Integrated Energy Microgrid Dynamic Scheduling Method Based on Artificial Intelligence

Abstract

The present invention belongs to the field of energy scheduling, specifically referring to a dynamic scheduling method for an integrated energy microgrid based on artificial intelligence. The method includes locally optimizing the commutation angle, optimizing the start-stop strategies of distributed energy units, searching for the optimal positions of the units, and globally optimizing the microgrid structure. This solution uses the ant colony optimization algorithm to reduce torque ripple, improve the efficiency of the units, avoid current regulation delay, increase the dynamic response speed, reduce power losses, and improve the voltage curve in the energy network system; creatively proposes an energy microgrid optimization method combining a genetic algorithm and a particle swarm optimization algorithm. The particle swarm optimization algorithm is responsible for global search, and the genetic algorithm is responsible for local optimization. The weight coefficients of the two are adjusted through fuzzy logic control to guide the particles to move towards the optimal solution, improve the efficiency of the algorithm for searching the layout of distributed energy units, and reduce the computational complexity.

Description

Technical Field

[0001] The present invention relates to the field of energy scheduling, and specifically refers to a dynamic scheduling method for an integrated energy microgrid based on artificial intelligence. Background Art

[0002] Improving energy utilization efficiency is an important way to save energy and reduce emissions. The dynamic scheduling method for an integrated energy microgrid based on artificial intelligence refers to a method that uses artificial intelligence technology to optimize the configuration of energy unit equipment and the topology of the energy network. In existing approximate solutions, for example, in the power optimization scheduling method, device and equipment in the new energy grid connection scenario of CN119010022B, for the technical problem of the defect in the prior art of ignoring the differences between units, this solution analyzes the frequency domain characteristics of the electric power data of each wind turbine in the wind power array, calculates the influence weight, measures the power volatility of the wind farm, and combines a neural network to predict the future output power, achieving the technical effects of improving the stability of the power grid and reducing the power grid fluctuation risk. However, firstly, there is a technical problem that it only relies on the frequency domain analysis of the electric power vector, does not consider the spatial distribution and topological structure factors between wind turbines, and does not consider the start-up response time and ramp rate of energy units, resulting in lag in energy scheduling.

[0003] In addition, for example, in a new energy microgrid group scheduling method considering stability constraints in CN118040789B, for the technical problem of low-frequency oscillation occurring due to a high proportion of new energy access in the microgrid in the prior art, the minimum damping ratio is calculated through the Lyapunov energy function, and the minimum damping ratio is used as a stability constraint, and a multi-time-domain rolling optimization scheduling scheme is adopted, achieving the technical effects of preventing low-frequency oscillation problems caused by a high proportion of new energy access in the microgrid and solving the negative impacts such as power fluctuation, line overload, voltage instability, and even system collapse caused by low-frequency oscillation. However, when multiple microgrids are interconnected through tie lines, there is a coupling effect among multiple sub-microgrids, and the dynamic response is more complex, resulting in a large error. Moreover, the state variables, power constraints, and damping constraints of each sub-microgrid will increase the difficulty of multi-dimensional optimization. Continuously collecting future multiple time scales faces the technical problem of excessively high computational complexity. Summary of the Invention

[0004] In view of the above situation, to overcome the defects of the prior art, the present invention provides a dynamic scheduling method for an integrated energy microgrid based on artificial intelligence. Aiming at the technical problems existing in the prior art that only rely on the frequency-domain analysis of the electric power vector, do not consider the spatial distribution and topological structure factors among wind turbines, do not consider the start-up response time and ramp rate of energy units, resulting in lagging energy scheduling, this solution uses the ant colony optimization algorithm to reduce torque ripple, improve the efficiency of the unit, avoid current regulation delay, improve the dynamic response speed, reduce power loss and improve the voltage curve in the energy network system, optimize the start-stop strategy of distributed energy units, and improve the reliability and safety of the network; when multiple microgrids are interconnected through tie lines, there are multi-sub-microgrid coupling effects, and the dynamic response is more complex, resulting in larger errors, and the state variables, power constraints and damping constraints of each sub-microgrid will increase the difficulty of multi-dimensional optimization. By continuously collecting multiple future time scales, it faces the technical problem of too high computational complexity. This solution creatively proposes an energy microgrid optimization method combining a genetic algorithm and a particle swarm optimization algorithm. The particle swarm optimization algorithm is responsible for global search, and the genetic algorithm is responsible for local optimization, which not only avoids the premature convergence of the particle swarm optimization algorithm but also prevents the genetic algorithm from overly disturbing the solution space. By adjusting the weight coefficients of the two through fuzzy logic control, it guides the particles to move towards the optimal solution, realizes more stable optimization, improves the efficiency of the algorithm for searching the layout of distributed energy units, and reduces the computational complexity.

[0005] The technical solution adopted by the present invention is as follows: The dynamic scheduling method for an integrated energy microgrid based on artificial intelligence provided by the present invention includes the following steps:

[0006] Step S1: Locally optimize the commutation angle to maintain the stability of the commutation process of distributed energy units. Specifically, in the integrated energy microgrid, the ant colony optimization method is used to offline find and store the individual optimal switching angles and angle change speeds of distributed energy units in a look-up table. The switching angles include the turn-on angle and the turn-off angle, and the combination of the switching angle and the angle change speed is recorded as a candidate solution;

[0007] Step S2: Optimize the start-stop strategy of distributed energy units. Specifically, during the operation of distributed energy units, the look-up table is searched online, and the direct instantaneous torque control technology is used to adjust the start-stop strategy of distributed energy units and optimize the current waveform by optimizing the switching angles and angle change speeds of distributed energy units;

[0008] Step S3: Search for the best positions of the units. Specifically, a method combining a genetic algorithm and a particle swarm optimization algorithm is used to iteratively optimize the positions and network topological relationships of distributed energy units in the integrated energy microgrid to obtain the best positions and network topological relationships;

[0009] Step S4: Globally optimize the microgrid structure, and optimize the integrated energy microgrid according to the best positions and network topological relationships.

[0010] Further, in step S1, the commutation angle is locally optimized, specifically including the following steps:

[0011] Step S11: Determine the ant colony optimization objective. Specifically, calculate the objective function of the ant colony optimization method, integrate and average the instantaneous torque within one electrical cycle to obtain the average torque, record the difference between the maximum torque value and the minimum torque value within the same electrical cycle as the torque ripple, assign a corresponding weight to the torque ripple, and subtract the weighted torque ripple from the average torque to obtain the torque quality. The calculation formula of the objective function used is as follows:

[0012] ;

[0013] In the formula, represents the objective function of the ant colony optimization method, represents the weight corresponding to the torque ripple, represents the torque quality, represents the reference torque quality for normalization processing, represents the unit efficiency of converting electrical energy into mechanical energy, represents the reference efficiency;

[0014] Step S12: Determine the constraint conditions of the ant colony optimization method to prevent the distributed energy unit from generating negative torque;

[0015] Step S13: Store the optimal solution. Specifically, use a lookup table to store the candidate solutions after each iteration update by the ant colony optimization method. By maximizing the torque quality and the unit efficiency, find the optimal candidate solution and store it in the lookup table.

[0016] Further, in step S12, determining the constraint conditions of the ant colony optimization method specifically includes the following steps:

[0017] Step S121: Angle constraint. Specifically, preset a minimum conduction angle threshold to ensure that the difference between the turn-off angle and the turn-on angle is greater than the minimum conduction angle threshold;

[0018] Step S122: Region division. Specifically, according to the inductance change rate, with time as the abscissa and inductance as the ordinate, divide the inductance positive slope region and the non-inductance positive slope region. The inductance positive slope region is the region where the inductance increases with the change of the rotor position;

[0019] Step S123: Inductance constraint. Specifically, set the turn-off angle before the end of the inductance positive slope region, that is, turn off the current before the inductance reaches the maximum value and turn on the current after the inductance reaches the maximum value.

[0020] Further, in step S3, searching for the optimal positions of the units specifically includes the following steps:

[0021] Step S31: Initialize the population. Specifically, regard the distributed energy units as nodes, and initialize the particle population of the particle swarm optimization algorithm. The particles represent potential node positions and topological relationships;

[0022] Step S32: Set the constraint conditions during the search process;

[0023] Step S33: Initialize the genetic weight parameters;

[0024] Step S34: Global search. Specifically, use the particle swarm optimization algorithm to calculate the fitness and perform a global search on the particle population;

[0025] Step S35: Local optimization. It is used to perform local optimization on the particle population using the genetic algorithm. Specifically, according to the genetic weight parameters, randomly select particles for crossover operation and mutation operation. The crossover operation refers to selecting two particles to exchange their information to generate new particles. The mutation operation refers to randomly changing the particle information;

[0026] Step S36: Adjust the influence of the genetic algorithm on the search process;

[0027] Step S37: Population merging. Specifically, merge the new particle populations generated by the particle swarm optimization algorithm and the genetic algorithm, update the particle population, and select particles according to the fitness as the particle population for the next generation.

[0028] Further, in step S32, setting the constraint conditions during the search process specifically includes the following steps:

[0029] Step S321: Power constraint. Specifically, ensure that the active power output of any node must be equal to the sum of the active power demand of this node and the total active power transmitted through all lines passing through this node. Ensure that the reactive power output of any node must be equal to the sum of the reactive power demand of this node and the total reactive power transmitted through all lines passing through this node;

[0030] Step S322: Capacity limit. Specifically, preset the specified voltage range and power capacity limit to ensure that the voltage is within the specified voltage range and ensure that the active power output and reactive power output of each node cannot exceed the power capacity limit.

[0031] Further, in step S34, the global search specifically includes the following steps:

[0032] Step S341: Update the velocity and position of the particles according to the historical best position and the group best position of each particle;

[0033] Step S342: Calculate the fitness of each particle using the following formula:

[0034] ;

[0035] In the formula, represents the fitness, represents the total number of nodes in the integrated energy microgrid, and respectively represent the active power at nodes and ; and respectively represent the reactive power at nodes and ; and represent the network correlation coefficients.

[0036] Furthermore, in step S36, adjust the influence of the genetic algorithm on the search process, specifically including the following steps:

[0037] Step S361: Calculate the contribution degree of the genetic algorithm to measure the optimization effect of the genetic algorithm. Specifically, divide the cumulative sum of the gain degrees brought by the genetic algorithm in the most recent random round of loop by the total number of particles affected by the genetic algorithm in the most recent random round of loop to obtain the optimization effect of the genetic algorithm. The cumulative sum of the gain degrees refers to the fitness difference of the particles whose fitness is improved and are affected by the genetic algorithm;

[0038] Step S362: Calculate the contribution degree of the particle swarm optimization algorithm. Repeat step S361 to measure the optimization effect of the particle swarm optimization algorithm and obtain the optimization effect of the particle swarm optimization algorithm;

[0039] Step S363: Rule learning. Specifically, use a fuzzy logic inference system to dynamically adjust the search process. If the optimization effect of the genetic algorithm is good, it means that the genetic algorithm makes a greater contribution to the optimization in the current stage, and the role of the genetic algorithm should be enhanced. If the optimization effect of the particle swarm optimization algorithm is good, it means that the particle swarm optimization algorithm makes a greater contribution, and the role of the genetic algorithm should be reduced. Update the genetic weight parameter. The update formula of the genetic weight parameter is as follows:

[0040] ;

[0041] In the formula, represents the genetic weight parameter, represents the fuzzy logic inference system, represents the optimization effect of the genetic algorithm.

[0042] The present invention provides a dynamic scheduling method for an integrated energy microgrid based on artificial intelligence. The beneficial effects achieved by the present invention using the above scheme are as follows:

[0043] (1) Aiming at the technical problems existing in the prior art that only rely on the frequency-domain analysis of the electric power vector, do not consider the spatial distribution and topological structure factors among wind turbines, do not consider the start-up response time and ramp rate of energy units, resulting in lagging energy scheduling, this solution uses the ant colony optimization algorithm to reduce torque ripple, improve the efficiency of the unit, avoid current regulation delay, improve the dynamic response speed, reduce power loss and improve the voltage curve in the energy network system, optimize the start-stop strategy of distributed energy units, and improve the reliability and security of the network;

[0044] (2)When multiple microgrids are interconnected through tie lines, there are multi-sub-microgrid coupling effects, and the dynamic response is more complex, resulting in larger errors. Moreover, the state variables, power constraints, and damping constraints of each sub-microgrid will increase the difficulty of multi-dimensional optimization. Continuously collecting multiple future time scales faces the technical problem of excessively high computational complexity. This solution creatively proposes an energy microgrid optimization method that combines a genetic algorithm and a particle swarm optimization algorithm. The particle swarm optimization algorithm is responsible for global search, and the genetic algorithm is responsible for local optimization. It not only avoids the premature convergence of the particle swarm optimization algorithm but also prevents the genetic algorithm from overly disturbing the solution space. By adjusting the weight coefficients of the two through fuzzy logic control, it guides the particles to move towards the optimal solution, realizes more stable optimization, improves the efficiency of the algorithm in searching for the layout of distributed energy units, and reduces the computational complexity. Description of the Drawings

[0045] Figure 1 It is a schematic flow chart of the integrated energy microgrid dynamic scheduling method based on artificial intelligence provided by the present invention;

[0046] Figure 2 It is a schematic diagram of step S1;

[0047] Figure 3 It is a schematic diagram of step S12;

[0048] Figure 4 It is a schematic diagram of step S3;

[0049] Figure 5 It is a schematic diagram of step S32;

[0050] Figure 6 It is a schematic diagram of step S34;

[0051] Figure 7 It is a schematic diagram of step S36.

[0052] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, and do not constitute a limitation to the present invention. Detailed Embodiment

[0053] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0054] Embodiment 1. Refer to Figures 1 to 7 , the integrated energy microgrid dynamic scheduling method based on artificial intelligence provided by the present invention includes the following steps:

[0055] Step S1: Locally optimize the commutation angle to maintain the stability of the commutation process of the distributed energy unit. Specifically, in the integrated energy microgrid, the ant colony optimization method is used to offline search for and store the individual optimal switching angles and angle change speeds of the distributed energy unit in a lookup table. The switching angles include the turn-on angle and the turn-off angle, and the combination of the switching angle and the angle change speed is denoted as a candidate solution.

[0056] Step S2: Optimize the start-stop strategy of the distributed energy unit. Specifically, during the operation of the distributed energy unit, the lookup table is searched online, and the direct instantaneous torque control technology is used to adjust the start-stop strategy of the distributed energy unit and optimize the current waveform by optimizing the switching angle and the angle change speed of the distributed energy unit.

[0057] Step S3: Search for the best position of the unit. Specifically, a method combining the genetic algorithm and the particle swarm optimization algorithm is used to cyclically optimize the position and network topology relationship of the distributed energy units in the integrated energy microgrid to obtain the best position and network topology relationship.

[0058] Step S4: Globally optimize the microgrid structure and optimize the integrated energy microgrid according to the best position and network topology relationship.

[0059] Embodiment 2. Refer to Figures 1 to 2 , based on the above embodiment, in step S1, locally optimize the commutation angle, which specifically includes the following steps:

[0060] Step S11: Determine the ant colony optimization objective. Specifically, calculate the objective function of the ant colony optimization method, integrate and average the instantaneous torque within one electrical cycle to obtain the average torque, denote the difference between the maximum torque value and the minimum torque value within the same electrical cycle as the torque ripple, assign a corresponding weight to the torque ripple, and subtract the weighted torque ripple from the average torque to obtain the torque quality. The calculation formula of the used objective function is as follows:

[0061] ;

[0062] In the formula, Represents the objective function of the ant colony optimization method, Represents the weight corresponding to torque ripple, Represents torque quality, Represents the reference torque quality for normalization processing, Represents the unit efficiency of converting electrical energy into mechanical energy, Represents the reference efficiency;

[0063] Step S12: Determine the constraint conditions of the ant colony optimization method to prevent the distributed energy unit from generating negative torque;

[0064] Step S13: Store the optimal solution. Specifically, use a lookup table to store the candidate solutions after each iteration update by the ant colony optimization method. Find the optimal candidate solution by maximizing the torque quality and the unit efficiency, and store it in the lookup table.

[0065] Example 3, refer to Figures 1 to 3 , This example is based on the above example. In step S12, determine the constraint conditions of the ant colony optimization method, which specifically include the following steps:

[0066] Step S121: Angle constraint. Specifically, preset a minimum conduction angle threshold to ensure that the difference between the turn-off angle and the turn-on angle is greater than the minimum conduction angle threshold;

[0067] Step S122: Region division. Specifically, according to the inductance change rate, with time as the abscissa and inductance as the ordinate, divide the inductance positive slope region and the non-inductance positive slope region. The inductance positive slope region is the region where the inductance increases with the change of the rotor position;

[0068] Step S123: Inductance constraint. Specifically, set the turn-off angle before the end of the inductance positive slope region, that is, turn off the current before the inductance reaches the maximum value and turn on the current after the inductance reaches the maximum value.

[0069] Example 4, refer to Figures 1 to 4 , This example is based on the above example. In step S3, search for the best position of the unit, which specifically includes the following steps:

[0070] Step S31: Initialize the population. Specifically, record the distributed energy unit as a node and initialize the particle population of the particle swarm optimization algorithm. The particles represent potential node positions and topological relationships;

[0071] Step S32: Set the constraint conditions during the search process;

[0072] Step S33: Initialize the genetic weight parameters;

[0073] Step S34: Global search, specifically, using the particle swarm optimization algorithm to calculate the fitness and perform global search on the particle population;

[0074] Step S35: Local optimization, which is used to perform local optimization on the particle population using the genetic algorithm. Specifically, according to the genetic weight parameter, randomly select particles for crossover operation and mutation operation. The crossover operation refers to selecting two particles to exchange their information to generate new particles, and the mutation operation refers to randomly changing the particle information;

[0075] Step S36: Adjust the influence of the genetic algorithm on the search process;

[0076] Step S37: Population merging, specifically, merging the new particle populations generated by the particle swarm optimization algorithm and the genetic algorithm, updating the particle population, and selecting particles according to the fitness as the particle population of the next generation.

[0077] Example Five, refer to Figures 1 to 5 , this example is based on the above example. In step S32, set the constraint conditions in the search process, which specifically include the following steps:

[0078] Step S321: Power constraint, specifically, ensure that the active power output of any node must be equal to the sum of the active power demand of the node plus the active power transmitted through all lines passing through the node, and ensure that the reactive power output of any node must be equal to the sum of the reactive power demand of the node plus the reactive power transmitted through all lines passing through the node;

[0079] Step S322: Capacity limit, specifically, preset the voltage regulation range and power capacity limit, ensure that the voltage is within the voltage regulation range, and ensure that the active power output and reactive power output of each node cannot exceed the power capacity limit.

[0080] Example Six, refer to Figures 1 to 6 , this example is based on the above example. In step S34, global search specifically includes the following steps:

[0081] Step S341: Update the velocity and position of the particle according to the historical best position and the group best position of each particle;

[0082] Step S342: Calculate the fitness of each particle, and the formula used is as follows:

[0083] ;

[0084] In the formula, represents the fitness, represents the total number of nodes in the integrated energy microgrid, and respectively represent the node and the active power at and respectively represent the reactive power at nodes and ; and represent the network correlation coefficient.

[0085] Example 7. Refer to Figures 1 to 7 . Based on the above example, further, in step S36, adjust the influence of the genetic algorithm on the search process, which specifically includes the following steps:

[0086] Step S361: Calculate the contribution degree of the genetic algorithm, which is used to measure the optimization effect of the genetic algorithm. Specifically, divide the cumulative sum of the gain degrees brought by the genetic algorithm in the recent random round of loops by the total number of particles affected by the genetic algorithm in the recent random round of loops to obtain the optimization effect of the genetic algorithm. The cumulative sum of the gain degrees refers to the fitness difference of the particles whose fitness is improved and are affected by the genetic algorithm;

[0087] Step S362: Calculate the contribution degree of the particle swarm optimization algorithm. Repeat step S361 to measure the optimization effect of the particle swarm optimization algorithm and obtain the optimization effect of the particle swarm optimization algorithm;

[0088] Step S363: Rule learning. Specifically, use a fuzzy logic inference system to dynamically adjust the search process. If the optimization effect of the genetic algorithm is good, it means that the genetic algorithm makes a greater contribution to optimization in the current stage, and the role of the genetic algorithm should be enhanced. If the optimization effect of the particle swarm optimization algorithm is good, it means that the contribution of the particle swarm optimization algorithm is greater, and the role of the genetic algorithm should be reduced. Update the genetic weight parameter. The genetic weight parameter update formula is as follows:

[0089] ;

[0090] In the formula, represents the genetic weight parameter, represents the fuzzy logic inference system, represents the optimization effect of the genetic algorithm.

[0091] Example 8. Refer to Figures 1 to 7 . Based on the above example, in step S1, this example uses a 12 / 8 pole three-phase switched reluctance motor as a distributed energy unit, sets the minimum conduction angle threshold to 15°, and sets the turn-off angle before the end of the positive slope region of the inductor. The inequality used is as follows:

[0092] ;

[0093] In the formula, represents the turn-off angle, is the rotor position corresponding to the maximum inductance.

[0094] Embodiment Nine, refer to Figures 1 to 7 , this embodiment is based on the above embodiment. In step S3,

[0095] The integrated energy microgrid includes 5 nodes, among which 3 nodes are 12 / 8 pole three-phase switched reluctance motors, and 2 nodes are load devices. The specified voltage range for each node is from 0.9 pu to 1.1 pu. For each node , the following power balance equations are established:

[0096] ;

[0097] In the formula, represents the active power output of node , represents the active power demand of node , represents the active power transmission from node to all the nodes connected to it;

[0098] ;

[0099] In the formula, represents the reactive power output of node , represents the reactive power demand of node , represents the reactive power transmission from node to all the nodes connected to it.

[0100] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or device.

[0101] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

[0102] The present invention and its embodiments have been described above. Such description is not restrictive. What is shown in the drawings is only one of the embodiments of the present invention, and the actual structure is not limited thereto. Generally speaking, if those of ordinary skill in the art are inspired by it and, without departing from the gist of the present invention, design similar structural modes and embodiments to this technical solution without creative efforts, they shall fall within the protection scope of the present invention.

Claims

1. A dynamic dispatching method for integrated energy microgrid based on artificial intelligence, characterized by: The method comprises the following steps: Step S1: Locally optimize the commutation angle, specifically, in the integrated energy microgrid, use the ant colony optimization method to offline search and use a lookup table to store the optimal switching angle and angle change speed of the individual distributed energy units, the switching angle includes a turn-on angle and a turn-off angle, and record the combination of the switching angle and the angle change speed as a candidate solution; Step S2: Optimizing the start-stop strategy of the distributed energy unit, specifically, during the operation of the distributed energy unit, searching the lookup table online, using direct instantaneous torque control technology, optimizing the switch angle and angle change speed of the distributed energy unit, adjusting the start-stop strategy of the distributed energy unit, and optimizing the current waveform; Step S3: searching for the optimal position of the unit, specifically, using a method combining a genetic algorithm and a particle swarm optimization algorithm to cyclically optimize the position and network topology of the distributed energy units in the integrated energy microgrid to obtain the optimal position and network topology; Step S4: Globally optimize the microgrid structure and optimize the integrated energy microgrid according to the optimal location and network topology.

2. The method for dynamic scheduling of integrated energy microgrid based on artificial intelligence according to claim 1 is characterized in that: In step S3, searching for the optimal position of the unit specifically includes the following steps: Step S31: Initializing the population, specifically, recording the distributed energy units as nodes, initializing the particle population of the particle swarm optimization algorithm, and the particles represent potential node positions and topological relationships; Step S32: setting constraints in the search process; Step S33: Initialize genetic weight parameters; Step S34: global search, specifically, using a particle swarm optimization algorithm to calculate fitness and perform a global search on the particle population; Step S35: local optimization, which is used to perform local optimization on the particle population using a genetic algorithm, specifically, randomly selecting particles for crossover and mutation operations according to genetic weight parameters; Step S36: adjusting the influence of the genetic algorithm on the search process; Step S37: merging populations, specifically, merging the new particle populations generated by the particle swarm optimization algorithm and the genetic algorithm to update the particle population.

3. The method for dynamic scheduling of integrated energy microgrid based on artificial intelligence according to claim 2 is characterized in that: In step S32, the constraints in the search process are set, which specifically includes the following steps: Step S321: power constraint, specifically, ensuring that the active power output of any node must be equal to the active power demand of the node plus the sum of the active power transmitted through all lines of the node, and ensuring that the reactive power output of any node must be equal to the reactive power demand of the node plus the sum of the reactive power transmitted through all lines of the node; Step S322: Capacity limitation.

4. The method for dynamic scheduling of integrated energy microgrid based on artificial intelligence according to claim 2 is characterized in that: In step S34, the global search specifically includes the following steps: Step S341: Update the speed and position of the particle; Step S342: Calculate the fitness of each particle using the following formula: ； In the formula, represents fitness, represents the total number of nodes in the integrated energy microgrid, and Respectively represent nodes and The active power at and Respectively represent nodes and The reactive power at and Represents the network correlation coefficient.

5. The method for dynamic scheduling of integrated energy microgrid based on artificial intelligence according to claim 4 is characterized in that: In step S36, the influence of the genetic algorithm on the search process is adjusted, which specifically includes the following steps: Step S361: Calculate the contribution of the genetic algorithm, specifically, divide the cumulative sum of the gain degree brought by the genetic algorithm in the most recent random round cycle by the total number of particles affected by the genetic algorithm in the most recent random round cycle to obtain the optimization effect of the genetic algorithm; Step S362: Calculate the contribution of the particle swarm optimization algorithm, repeat step S361, and obtain the particle swarm optimization effect; Step S363: rule learning, specifically, using a fuzzy logic reasoning system to dynamically adjust the search process, and updating the genetic weight parameters by comparing the optimization effect of the genetic algorithm and the particle swarm optimization effect. The genetic weight parameter update formula is as follows: ； In the formula, represents the genetic weight parameter, represents the fuzzy logic reasoning system, It represents the optimization effect of the genetic algorithm.

Images