Intelligent dispatching system of digital micro-grid

Abstract

The invention provides an intelligent scheduling system for a digital micro-grid. The intelligent scheduling system comprises a distributed power supply module, an energy storage module, a power distribution scheduling module, an energy management module, a load module, a monitoring protection module and an on-grid and off-grid switching device. The energy management module comprises an energy manager and a communication controller. The monitoring protection module comprises an intelligent control unit and a protection unit. The power distribution scheduling module comprises a data collection unit, a data sensing unit, a data processing modeling unit, a dynamic decision-making unit and a multi-mode execution unit. The data processing modeling unit performs modeling according to the information collected by the data collection unit, the dynamic decision-making unit performs optimization analysis and adjusts the operation mode of the power electronic interface equipment according to the establishment of a model, and nodes are preset in a power grid to centrally manage and configure the output of the distributed power supply module. The power distribution scheduling module comprises a centralized optimization layer and a distributed autonomous layer. The method improves the utilization rate of uncertain renewable energy sources, reduces the light abandoning rate, and guarantees the continuity of load power supply.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid dispatching, and in particular to a digital microgrid intelligent dispatching system. The present invention also relates to a digital microgrid intelligent dispatching method. Background Art

[0002] A microgrid is a small energy and power system that is composed of distributed power sources, energy storage, AC / DC distribution networks, power loads, control and protection systems, etc., and can achieve a basic balance of power and electricity within the power supply range. Microgrids can provide industrial and commercial users in urban power grids with optimized balance of sources, grids, loads and storage within the property boundary and emergency power supply. At the same time, they can also provide distributed energy and power supply for power users in weak off-grid areas and achieve self-sufficient energy independence.

[0003] The microgrid management system is an intelligent system that integrates software and hardware for efficient management and optimized control of microgrid systems. Based on electrical automation technologies such as data acquisition, safety protection, power prediction, dynamic optimization, and dispatch control, the microgrid system monitors the operating status, power generation status, load power consumption behavior, on-site environmental parameters, grid dispatch instructions and other constraints of various energy equipment such as photovoltaics, energy storage, diesel generation, fuel generation, wind power, charging piles, and loads in real time, and dynamically forms operation control and dispatch strategies to ensure that the microgrid system always operates in the optimal state. At the same time, through a variety of human-computer interaction methods, efficient operation and maintenance management can be achieved.

[0004] With the development of science and technology, the demand for electric energy is getting higher and higher, and the application of microgrid technology in electric energy dispatching plays a vital role. First, the demand for reliable electricity consumption of enterprises continues to increase. Under the dual carbon policy, the energy consumption control of enterprises is strict, and the power is cut off, which reduces the reliability of the main power grid. In traditional methods, the energy management problem of microgrids usually uses a model-based framework to describe the dynamics of microgrids. Then, the uncertainty is estimated by using a predictor, and the optimal dispatch is obtained by using an optimization problem solver. Although these methods have proven their effectiveness in the field of energy management, they are highly dependent on experts to accurately model the dynamics of microgrids. Since the scale, capacity and dynamics of microgrids may change over time, the uncertainty curves of power generation and demand will change significantly, which greatly limits the applicability of the above methods to large-scale microgrid energy real-time problems, resulting in low model stability and inaccurate and stable dispatch.

[0005] The existing technology adopts the method of installing energy storage devices in each wind power and photovoltaic power station to adjust the volatility and randomness of the regulator itself and avoid the occurrence of wind and solar power abandonment problems. However, such a setting increases the cost of investment in renewable energy equipment to alleviate the prediction error caused by the uncertainty of renewable energy. This method increases the cost of energy utilization, and the energy consumption is almost equal to the energy utilization, which fails to achieve the purpose of fully utilizing renewable energy. Summary of the invention

[0006] In view of this, the present invention aims to propose a digital microgrid intelligent dispatching system to solve the problem of intermittent power dispatching, improve the utilization rate of uncertain renewable energy, reduce the abandonment rate, and ensure the continuity of load power supply.

[0007] To achieve the above object, the technical solution of the present invention is achieved as follows:

[0008] A digital microgrid intelligent dispatching system, comprising a distributed power supply module, an energy storage module, a power distribution dispatching module, an energy management module, a load module, a monitoring and protection module, and an on-grid and off-grid switching device interconnected by an industrial bus;

[0009] The energy management module includes an energy manager and a communication controller;

[0010] The monitoring and protection module includes an intelligent control unit and a protection unit;

[0011] The protection unit is used to detect the power parameters of the microgrid and perform fault protection, and the protection unit is connected to the intelligent control unit by electrical signals;

[0012] The power distribution scheduling module includes a data collection unit, a data perception unit, a data processing modeling unit, a dynamic decision unit, and a multi-modal execution unit;

[0013] The data processing modeling unit performs modeling according to the information collected by the data collection unit, and the dynamic decision-making unit performs optimization analysis and adjusts the operation mode of the power electronic interface device according to the establishment of the model, and centrally manages and configures the output of the distributed power supply module at a preset node in the power grid;

[0014] The power distribution scheduling module includes a centralized optimization layer and a distributed autonomous layer.

[0015] Furthermore, the data perception unit includes a photovoltaic storage collaborative prediction module integrating a hybrid LSTM-wavelet algorithm and a multi-channel load feature extraction device;

[0016] The multi-channel load feature extraction device is equipped with a synchronous sampling circuit of ≥1kHz, which is used to capture the dynamic parameters of harmonic distortion rate and power factor in real time.

[0017] Furthermore, the dynamic decision-making unit includes a non-cooperative game modeling engine and a digital twin verification platform;

[0018] The non-cooperative game modeling engine includes a three-party game tree of new energy units, energy storage modules, and flexible loads, and uses a Q-learning reinforcement learning algorithm to solve the Nash equilibrium point;

[0019] The digital twin verification platform includes a real-time simulator based on RT-LAB, which is configured to verify the transient stability of the scheduling strategy in the digital mirror.

[0020] Further, the multi-modal execution unit includes an FPGA-based network topology controller and a black start power optimization distributor;

[0021] The FPGA-based network topology controller is configured to complete the switching operation of the tie switch matrix within 80ms;

[0022] The black start power optimization allocator has a built-in load priority sorting table based on a dynamic programming algorithm.

[0023] Furthermore, the data collection unit includes a first monitoring component for monitoring the distributed power supply module, a second monitoring component for monitoring the load module, and a third monitoring component for monitoring the energy storage module.

[0024] Furthermore, the intelligent control unit is used to collect the microgrid parameter data and transmit the parameter data to the monitoring center, and the monitoring center is used to analyze the relationship between the parameter data and the load data of the second monitoring component;

[0025] The monitoring center is connected to the intelligent control unit for interactive communication, and the monitoring center is connected to the second monitoring component for electrical signals.

[0026] Furthermore, the energy manager includes an implementation monitoring unit, an energy consumption analysis unit, an intelligent prediction unit, a coordination control unit, an economic dispatch unit, and a demand response unit.

[0027] Furthermore, the distributed autonomous layer includes a virtual power plant collaborative platform, an intelligent edge controller and an integrated photovoltaic and storage machine.

[0028] Furthermore, the energy storage module includes a battery pack, an energy storage converter, a battery management system, and an energy management system;

[0029] The intelligent control unit includes a power monitoring unit, a load monitoring unit, and an information processing unit;

[0030] The intelligent control unit is connected to the communication controller by electrical signals.

[0031] The digital microgrid intelligent dispatching method described in the present invention integrates time series characteristics and frequency domain fluctuation analysis through a hybrid LSTM-wavelet algorithm, and the photovoltaic processing prediction error is significantly reduced compared with the traditional LSTM, solving the problem of intermittent power dispatching. And by setting up a data collection unit to collect and model the influencing factors of wind power generation and the uncertain factors of the load, the dynamic decision-making unit performs optimization analysis based on the model, adjusts the operating status of the electronic interface equipment, centrally manages and configures the output of the distributed power module for the key preset nodes in the power grid, fully allocates the output of renewable energy during the use of the load, and realizes continuous power supply with the assistance of the energy storage device.

[0032] In addition, by setting up a three-party game model and introducing flexible load demand response, the efficiency of peak shaving and valley filling is effectively improved, and the electricity cost on the user side is reduced. The digital twin platform verifies transient stability through RT-LAB simulation to avoid voltage limit accidents caused by scheduling strategy conflicts.

[0033] In addition, the FPGA controller supports nanosecond-level instruction parsing, and the network topology reconstruction speed is twice as fast as the traditional DSP solution, ensuring that the power supply interruption time of key loads is less than 100ms to meet national standards. The dynamic programming algorithm optimizes the black start power allocation, and the success rate of power restoration for important loads in island mode is ≥99.7%.

[0034] Another object of the present invention is to provide a digital microgrid intelligent dispatching method, the dispatching method comprising the following steps:

[0035] Step S1: Establish a security domain model of the distribution network node containing distributed generation, and define voltage, frequency and line capacity constraints;

[0036] Step S2: Select key nodes based on the fuzzy comprehensive evaluation method, generate an initial configuration plan, and synchronously collect the IV curve of the photovoltaic string and the PCS operating status through the edge computing node;

[0037] Step S3: Call the federated learning framework to update the game model parameters and generate a scheduling plan that satisfies ΔV / Δt≤5%·s-1;

[0038] Step S4: simulate and execute the scheduling plan in the digital twin environment, trigger the backtracking mechanism when the transient overvoltage is detected to be greater than 1.1 pu, and verify the feasibility of the plan through the digital twin system;

[0039] Step S5: If the verification is successful, the network topology reconstruction and power allocation are implemented to the multimodal execution unit through the GOOSE protocol. If it fails, return to step S3 for iterative optimization until the verified scheduling instruction is issued after the verification is successful.

[0040] The digital microgrid intelligent dispatching method described in the present invention realizes cross-microgrid data sharing through federated learning, reduces the risk of data leakage in model training, shortens the iteration cycle, and improves data security and efficiency. Risk prejudgment is also achieved, and the overvoltage backtracking mechanism can block the voltage collapse chain reaction in advance, and the transient instability recovery time is less than 500ms. GOOSE protocol transmission supports synchronous operation of multiple execution units, has a high success rate of instruction execution, and has good industrial-grade real-time performance. The dispatching method covers core technical nodes such as LSTM-wavelet algorithm optimization, FPGA hardware design, and GOOSE protocol interaction through the steps of perception, decision-making, and execution, ensuring the efficiency, security, and automation level of the dispatching process. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the accompanying drawings:

[0042] Figure 1 A system architecture diagram of a digital microgrid intelligent dispatching system according to an embodiment of the present invention;

[0043] Figure 2 Schematic diagram of a microgrid control system strategy according to an embodiment of the present invention. DETAILED DESCRIPTION

[0044] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0045] In the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper", "lower", "inner", "back", etc. are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0046] In addition, in the description of the present invention, unless otherwise clearly defined, the terms "installed", "connected", "connected", and "connector" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood in combination with specific circumstances.

[0047] The present invention will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0048] The present embodiment provides a digital microgrid intelligent dispatching system, which includes a distributed power supply module, an energy storage module, a power distribution dispatching module, an energy management module, a load module, a monitoring and protection module, and an on-grid and off-grid switching device interconnected by an industrial bus. The energy management module includes an energy manager and a communication controller. The monitoring and protection module includes an intelligent control unit and a protection unit. The protection unit is used to detect microgrid power parameters and perform fault protection, and the protection unit is connected to the intelligent control unit by electrical signals.

[0049] The distribution scheduling module includes a data collection unit, a data perception unit, a data processing modeling unit, a dynamic decision unit, and a multi-modal execution unit. The data processing modeling unit performs modeling based on the information collected by the data collection unit, and the dynamic decision unit performs optimization analysis and adjusts the operating mode of the power electronic interface device based on the establishment of the model, and centrally manages and configures the output of the distributed power supply module at the preset nodes in the power grid. The distribution scheduling module includes a centralized optimization layer and a distributed autonomous layer.

[0050] The digital microgrid intelligent dispatching method described in this embodiment integrates time series characteristics and frequency domain fluctuation analysis through a hybrid LSTM-wavelet algorithm, and the photovoltaic processing prediction error is significantly reduced compared with the traditional LSTM, solving the problem of intermittent power dispatching. And by setting up a data collection unit to collect and model the influencing factors of wind power generation and the uncertain factors of the load, the dynamic decision-making unit performs optimization analysis based on the model, adjusts the operating status of the electronic interface equipment, centrally manages and configures the output of distributed power modules for key preset nodes in the power grid, fully allocates the output of renewable energy during load use, and achieves continuous power supply with the assistance of energy storage devices.

[0051] As a system that integrates power generation, distribution, storage, and consumption, there is a lot of uncertainty in the planning and operation of microgrids. According to different sources, the uncertainty involved in microgrids can be roughly divided into renewable energy and load uncertainty, energy storage and demand response uncertainty, etc. In the process of microgrid operation optimization and scheduling, these uncertainties need to be considered separately to achieve better optimization purposes.

[0052] Data processing modeling evaluates the impact of different irradiances on the current scheduling of microgrids under different dates and seasons based on the uncertainty of renewable energy and loads. The scenario analysis method is used to model the uncertainty of the output function of the photovoltaic power generation system, the demand forecast error, and the grid quotation changes of the microgrid optimal energy management. The bat algorithm is used to solve and reduce the total operating cost of the grid-connected microgrid.

[0053] As a preferred implementation, the data perception unit includes a photovoltaic energy storage collaborative prediction module integrating a hybrid LSTM-wavelet algorithm and a multi-channel load feature extraction device. The multi-channel load feature extraction device is configured with a synchronous sampling circuit of ≥1kHz for real-time capture of harmonic distortion rate and power factor dynamic parameters.

[0054] Among them, the hybrid LSTM-wavelet algorithm meets the following requirements: when the irradiance change gradient ΔG / Δt is less than 10%·min -1 Under the working condition, the normalized root mean square error of the prediction result satisfies:

[0055] RMSE / P rated ≤3%;

[0056] Where P rated is the rated power value of the PV array.

[0057] With this setting, the gradient of irradiance change is constrained, and the normalized RMSE of the hybrid algorithm is ≤3%. The irradiance mutation noise is filtered through wavelet decomposition, and the prediction stability is maintained in extreme weather (such as rapid cloud movement). The prediction fluctuation error is significantly reduced in the scenario of rapid cloud movement, avoiding the power jump problem caused by noise in the transmission LSTM. In addition, the normalized error index is adapted to microgrids of different capacities, and the normalized index is compatible with 5kW to 10MW photovoltaic systems, supporting the standardized deployment of rural microgrids and urban energy hubs, and facilitating cross-scenario promotion and application.

[0058] In addition, the dynamic decision-making unit of this embodiment includes a non-cooperative game modeling engine and a digital twin verification platform. The non-cooperative game modeling engine includes a three-party game tree of new energy units, energy storage modules, and flexible loads, and uses a Q-learning reinforcement learning algorithm to solve the Nash equilibrium point. The digital twin verification platform includes a real-time simulator based on RT-LAB, which is configured to verify the transient stability of the scheduling strategy in a digital mirror.

[0059] As mentioned above, by setting up a three-party game model to introduce flexible load demand response, the efficiency of peak shaving and valley filling is effectively improved, and the electricity cost on the user side is reduced. The digital twin platform verifies transient stability through RT-LAB simulation to avoid voltage limit accidents caused by scheduling strategy conflicts.

[0060] Furthermore, in the non-cooperative game modeling engine, the utility function of the new energy unit is modeled as:

[0061] U gen =λ·P grid -η·(P curt / P max ) where λ is the real-time electricity price, η is the energy abandonment penalty coefficient, P curt is the abandoned optical power;

[0062] The utility function of the energy storage module is modeled as:

[0063] U ess = k·DoD

[0064] Where k is the dynamic attenuation coefficient based on the Arrhenius life model, and DoD is the depth of discharge.

[0065] By setting the energy abandonment penalty coefficient η and linking it with the real-time electricity price λ, empirical evidence shows that the wind and solar power consumption rate has increased to 97.5%, and the abandonment rate is less than 2%, which has increased the utilization rate of renewable energy, reduced consumption, and improved economic efficiency. In addition, due to the setting of the energy storage module to quantify the temperature-charge and discharge cycle correlation, the calendar life of the energy storage battery is extended.

[0066] Preferably, the multimodal execution unit includes an FPGA-based network topology controller and a black start power optimization distributor. The FPGA-based network topology controller is configured to complete the switching operation of the interconnection switch matrix within 80ms. The black start power optimization distributor has a built-in load priority sorting table based on a dynamic programming algorithm. The FPGA controller supports nanosecond instruction parsing, and the network topology reconstruction speed is doubled compared to the traditional DSP solution, ensuring that the power supply interruption time of critical loads is less than 100ms to meet national standards. The dynamic programming algorithm optimizes the black start power distribution, and the success rate of power restoration of important loads in island mode is ≥99.7%.

[0067] Among them, the data collection unit includes a first monitoring component for monitoring the distributed power module, a second monitoring component for monitoring the load module, and a third monitoring component for monitoring the energy storage module. The first monitoring component includes electrical parameter sensors for monitoring wind power, photovoltaics, and diesel motors, which are used to monitor real-time electrical parameters such as voltage, current, power, and harmonics. Environmental sensors are used to detect environmental parameters such as temperature, humidity, light (photovoltaic system), and wind speed (wind power system). Battery status sensors on wind power and photovoltaics are used to monitor SOC (battery state of charge), SOH (health state), temperature, etc. RTU (remote terminal unit) is used to collect and preprocess local sensor data. Smart meters are used for high-precision metering of electrical energy input / output and bidirectional flow.

[0068] The second monitoring component includes smart meters that measure the voltage, current, power, power factor and other parameters of the load in real time. The power quality analyzer monitors power quality issues such as wave distortion rate, voltage sag / swell, and frequency deviation. The distributed synchronous phasor measurement unit PMU uses high-precision phasor measurement based on GPS synchronization to capture dynamic changes in loads. Temperature / infrared sensors monitor the temperature rise of equipment such as cable joints and transformers to prevent failures caused by overload. The overload protection of the smart circuit breaker is linked to the load status, automatically cutting off non-critical loads, and reporting the opening and closing status through wireless communication.

[0069] The third monitoring component includes battery status monitoring, safety monitoring and protection.

[0070] The battery status monitoring is equipped with a high-precision battery parameter sensor, and its key parameters are:

[0071] SOC (state of charge): error ≤ 1% (based on electrochemical impedance spectroscopy + AI correction algorithm).

[0072] SOH (State of Health): Predict life attenuation through internal resistance, number of cycles, and temperature rise data.

[0073] Temperature / Pressure: Fiber optic sensors monitor hot spots inside the battery in real time.

[0074] The multi-dimensional battery diagnostic instrument is used for cell-level balancing control and prevention of overcharging or over-discharging; lithium plating detection and capacity decay warning.

[0075] Among them, safety monitoring and protection are equipped with thermal runaway warning system and safety linkage device.

[0076] The thermal runaway warning system includes gas sensors and infrared thermal imaging cameras to detect CO, H 2 The intelligent fire fighting module includes automatic spraying of aerosol / liquid coolant, which is linked with the microgrid EMS to quickly cut off the faulty energy storage unit.

[0077] Preferably, the intelligent control unit is used to collect microgrid parameter data and transmit the parameter data to the monitoring center, which is used to analyze the relationship between the parameter data and the load data of the second monitoring component. The monitoring center and the intelligent control unit are connected in interactive communication, and the monitoring center and the second monitoring component are connected in electrical signals.

[0078] By collecting various parameter data of the microgrid and transmitting them to the monitoring center for processing. The monitoring center will analyze these data and realize orderly control of the microgrid through the intelligent control unit according to the actual microgrid load demand. For example, when the load changes, the intelligent control unit can automatically adjust the output power of the distributed power source to ensure that the microgrid can provide sufficient power services. The protection unit determines whether the microgrid is overloaded, short-circuited or other faults by real-time monitoring of the microgrid power parameters. Once an abnormality is detected, the protection unit will take protective action in time to prevent damage to the microgrid electrical equipment.

[0079] Furthermore, the energy manager includes an implementation monitoring unit, an energy consumption analysis unit, an intelligent prediction unit, a coordination control unit, an economic dispatch unit, and a demand response unit. It meets the requirements of comprehensive microgrid operation monitoring, intelligent safety analysis, forward-looking adjustment and control, and dynamic panoramic analysis, and completes the flexible interaction and economic optimization operation between photovoltaic storage and charging resources under different goals, so as to maximize energy, economic and environmental benefits. The monitoring, analysis, prediction, regulation, and economic dispatch data in the energy manager are sent to the distribution network dispatch module for re-judgment, and the demand response is executed when the two decisions are the same.

[0080] In addition, preferably, the distributed autonomous layer includes a virtual power plant collaborative platform, an intelligent edge controller and an integrated photovoltaic and storage machine.

[0081] In addition, the energy storage module includes a battery pack, an energy storage converter, a battery management system, and an energy management system. The intelligent control unit includes a power monitoring unit, a load monitoring unit, and an information processing unit. The intelligent control unit is connected to the communication controller by electrical signals.

[0082] This embodiment also relates to a digital microgrid intelligent dispatching method, which includes the following steps:

[0083] Step S1: Establish a security domain model of the distribution network node containing distributed generation, and define voltage, frequency and line capacity constraints;

[0084] Step S2: Select key nodes based on the fuzzy comprehensive evaluation method, generate an initial configuration plan, and synchronously collect the IV curve of the photovoltaic string and the PCS operating status through the edge computing node;

[0085] Step S3: Call the federated learning framework to update the game model parameters and generate a scheduling plan that satisfies ΔV / Δt≤5%·s-1;

[0086] Step S4: simulate and execute the scheduling plan in the digital twin environment, trigger the backtracking mechanism when the transient overvoltage is detected to be greater than 1.1 pu, and verify the feasibility of the plan through the digital twin system;

[0087] Step S5: If the verification is successful, the network topology reconstruction and power allocation are implemented to the multimodal execution unit through the GOOSE protocol. If it fails, return to step S3 for iterative optimization until the verified scheduling instruction is issued after the verification is successful.

[0088] The digital microgrid intelligent dispatching method of this embodiment is divided into multiple dispatching cycles. First, the electricity price of purchasing electricity directly from the power grid and fully utilizing renewable energy is taken into account. Then, the minimum average load unit price is used as the optimization target, and the power load, installed capacity of wind and solar energy, and battery charge and discharge limits are used as constraints. After the battery participates in the regulation and introduces the decision factors and utilization rates regarding whether the wind turbine and photovoltaic are enabled, the total power supply cost is calculated.

[0089] The load forecast for each period in the microgrid and the maximum output of wind power and photovoltaic power in each period show that in the future, wind turbine processing will be mainly concentrated from 0:00 to 12:00, and photovoltaic processing will be mainly concentrated from 6:00 to 18:00. From the perspective of load demand, there is a steady growth trend at 0:00 and a peak at 19:00, which means that the relationship between load demand and wind turbine processing is negatively correlated, while the correlation with photovoltaic output is weak. Gas turbines have more output during peak hours.

[0090] According to the load demand in each time period, the total load demand for the whole day is calculated. According to the relationship between cost, power unit price, time period, and the principle that power generation and load are equal, the power supply composition of the load when there is no renewable energy and when renewable energy is fully utilized is obtained, when the full-day load needs to be purchased.

[0091] When renewable energy is not enough to meet the load demand, electricity is purchased from the main grid. When renewable energy is greater than the load demand, the remaining electricity can be sold to the main grid. When wind and solar power abandonment is allowed, and renewable energy is used but not at full capacity, all factors must be considered comprehensively to reduce the total cost. The price of renewable energy power generation can be compared with the grid price in each period.

[0092] The power generation status of renewable energy in each time period is obtained, and then the cost of renewable energy power generation is calculated. Then, the load requirement is taken as the algebraic sum of each item to obtain the exchange power between the microgrid and the main grid in each time period.

[0093] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A digital microgrid intelligent dispatching system, characterized by: It includes distributed power supply modules, energy storage modules, power distribution scheduling modules, energy management modules, load modules, monitoring and protection modules, and on-grid and off-grid switching devices interconnected by industrial buses; The energy management module includes an energy manager and a communication controller; The monitoring and protection module includes an intelligent control unit and a protection unit; The protection unit is used to detect the power parameters of the microgrid and perform fault protection, and the protection unit is connected to the intelligent control unit by electrical signals; The power distribution scheduling module includes a data collection unit, a data perception unit, a data processing modeling unit, a dynamic decision unit, and a multi-modal execution unit; The data processing modeling unit performs modeling according to the information collected by the data collection unit, and the dynamic decision-making unit performs optimization analysis and adjusts the operation mode of the power electronic interface device according to the establishment of the model, and centrally manages and configures the output of the distributed power supply module at a preset node in the power grid; The power distribution scheduling module includes a centralized optimization layer and a distributed autonomous layer.

2. The digital microgrid intelligent dispatching system according to claim 1 is characterized in that: The data perception unit includes a photovoltaic storage collaborative prediction module and a multi-channel load feature extraction device integrating a hybrid LSTM-wavelet algorithm; The multi-channel load feature extraction device is equipped with a synchronous sampling circuit of ≥1kHz, which is used to capture the dynamic parameters of harmonic distortion rate and power factor in real time.

3. The digital microgrid intelligent dispatching system according to claim 2 is characterized in that: The dynamic decision-making unit includes a non-cooperative game modeling engine and a digital twin verification platform; The non-cooperative game modeling engine includes a three-party game tree of new energy units, energy storage modules, and flexible loads, and uses a Q-learning reinforcement learning algorithm to solve the Nash equilibrium point; The digital twin verification platform includes a real-time simulator based on RT-LAB, which is configured to verify the transient stability of the scheduling strategy in the digital mirror.

4. The digital microgrid intelligent dispatching system according to claim 3 is characterized in that: The multi-modal execution unit includes a network topology controller based on FPGA and a black start power optimization distributor; The FPGA-based network topology controller is configured to complete the switching operation of the tie switch matrix within 80ms; The black start power optimization allocator has a built-in load priority sorting table based on a dynamic programming algorithm.

5. The digital microgrid intelligent dispatching system according to claim 1 is characterized in that: The data collection unit includes a first monitoring component for monitoring the distributed power supply module, a second monitoring component for monitoring the load module, and a third monitoring component for monitoring the energy storage module.

6. The digital microgrid intelligent dispatching system according to claim 5 is characterized in that: The intelligent control unit is used to collect the microgrid parameter data and transmit the parameter data to the monitoring center, and the monitoring center is used to analyze the relationship between the parameter data and the load data of the second monitoring component; The monitoring center is connected to the intelligent control unit for interactive communication, and the monitoring center is connected to the second monitoring component for electrical signals.

7. The digital microgrid intelligent dispatching system according to claim 1 is characterized in that: The energy manager includes an implementation monitoring unit, an energy consumption analysis unit, an intelligent prediction unit, a coordination control unit, an economic dispatch unit, and a demand response unit.

8. The digital microgrid intelligent dispatching system according to claim 1 is characterized in that: The distributed autonomous layer includes a virtual power plant collaborative platform, an intelligent edge controller and an integrated photovoltaic and storage machine.

9. The digital microgrid intelligent dispatching system according to claim 1 is characterized in that: The energy storage module includes a battery pack, an energy storage converter, a battery management system, and an energy management system; The intelligent control unit includes a power monitoring unit, a load monitoring unit, and an information processing unit; The intelligent control unit is connected to the communication controller by electrical signals.

10. A digital microgrid intelligent dispatching method, characterized in that: The steps include: Step S1: Establish a security domain model of the distribution network node containing distributed generation, and define voltage, frequency and line capacity constraints; Step S2: Select key nodes based on the fuzzy comprehensive evaluation method, generate an initial configuration plan, and synchronously collect the IV curve of the photovoltaic string and the PCS operating status through the edge computing node; Step S3: Call the federated learning framework to update the game model parameters and generate a scheduling plan that satisfies ΔV / Δt≤5%·s-1; Step S4: simulate and execute the scheduling plan in the digital twin environment, trigger the backtracking mechanism when the transient overvoltage is detected to be greater than 1.1 pu, and verify the feasibility of the plan through the digital twin system; Step S5: If the verification is successful, the network topology reconstruction and power allocation are implemented to the multimodal execution unit through the GOOSE protocol. If it fails, return to step S3 for iterative optimization until the verified scheduling instruction is issued after the verification is successful.

Images