limitations of conventional fixed scenes or fixed weights, as detailed below: First, the photovoltaic output prediction value is received from the trend prediction module 20. Predicted remaining energy storage capacity Forecast values ​​of charging and swapping load and the power of the tie line output by the microgrid status real-time sensing unit 1 At the same time, retrieve the rated power of the charging and swapping equipment. Transmission limits of tie lines ; Then, the operating scenarios are divided based on the multi-physical quantity coupling criterion: when and (When the energy storage is not fully charged but has surplus energy storage capacity), it is determined to be a photovoltaic energy surplus scenario; when (Based on the full-load operation characteristics of charging and swapping equipment, 80% is the peak threshold) and When the photovoltaic output cannot meet the preset proportion of load demand, it is determined to be a peak charging and swapping load scenario; when (Based on the secure transmission characteristics of tie lines, 85% is the power grid constraint threshold) and When the preset ratio of the impact of photovoltaic output on the tie line is reached, it is determined to be a grid interaction constraint scenario; Finally, the weights of core metrics are dynamically allocated based on the characteristics of the scenario. Photovoltaic energy surplus scenarios (In scenarios with surplus photovoltaic energy, the allocation weight is based on photovoltaic consumption as the core), while in peak charging and swapping load scenarios, the allocation weight is... (With load guarantee as the core, the weight of increasing the demand intensity of charging and battery swapping loads is increased), taking the power grid interaction constraint scenario as... (With energy storage regulation as the core, strengthen the adaptation weight of the remaining energy storage capacity to the grid interaction constraints), and output the dynamic weight to the weight constraint module 23.

[0034] In the scenario and weight adaptation of photovoltaic charging and swapping microgrids, traditional technologies use fixed weight allocation, which is difficult to adapt to the multi-scenario needs of "PV surplus, peak load, and grid constraints"—fixed weights waste absorption potential when PV is in surplus, and lead to insufficient PV absorption during peak load. The scenario adaptation module 22 in this embodiment uses a method of "multi-scenario coupling criteria + dynamic core indicator weights and allocation," allowing scenario determination and weight allocation to focus on different operational needs. Furthermore, through dynamic weights, it acts like a "smart allocator," automatically adjusting the weight ratio of PV, energy storage, and load according to the scenario type (PV surplus, peak load, grid constraints). Its core function is to adapt the PV to various scenarios, including those with PV surplus. Adaptation and enhancement to maximize midday photovoltaic power consumption; during peak load scenarios Improved compatibility to ensure charging needs are met; in scenarios constrained by the power grid. The improved adaptation and stable grid interaction, combined with the combination of multiple scenarios and dynamic weights, also reduce the risk of strategy failure due to misjudgment of a single weight, improve the accuracy of scenario adaptation, and provide scenario-based weight support for the calculation of photovoltaic consumption priority.

[0035] In this embodiment, the weight constraint module 23 achieves weight smoothing through first-order inertial filtering and monitors the weight convergence state through a state timer to avoid sudden weight changes affecting the stability of the determination. The weight constraint module 23 implements weight smoothing and convergence determination through the following steps: S23.1, Receive the output of the scene adaptation module 22 Calculate weights in real time retrieval Time-based historical weighting Calculate the change in weights ; S23.2, Set the basic filter coefficients based on the microgrid response characteristics. and the maximum allowable change in weight (Determined based on the response speed of microgrid devices), the result is obtained through correlation calculation between the weight change and the maximum allowable change. Time-adaptive filter coefficients ; S23.3, based on , and Perform weighted smoothing calculation to obtain Final weight at time ; S23.4 Start the state timer and calculate the weight change rate. When the rate of change is lower than the convergence threshold set based on the circuit response characteristics within a preset number of consecutive data periods, the weight is determined to have reached a convergence state, and the current weight is locked. The result is output to the priority calculation module 24; if convergence is not achieved, steps S23.1 to S23.3 are repeated.

[0036] Specifically, the weight constraint module 23 solves the problem of control strategy instability caused by conventional weight abrupt changes through a dual mechanism of weight smoothing by first-order inertial filtering and convergence monitoring by a state timer, as follows: First, receive the output from the scene adaptation module 22. Calculate weights in real time (Corresponding to the characteristic weight of photovoltaic power output fluctuation) Remaining energy storage capacity weight Weight of charging and swapping load demand intensity ), retrieve Time-based historical weighting Calculate the change in weights ; Then, set the basic filter coefficients. Maximum allowable change in weight Calculated using the following formula Time-adaptive filter coefficients ( ): ; Next, based on , and Calculated using the following formula Final weight at time : ; Finally, start the state timer and calculate the weight change rate. A convergence threshold of 0.05 is set (based on the stability requirements of the microgrid control strategy and the circuit response characteristics, to ensure the smoothness and reliability of technical processes such as weight adjustment). When the rate of change is lower than the threshold for three consecutive scheduling cycles, the weights are considered to have converged and are locked. The result is output to the priority calculation module 24; otherwise, the above process is repeated.

[0037] In the weight control of photovoltaic charging and swapping microgrids, traditional technologies either use fixed weight outputs or lack convergence judgment for weight adjustments, making it difficult to meet the requirements of "smooth weight transition and rapid response to scene changes"—sudden weight changes during scene switching can cause control strategy oscillations, while the lack of convergence judgment can lead to long-term weight fluctuations. The weight constraint module 23 in this embodiment uses a combination of "first-order inertial filtering smoothing + state timer convergence monitoring" to ensure both smoothness and convergence in weight adjustments. Furthermore, by using a convergence threshold (e.g., 0.05), it acts like a "smart stabilizer," locking the weight when the weight change rate is below the threshold and dynamically adjusting it when it is above. Its core function is to ensure a smooth weight transition during scene switching, avoiding frequent switching of the charging and swapping microgrid control mode; simultaneously, it quickly locks the weight when it is stable, ensuring the timeliness of photovoltaic consumption priority calculation, reducing the risk of strategy instability due to missing adjustment logic in single weight control, improving the reliability of the weight system, and providing stable weight support for photovoltaic consumption priority calculation in multiple scenarios.

[0038] In this embodiment, the priority calculation module 24 is used to integrate the confidence factor, the smoothed converged core indicator weights, and the corresponding indicator representation values ​​to calculate and generate a corrected photovoltaic grid connection priority result. The priority calculation module 24 generates the photovoltaic grid connection priority result through the following steps: S24.1 Receive the confidence factor output by the confidence correction module 21 The converged core index weights output by the weight constraint module 23 , , The original physical quantities output by the real-time microgrid status sensing unit 1, including photovoltaic power output fluctuations, remaining energy storage capacity, and charging / swapping load demand intensity. S24.2. Normalize the above original physical quantities based on their respective physical extreme values ​​to obtain the photovoltaic power output fluctuation characterization values. Energy storage remaining capacity characterization value Characteristic value of charging and swapping load demand intensity ,and All are in the 0-1 range; S24.3 Calculate the predicted photovoltaic output value With the current discharge power of energy storage The sum of the two, and the predicted charging and swapping load. Perform a difference comparison; if Set the energy balance correction coefficient ;otherwise, The value is adaptively adjusted according to the energy gap; S24.4. Photovoltaic grid integration priority results are obtained through deep fusion calculation of multi-dimensional factors. The result is then transmitted to the charging and swapping microgrid multimodal control unit 3.

[0039] Specifically, the priority calculation module 24 generates a corrected photovoltaic consumption priority result through multi-dimensional factor fusion, as follows: First, the confidence factor output by the confidence correction module 21 is received. The converged core index weights output by the weight constraint module 23 , , And the photovoltaic power output fluctuations output by the microgrid status real-time sensing unit 1 Remaining energy storage capacity Charging and swapping load demand intensity The original physical quantity; Then, the original physical quantities are normalized based on their respective physical extreme values ​​to obtain the photovoltaic power output fluctuation characterization values. Energy storage remaining capacity characterization value Characteristic value of charging and swapping load demand intensity ,and All values ​​fall within the 0-1 range. The normalization formula (taking photovoltaic power output fluctuation as an example) is: ; in, , These represent the minimum and maximum measured values ​​of photovoltaic output, respectively. This represents the measured value of the actual output power of the photovoltaic module; and The same logical normalization is used as described above; Next, the predicted photovoltaic output value is calculated. With the current discharge power of energy storage The sum of the two, and the predicted charging and swapping load. Perform a difference comparison: Based on the current energy storage SOC and rated energy storage power Confirmed, the formula is ( For example, the discharge coefficient corresponding to SOC. , hour (This coefficient is based on the technical specifications of the energy storage converter manufacturer). like Set the energy balance correction coefficient ;otherwise, Values ​​are adaptively adjusted based on the energy gap: When the energy gap accounts for 10%-20%, and When (energy storage and discharge capacity are sufficient), ; When the energy gap accounts for More than 20%, or When (energy storage capacity is insufficient for discharge), ; Finally, the photovoltaic grid integration priority results were obtained through deep fusion calculation of multi-dimensional factors. The formula is: ; in, Indicates the first The core index weights of each physical quantity are output by the weight constraint module 23. These correspond to three physical quantities: photovoltaic power output fluctuation characteristics, remaining energy storage capacity, and charging / swapping load demand intensity, respectively. Indicates the first The normalized representation value of each physical quantity. The original physical quantities (such as photovoltaic output, remaining energy storage capacity, and charging / swapping load demand intensity) output by the real-time microgrid status sensing unit 1 are normalized based on their respective physical extreme values ​​(minimum and maximum measured values), and the value range is... Its function is to unify the magnitude of different physical quantities, so that multi-dimensional factors have comparable contributions in priority calculation. Indicates the first The confidence factor for each physical quantity is output by the confidence correction module 21. The mean absolute error (MAE) between the predicted and measured values ​​is calculated using a sliding window, combined with the rated operating range of the physical quantity. and adaptive correction coefficient The range of values ​​is obtained through dynamic normalization. Used to adjust the confidence level of predicted values. The closer it is to 1, the higher the reliability of the predicted value of the corresponding physical quantity, and the more reliable its contribution to the priority calculation; The result is then transmitted to the charging and swapping microgrid multimodal control unit 3.

[0040] In the priority calculation of photovoltaic charging and swapping microgrids, traditional technologies either rely solely on a single dimension (such as photovoltaic output alone) or ignore the energy balance state, making it difficult to adapt to the multi-objective requirements of "simultaneously considering photovoltaic consumption, load guarantee, and grid constraints." Ignoring the energy gap in photovoltaic consumption calculation leads to infeasibility of the strategy, while the lack of multiple dimensions in load guarantee results in a one-sided priority. The priority calculation module 24 in this embodiment employs a method of "deep fusion of multi-dimensional factors, characterization values, confidence levels + energy balance correction coefficients," allowing multiple factors and energy balance to focus on different decision dimensions, and further... Like a "smart balancer," it automatically adjusts the correction intensity based on the energy difference between photovoltaic, energy storage, and load. Its core function is to balance energy levels when there is a surplus. Adapting to ensure priority for photovoltaic power consumption, during energy gaps The feasibility of the adaptation and protection strategy is ensured. At the same time, the combination of multiple factors and energy balance reduces the risk of priority failure caused by energy imbalance in single-dimensional calculation, improves the rationality of priority calculation, and provides precise priority support for the multi-modal control of charging and swapping microgrids.

[0041] The charging and swapping microgrid multi-mode control unit 3 receives the photovoltaic consumption priority result from the photovoltaic consumption priority dynamic determination unit 2, switches to three control modes corresponding to photovoltaic priority consumption, charging and swapping load guarantee, and grid interaction stability, and outputs control commands to the charging and swapping resource flexible scheduling unit 4. In this embodiment, the charging and swapping microgrid multimodal control unit 3 includes a mode determination module 30, a multimodal control module 31, and an instruction output module 32, wherein: The mode determination module 30 receives the photovoltaic consumption priority result Priority output by the photovoltaic consumption priority dynamic determination unit 2, and determines the corresponding control mode triggering conditions based on the physical balance relationship between photovoltaic consumption demand, charging and swapping load demand and grid interaction constraints reflected by the photovoltaic consumption priority result Priority. Specifically, the mode determination module 30 analyzes the physical balance relationship between photovoltaic consumption demand, charging and swapping load demand, and grid interaction constraints reflected by the photovoltaic consumption priority result Priority through a preset photovoltaic consumption priority range, and determines the control mode triggering conditions: when At that time, it was determined that the demand for photovoltaic (PV) grid connection was the primary factor, triggering the PV priority grid connection control mode; when When the demand for charging and swapping loads is determined to be dominant, the charging and swapping load guarantee control mode is triggered. when When the condition is determined to be dominated by grid interaction constraints, the grid interaction stability control mode is triggered. The mode determination module 30 uses the embedded processor's running range determination algorithm to classify and identify Priority in real time, and outputs the mode trigger signal to the multi-mode control module 31.

[0042] The multimodal control module 31 switches to the corresponding control mode according to the triggering condition, specifically including: When Priority primarily reflects the demand for photovoltaic (PV) grid integration, the system switches to PV priority integration control mode. By adjusting the charging and discharging power of energy storage and optimizing the access sequence of charging and swapping equipment, the system maximizes the integration of PV output. When Priority primarily reflects the charging and swapping load demand, the system switches to the charging and swapping load guarantee control mode. This mode stabilizes the energy storage discharge output and dynamically adjusts the tie-line power supplement to meet the rated operating power requirements of the charging and swapping equipment. When Priority primarily reflects grid interaction constraints, the system switches to grid interaction stability control mode. By suppressing photovoltaic output fluctuations and adjusting energy storage to smooth power, the power fluctuations of the tie line are controlled within the transmission limit. Specifically, the multimodal control module 31 switches to the corresponding control mode and generates control logic based on the mode trigger signal, as follows: Photovoltaic priority consumption control mode: collecting real-time photovoltaic output. and charging / swapping real-time load Calculate the power difference .like Adjust the energy storage system to enter charging mode, and the charging power follows. (Not exceeding the maximum charging power limit of energy storage); at the same time, optimize the access sequence of charging and swapping equipment, prioritize the scheduling of charging and swapping tasks with low power demand, and free up capacity for photovoltaic consumption.

[0043] Charging and swapping load protection control mode: monitoring real-time energy storage discharge power and charging / swapping load demand ,when When the power supply logic of the tie line is triggered, the input power of the tie line is dynamically adjusted to make up for the gap; at the same time, the current discharge power of the energy storage is kept stable, and a constant power control strategy is adopted to meet the rated operating power requirements of the charging and swapping equipment.

[0044] Power grid interactive stability control mode: Real-time power acquisition of tie lines Calculate its relationship with the transmission limit Fluctuation If the fluctuation exceeds the allowable range, the energy storage system outputs a smoothing power (not exceeding the maximum regulating power of the energy storage), while a sliding window filtering algorithm is used to suppress fluctuations in photovoltaic power output, keeping the power fluctuations of the tie line within the transmission limit.

[0045] The instruction output module 32 standardizes and processes the power adjustment instructions and equipment operation status adjustment instructions under each control mode, and then outputs them to the flexible scheduling unit 4 for charging and swapping resources.

[0046] Specifically, the instruction output module 32 uses industry-standard communication protocols (such as Modbus-RTU or OPCUA) to standardize and encapsulate the control logic. The instruction format includes the target device identifier (such as the energy storage converter ID, charging / swapping device ID), the control parameter type (such as power instruction, timing instruction), the target value, and the response time limit. The encapsulated instruction is output to the flexible scheduling unit 4 for charging / swapping resources via industrial Ethernet or RS485 bus.

[0047] The flexible scheduling unit 4 for charging and swapping resources receives control commands from the multi-mode control unit 3 of the charging and swapping microgrid, adjusts the power output of the charging and swapping equipment, the control parameters of the energy storage battery charging and discharging circuit, and the load access sequence, so as to realize the flexible control of the charging and swapping microgrid with photovoltaic consumption priority orientation.

[0048] In this embodiment, the flexible scheduling unit 4 for charging and swapping resources includes an instruction parsing module 40, a resource collaborative scheduling module 41, and an execution feedback module 42, wherein: The instruction parsing module 40 receives standardized control instructions output by the multi-mode control unit 3 of the charging and swapping microgrid, and separates them into power adjustment instructions and equipment operation status adjustment instructions according to the instruction type. It also parses the target parameter range and response time limit corresponding to the instruction. Based on the priority orientation of photovoltaic consumption and combined with the real-time operation status of the microgrid, the resource collaborative scheduling module 41 dynamically adjusts the output power of the charging and swapping equipment, the control parameters of the energy storage battery charging and discharging circuit (including converter modulation parameters and charging and discharging current constraint thresholds), and optimizes the access sequence of the charging and swapping loads. Specifically, the resource coordination and scheduling module 41 performs resource scheduling based on the parsed instructions and the real-time operating status of the microgrid, as follows: Power output adjustment of charging and swapping equipment: By adjusting the duty cycle of the PWM (Pulse Width Modulation) signal of the equipment inverter, the output power is adjusted to achieve precise control of the power of the charging and swapping equipment; Adjustment of control parameters for energy storage battery charging and swapping circuit: Converter modulation parameters: Adjust the carrier frequency or modulation amplitude to change the output power to match the control command; Charge and discharge current constraint threshold: dynamically adjusted based on the current SOC (state of charge) of the energy storage battery to avoid overcharging and over-discharging of the battery; Charging and swapping load access timing optimization: A priority queue mechanism is adopted to classify charging and swapping tasks according to power demand and urgency. Low power demand tasks are scheduled first (which is conducive to photovoltaic consumption), and urgent tasks are scheduled immediately. At the same time, timing adjustment logs are recorded.

[0049] The execution feedback module 42 feeds back the real-time operating parameters (including actual power output and loop status) of the charging and swapping equipment and energy storage battery to the charging and swapping microgrid multi-mode control unit 3, forming a control closed loop.

[0050] Specifically, the execution feedback module 42 collects real-time operating parameters of the charging and swapping equipment and energy storage batteries through distributed sensors: Charging and swapping equipment: Collects actual output power (calculated through current and voltage sensors) and equipment operating status (such as "running" or "standby"); Energy storage battery: collects the voltage, current, temperature of the charging and discharging circuit, as well as the battery SOC (calculated by ampere-hour integration method combined with open circuit voltage correction); these parameters are fed back to the mode determination module 30 of the multi-mode control unit 3 of the charging and swapping microgrid at fixed intervals to update the real-time operating status of the microgrid and form a control closed loop of "decision-execution-feedback".

[0051] Those skilled in the art will understand that the process of implementing all or part of the steps of the above embodiments can be carried out by hardware or by a program instructing the relevant hardware.

[0052] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi-modal flexible control system for a charging and swapping microgrid that fuses photovoltaic curtailment priority, characterized in that, include: The microgrid status real-time sensing unit (1) collects the photovoltaic module voltage and current, charging and swapping load current, energy storage battery terminal voltage and current, and microgrid and main grid interconnection line power signal in real time within the charging and swapping microgrid. It performs noise reduction and outlier removal preprocessing on the collected signals and transmits the preprocessed data to the photovoltaic consumption priority dynamic judgment unit (2). The photovoltaic consumption priority dynamic determination unit (2) receives preprocessed data and performs short-term trend prediction on photovoltaic output, remaining energy storage capacity and charging / swapping load; generates prediction confidence factor through sliding window error estimation, applies smooth convergence constraint to weight adjustment by combining first-order inertial filtering and state timer, dynamically distinguishes microgrid operation scenarios and adapts the weight ratio of core indicators, generates photovoltaic consumption priority result after confidence correction, and transmits it to the charging / swapping microgrid multimodal control unit (3); the photovoltaic consumption priority dynamic determination unit (2) includes a trend prediction module (20), a confidence correction module (21), a scenario adaptation module (22), a weight constraint module (23) and a priority calculation module (24), wherein: The trend prediction module (20) is used to receive the preprocessed data output by the microgrid status real-time sensing unit (1), perform short-term trend prediction on photovoltaic output, energy storage remaining capacity and charging / swapping load and output the predicted value. The confidence correction module (21) is used to receive the predicted value from the trend prediction module (20) and the measured value in the preprocessed data, and generate a confidence factor through error calculation to provide a basis for correcting the priority results; The scenario adaptation module (22) distinguishes microgrid operation scenarios based on the dynamic correlation between trend prediction values ​​and measured values, and dynamically allocates the weights of core indicators such as photovoltaic power output fluctuation characteristics, energy storage remaining capacity, and charging and swapping load demand intensity according to scenario characteristics; the scenario adaptation module (22) receives the photovoltaic power output prediction values ​​output by the trend prediction module (20). Predicted remaining energy storage capacity Forecast values ​​of charging and swapping load and the power of the tie line in the preprocessed data And retrieve the rated power of the charging and swapping equipment. Transmission limits of tie lines Microgrid operation scenarios are divided based on multi-physical quantity coupling criteria, specifically including: when and When the battery is not fully charged, it is considered a scenario with surplus photovoltaic energy. when and The ratio reaches the peak load threshold set based on the equipment operating characteristics, and Less than When the preset ratio is reached, it is determined to be a peak charging / swapping load scenario; when and The ratio reaches the power grid constraint threshold set based on tie-line transmission characteristics, and Greater than When the preset ratio is reached, it is determined to be a power grid interaction constraint scenario; The scenario adaptation module (22) dynamically allocates the weights of core indicators based on the energy flow characteristics of each scenario. ,in Weights for photovoltaic power output fluctuation characteristics, Weighting of remaining energy storage capacity Weighting the demand intensity of charging and swapping loads, and satisfying In scenarios where there is a surplus of photovoltaic energy, the focus is on photovoltaic energy consumption. To achieve the highest proportion, during peak charging and battery swapping load scenarios, the focus is on load assurance. With the highest proportion, energy storage regulation is the core in grid interaction constraint scenarios. To achieve the highest percentage, the dynamic weight will be used in the final calculation. , , Output to the weight constraint module (23); The weight constraint module (23) achieves weight smoothing through first-order inertial filtering and monitors the weight convergence state through a state timer to avoid weight abrupt changes affecting the stability of the judgment. The priority calculation module (24) is used to integrate the confidence factor, the weight of the core index after smooth convergence and the corresponding index characterization value to calculate and generate the corrected photovoltaic consumption priority result; The charging and swapping microgrid multi-mode control unit (3) receives the photovoltaic consumption priority result from the photovoltaic consumption priority dynamic determination unit (2), switches to three control modes: photovoltaic priority consumption, charging and swapping load guarantee, and grid interaction stability, and outputs control commands to the charging and swapping resource flexible scheduling unit (4). The flexible scheduling unit (4) for charging and swapping resources receives control commands from the multi-mode control unit (3) of the charging and swapping microgrid, adjusts the power output of the charging and swapping equipment, the control parameters of the energy storage battery charging and discharging circuit, and the access sequence of the charging and swapping load, so as to realize the flexible control of the charging and swapping microgrid with photovoltaic consumption priority orientation.

2. The multi-modal flexible control system for charging and swapping microgrids integrating photovoltaic absorption priority as described in claim 1, characterized in that, The real-time status sensing unit (1) of the microgrid includes a distributed acquisition module (10), a signal preprocessing module (11), and a data transmission module (12), wherein: The distributed acquisition module (10) consists of a voltage sensor, a current sensor, and a power sensor. The voltage sensor is used to acquire the terminal voltage of the photovoltaic module and the terminal voltage of the energy storage battery. The current sensor is used to acquire the charging and swapping load current and the charging and discharging current of the energy storage battery. The power sensor is used to acquire the instantaneous power of the interconnection line between the microgrid and the main grid. The signal preprocessing module (11) uses mean filtering to reduce noise by averaging multiple sets of signals from the same source to reduce signal fluctuations caused by electromagnetic interference. The outlier removal process of the signal preprocessing module (11) is achieved by judging whether the amplitude of the acquired signal exceeds the normal operating range of the corresponding physical quantity. The outlier determination is based on the rated parameters and physical characteristics of the equipment. The data transmission module (12) uses industrial Ethernet to transmit the pre-processed digital signal to the photovoltaic consumption priority dynamic determination unit (2) in real time according to a standardized format.

3. The multi-modal flexible control system for charging and swapping microgrids integrating photovoltaic absorption priority as described in claim 2, characterized in that, The trend prediction module (20) performs short-term trend prediction and outputs the predicted value, including the following steps: S20.1 Receive the preprocessed data output by the microgrid status real-time sensing unit (1), extract the photovoltaic module voltage and current, charging and swapping load current, energy storage battery terminal voltage and current signals, and establish a multi-physical quantity cross-correlation model: respectively construct the physical correlation model between photovoltaic power output and light intensity, module temperature, dynamic mapping relationship between energy storage remaining capacity and charging and discharging current and internal resistance, and coupling model between charging and swapping load and equipment operating status; S20.2 Select a sliding data window of a preset length and extract the measured values ​​of the core physical quantities within the window. and measured values ​​of related physical quantities ; S20.3 Adaptively determine dynamic coefficients based on the coupling strength of each physical quantity The result is obtained through sliding window fusion calculation. Time prediction value ,in Corresponding to the predicted photovoltaic power output values Predicted remaining energy storage capacity Or the predicted value of charging and swapping load ;in satisfy ; S20.4, will , , Output to the confidence correction module (21).

4. The multi-modal flexible control system for charging and swapping microgrids integrating photovoltaic absorption priority as described in claim 3, characterized in that, The confidence correction module (21) generates the confidence factor by the following steps: S21.1 Receive the predicted values ​​of each physical quantity output by the trend prediction module (20) Synchronously retrieve the corresponding measured values ​​of physical quantities output by the real-time sensing unit (1) of the microgrid status. Select a sliding error calculation window of a preset length; where, For indexing historical moments within the window, and ; S21.2 Calculate the mean absolute error (MAE) between the predicted and measured values ​​within the window using error statistics methods; S21.3 Obtain the rated operating range of the corresponding physical quantity Calculate the reasonable operating range span of the corresponding physical quantity. And adaptively adjust the correction coefficient based on the fluctuation characteristics of physical quantities. ; S21.

4. Convert MAE into a confidence factor in the 0-1 interval using a dynamic normalization method. ,Will Output to the priority calculation module (24).

5. The multimodal flexible control system for charging and swapping microgrids integrating photovoltaic absorption priority as described in claim 4, characterized in that, The weight constraint module (23) implements weight smoothing and convergence determination through the following steps: S23.1, Receive the output of the scene adaptation module (22) Calculate weights in real time retrieval Time-based historical weighting Calculate the change in weights ; S23.2, Set the basic filter coefficients based on the microgrid response characteristics. and the maximum allowable change in weight By calculating the correlation between the weight change and the maximum allowable change, we obtain Time-adaptive filter coefficients ; S23.3, based on , and Perform weighted smoothing calculation to obtain Final weight at time ; S23.4 Start the state timer and calculate the weight change rate. When the rate of change is lower than the convergence threshold set based on the circuit response characteristics within a preset number of consecutive data periods, the weight is determined to have reached a convergence state, and the current weight is locked. And output to the priority calculation module (24); if it does not converge, repeat steps S23.1~S23.

3.

6. The multimodal flexible control system for charging and swapping microgrids integrating photovoltaic absorption priority as described in claim 5, characterized in that, The priority calculation module (24) generates photovoltaic consumption priority results through the following steps: S24.1 Receive the confidence factor output by the confidence correction module (21) The converged core index weights output by the weight constraint module (23) , , , and the original physical quantities of photovoltaic power output fluctuation, energy storage remaining capacity, and charging and swapping load demand intensity output by the microgrid status real-time sensing unit (1); S24.

2. Normalize the original physical quantities in S24.1 based on their respective physical extreme values ​​to obtain the photovoltaic power output fluctuation characterization values. Energy storage remaining capacity characterization value Characteristic value of charging and swapping load demand intensity ,and All are in the 0-1 range; S24.3 Calculate the predicted photovoltaic output value With the current discharge power of energy storage The sum of the two, and the predicted charging and swapping load. Perform a difference comparison; if Set the energy balance correction coefficient ; otherwise, The value is adaptively adjusted according to the energy gap; S24.

4. Photovoltaic grid integration priority results are obtained through deep fusion calculation of multi-dimensional factors. The result is then transmitted to the multi-modal control unit of the charging and swapping microgrid (3).

7. The multi-modal flexible control system for charging and swapping microgrids integrating photovoltaic absorption priority as described in claim 6, characterized in that, The charging and swapping microgrid multimodal control unit (3) includes a mode determination module (30), a multimodal control module (31), and an instruction output module (32), wherein: The mode determination module (30) receives the photovoltaic consumption priority result Priority output by the photovoltaic consumption priority dynamic determination unit (2), and determines the corresponding control mode triggering conditions based on the physical balance relationship between photovoltaic consumption demand, charging and swapping load demand and grid interaction constraints reflected by the photovoltaic consumption priority result Priority. The multimodal control module (31) switches to the corresponding control mode according to the triggering condition, specifically including: When Priority primarily reflects the demand for photovoltaic (PV) power consumption, the system switches to a PV-priority consumption control mode. This mode maximizes the consumption of PV power by adjusting the charging and discharging power of energy storage and optimizing the connection sequence of charging and swapping loads. When Priority primarily reflects the charging and swapping load demand, the system switches to the charging and swapping load guarantee control mode. This mode stabilizes the energy storage discharge output and dynamically adjusts the tie-line power supplement to meet the rated operating power requirements of the charging and swapping equipment. When Priority primarily reflects grid interaction constraints, the system switches to grid interaction stability control mode. By suppressing photovoltaic output fluctuations and adjusting energy storage to smooth power, the power fluctuations of the tie line are controlled within the transmission limit. The instruction output module (32) standardizes and processes the power adjustment instructions and equipment operation status adjustment instructions under each control mode, and outputs them to the flexible scheduling unit (4) for charging and swapping resources.

8. The multi-modal flexible control system for charging and swapping microgrids integrating photovoltaic absorption priority as described in claim 7, characterized in that, The flexible scheduling unit (4) for charging and swapping resources includes an instruction parsing module (40), a resource collaborative scheduling module (41), and an execution feedback module (42), wherein: The instruction parsing module (40) receives the standardized control instructions output by the multi-mode control unit (3) of the charging and swapping microgrid, and splits them into power adjustment instructions and equipment operation status adjustment instructions according to the instruction type. It also parses the target parameter range and response time limit corresponding to the instruction. The resource collaborative scheduling module (41) is based on the priority guidance of photovoltaic consumption and combined with the real-time operation status of the microgrid. It dynamically adjusts the output power of the charging and swapping equipment, the control parameters of the energy storage battery charging and discharging circuit, and optimizes the access sequence of the charging and swapping load. The execution feedback module (42) feeds back the real-time operating parameters of the charging and swapping equipment and the energy storage battery to the charging and swapping microgrid multimodal control unit (3) to form a control closed loop.

Images