Distributed resource optimization aggregation method for distribution network based on ant colony simulated annealing algorithm

Abstract

The application belongs to the technical field of power systems and discloses a power distribution network distributed resource optimization aggregation method based on an ant colony simulated annealing algorithm; the method first acquires parameters of distributed energy in a virtual power plant, constructs an objective function with the aim of minimizing the total cost of the virtual power plant, sets a constraint condition, and finally solves the objective function by combining an ant colony algorithm and a simulated annealing algorithm to output an optimal solution, namely, an optimal aggregation method of the distributed resources of the power distribution network; the application can jump out of the trap of a local optimal solution, enhance the global search capability, improve the convergence speed, and significantly improve the probability of finding a global optimal solution; the cooling strategy of the simulated annealing algorithm accelerates the late convergence, the pheromone mechanism of the ant colony algorithm strengthens high-quality solutions, and the cooperation of the two can improve the economy and stability of the scheme; by setting the constraint, the aggregation scheme can meet various operation requirements, reduce the scheduling cost, improve the resource utilization rate, realize the efficient utilization and optimal allocation of resources, and the like.

Description

Technical Field

[0001] This invention belongs to the field of power system technology, specifically relating to a distributed resource optimization and aggregation method for distribution networks based on ant colony simulated annealing algorithm. Background Technology

[0002] With the advancement of global energy transition, the proportion of distributed energy resources in the power system is constantly increasing. Virtual power plants (VPPs), as a new power management model integrating distributed energy resources, adjustable loads, and energy storage devices, have emerged. Through information technology and intelligent control, they achieve unified scheduling and management of various energy sources to improve energy utilization efficiency and enhance power system stability. However, existing aggregation methods do not fully consider various practical factors such as the limitations of grid interaction and equipment operating constraints during the operation of virtual power plants. This leads to the infeasibility of proposed aggregation schemes in actual operation, failing to meet the operational requirements of virtual power plants. Furthermore, the intermittency and uncertainty of distributed energy generation within virtual power plants, as well as the diversity and complexity of adjustable loads, make resource scheduling a key challenge. How to rationally allocate these resources to meet electricity demand, optimize energy utilization, and reduce operating costs has become an urgent problem to be solved.

[0003] In the virtual power plant resource scheduling problem, traditional optimization algorithms such as linear programming and integer programming have limitations when dealing with large-scale optimization problems under complex constraints. These algorithms are usually based on precise mathematical models, have high computational complexity, and long solution times, making them difficult to adapt to the real-time scheduling requirements of virtual power plants. Moreover, as the problem size increases or the constraints become more complex, traditional algorithms are prone to getting trapped in local optima and failing to find the globally optimal scheduling scheme. Taking linear programming as an example, its assumptions are relatively idealized and cannot accurately describe the complex characteristics and uncertainties of distributed energy resources and loads in virtual power plants, resulting in poor performance in practical applications.

[0004] To address the shortcomings of traditional algorithms, intelligent optimization algorithms are increasingly being applied in the field of virtual power plant resource scheduling. Among them, the Ant Colony Algorithm (ACO) possesses strong global search capabilities and a positive feedback mechanism, enabling it to handle complex combinatorial optimization problems to a certain extent. However, it suffers from slow convergence, is prone to getting trapped in local optima, and the selection of parameters such as the pheromone evaporation coefficient and pheromone importance factor significantly impacts its performance. Furthermore, these parameters are typically set empirically, making it difficult to adapt to different virtual power plant scenarios. While the Simulated Annealing Algorithm (SA) has strong global optimization capabilities, it lacks in-depth utilization of the problem structure during the search process, resulting in relatively low search efficiency. Current technologies do not fully combine the advantages of both, leading to virtual power plant resource aggregation schemes failing to meet the demands for high efficiency and stability in practical operation. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention proposes a distributed resource optimization and aggregation method for power distribution networks based on ant colony simulated annealing algorithm, which is implemented through the following steps:

[0006] A method for optimizing and aggregating distributed resources in a power distribution network based on ant colony simulated annealing algorithm, the method comprising the following steps:

[0007] Step 1. Obtain the parameters of distributed energy resources in the virtual power plant;

[0008] Step 2. Construct an objective function with the goal of minimizing the total cost of the virtual power plant;

[0009] Step 3. Set constraints;

[0010] Step 4. Combine the ant colony algorithm and the simulated annealing algorithm to solve the objective function and output the optimal solution, which is the optimal aggregation method for distributed resources in the distribution network.

[0011] Furthermore, the parameters of the distributed energy in the virtual power plant obtained in step 1 include: the power generation cost coefficient, power generation capacity, minimum and maximum power generation capacity of each distributed energy in each time period; the interaction power between the virtual power plant and the grid in each time period; the minimum and maximum power of interaction with the grid; the electricity price purchased from the grid; the electricity price sold to the grid; and the adjustment cost coefficient, original load demand, adjustment amount, minimum and maximum adjustment amount of each adjustable load in each time period.

[0012] Furthermore, the total cost of the virtual power plant described in step 2 includes the generation cost of distributed energy, the interaction cost with the power grid, and the adjustment cost of the adjustable load.

[0013] Furthermore, the objective function described in step 2 is as follows:

[0014]

[0015] Where C represents the total cost of the virtual power plant over the entire dispatch cycle, T represents the set of dispatch periods, t∈T, I represents distributed energy resources, i∈I, J represents adjustable load, j∈J, x ij (t) is the decision variable, representing the proportion of distributed energy source i supplying power to adjustable load j during time period t, C gen,i P represents the generation cost coefficient (yuan / kWh) of distributed energy source i. gen,i (t) represents the power generation of distributed energy source i during time period t, C buy (t) represents the electricity price (yuan / kWh) purchased from the grid during time period t. grid (t) represents the interaction power between the virtual power plant and the power grid during time period t, C sell (t) represents the electricity price (yuan / kWh) sold to the grid during time period t, C adj,jΔL represents the adjustment cost coefficient (yuan / kWh) of the adjustable load j. j (t) represents the adjustment amount of the adjustable load j in time period t.

[0016] Furthermore, the constraints described in step 3 are as follows:

[0017] (1) The total power generation in each time period is equal to the total load demand;

[0018] (2) The power generation capacity of distributed energy sources in each time period is within the range of their allowable minimum and maximum power generation capacity;

[0019] (3) The interaction power between the virtual power plant and the power grid in each time period is within the specified minimum and maximum interaction power range;

[0020] (4) The adjustment amount of the adjustable load in each time period is within the range of its allowable minimum and maximum adjustment amount.

[0021] Furthermore, step 4, which combines the ant colony algorithm and the simulated annealing algorithm to solve the objective function and output the optimal solution, is as follows: First, the ant colony algorithm is used to generate an initial solution. Based on the objective function, the path selection is optimized through the pheromone update mechanism. Second, the simulated annealing algorithm is used to generate a new solution by perturbing the neighborhood of the solution generated by the ant colony algorithm. The new solution is accepted or rejected according to the Metropolis criterion, and the solution is gradually converged to the optimal solution.

[0022] Furthermore, the initial solution is generated using the ant colony algorithm, and the path selection is optimized through a pheromone update mechanism based on the objective function, as detailed below:

[0023] First, set the basic parameters required for the ant colony algorithm, construct the pheromone matrix of distributed energy and adjustable load, and randomly place each ant in a different initial load aggregation state.

[0024] Next, each ant starts from its initial position and selects the next scheduling state according to the path selection probability formula until the path is constructed.

[0025] Finally, the objective function value corresponding to each path is obtained, and the pheromone is updated according to the objective function value until the pheromone converges.

[0026] Furthermore, the formula for updating pheromones is as follows:

[0027]

[0028] Where, τ ij (t) represents the pheromone concentration at time t, ρ represents the pheromone evaporation coefficient, and m represents the number of ants. This represents the pheromone increment left by ant Q on path (i,j). The lower the cost, the greater the pheromone increment. The larger; L Q Let denot Q be the length (cost) of the path traveled by ant Q.

[0029] Furthermore, the simulated annealing algorithm generates a new solution by perturbing the neighborhood of the solution generated by the ant colony algorithm, thereby obtaining the change in the objective function value. If the change is less than or equal to 0, the new solution is accepted directly; if the change is greater than 0, the new solution is accepted or rejected according to the Metropolis criterion. The algorithm continues to iterate until the current temperature reaches the threshold, at which point a cooling operation is performed to obtain the optimal solution.

[0030] Furthermore, the specific steps for accepting or rejecting a new solution based on the Metropolis criterion are as follows: accept the new solution with probability p, generate a random number k, and determine whether k is less than p. If so, accept the new solution; otherwise, reject the new solution.

[0031] Advantages of this invention:

[0032] The proposed distributed resource optimization aggregation method for power distribution networks based on ant colony simulated annealing algorithm combines the extensive search capability of simulated annealing with the positive feedback mechanism of ant colony algorithm. This allows the method to escape the "trap" of local optima, enhance global search capability, improve convergence speed, and significantly increase the probability of finding the global optimum. The cooling strategy of simulated annealing accelerates later convergence, while the pheromone mechanism of ant colony algorithm strengthens high-quality solutions. The synergy of these two approaches improves the economy and stability of the solution. By setting constraints, the aggregation scheme is ensured to meet various operational requirements while reducing scheduling costs, improving resource utilization, and achieving efficient resource utilization and optimized allocation. Attached Figure Description

[0033] Figure 1 This is a flowchart of a distributed resource optimization and aggregation method for power distribution networks based on ant colony simulated annealing algorithm according to an embodiment of the present invention;

[0034] Figure 2 This is a comparison chart of the results before and after algorithm optimization in one embodiment of the present invention. Detailed Implementation

[0035] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings.

[0036] In this embodiment of the invention, a virtual power plant aggregates three types of distributed energy sources: rooftop solar photovoltaic, small wind turbine generators, and ground source heat pump systems; and four types of adjustable loads: small industrial production equipment, commercial building air conditioning systems, commercial lighting systems, and agricultural irrigation equipment. The method of this invention is used to obtain the optimal aggregation method for distributed resources in the distribution network. The flowchart of the method is as follows: Figure 1 As shown, it includes the following steps:

[0037] Step 1. Obtain the parameters of distributed energy resources in the virtual power plant, as follows:

[0038] Important information is collected from the reported virtual power plants regarding I distributed energy sources and J adjustable loads within a scheduling period set of T, including:

[0039] (1) Power generation of each distributed energy source at different times: This parameter reflects the power generation capacity of different distributed energy sources at different times. It is a key variable for calculating total power generation, assessing power generation costs, and ensuring power balance, namely: P gen,i (t): The power generation of distributed energy source i in time period t;

[0040] (2) The power exchange between the virtual power plant and the power grid at different times: positive values ​​represent electricity purchased from the grid, and negative values ​​represent electricity sold to the grid. It reflects the power transaction between the virtual power plant and the power grid and has an important impact on cost calculation and power balance, namely: P grid (t): The power of interaction between the virtual power plant and the power grid at time t;

[0041] (3) Adjustment amount of each adjustable load in each time period: This parameter reflects the power consumption adjustment amount of the adjustable load in different time periods, which is used to optimize load allocation and reduce costs, i.e.: ΔL j (t): The adjustment amount of the adjustable load j in time period t;

[0042] (4) Generation cost coefficient for each distributed energy source: This coefficient is used to calculate the generation cost of distributed energy sources. Different distributed energy sources have different generation cost coefficients, namely: C gen,i : The generation cost coefficient of distributed energy source i (yuan / kWh);

[0043] (5) Electricity price purchased from the grid at different times: This is the price at which the virtual power plant purchases electricity from the grid at different times, and is an important parameter for calculating grid interaction costs, namely: C buy (t): The price of electricity purchased from the grid during time period t (yuan / kWh);

[0044] (6) Electricity price sold to the grid at different times: This represents the price at which the virtual power plant sells electricity to the grid at different times, and is also used to calculate the grid interaction cost, i.e.: C sell (t): The price of electricity sold to the grid during time period t (yuan / kWh);

[0045] (7) Adjustment cost coefficient for each adjustable load: It reflects the cost incurred in adjusting a unit of electricity for a certain adjustable load and is used to calculate the load adjustment cost, i.e.: C adj,j : Adjustment cost coefficient of adjustable load j (yuan / kWh);

[0046] (8) Minimum and maximum power generation of each distributed energy source: These two parameters limit the power generation range of the distributed energy source, ensuring that the power generation is within the range that the equipment can withstand and that operation is safe, i.e.: P gen,min,i P gen,max,i : These represent the minimum and maximum power generation of distributed energy source i, respectively;

[0047] (9) Minimum and maximum power interaction with the power grid: These limitations are mainly determined by the grid capacity, safe operation requirements, and agreements signed with the virtual power plant, i.e.: P grid,min : Minimum power exchanged with the power grid, P grid,max : The maximum value of the power interacting with the power grid;

[0048] (10) Minimum and maximum adjustment amounts for each adjustable load: These two parameters specify the adjustment range of the adjustable load, ensuring that the adjustment operation meets the actual situation and user needs, i.e.: ΔL min,j ΔL max,j : These represent the minimum and maximum adjustment amounts of the adjustable load j, respectively;

[0049] (11) Original load demand of each adjustable load in each time period: This represents the normal power demand of a certain adjustable load in any time period without any adjustment operations. This is the basic data for calculating load regulation and power balance, i.e.: The original load demand (kW) of the adjustable load in time period t.

[0050] Where t∈T, i∈I (i=1,2,…,I), j∈J (j=1,2,…,J).

[0051] Step 2. Construct the objective function with the goal of minimizing the total cost of the virtual power plant, as follows:

[0052] Define decision variable x ij (t): represents the proportion of time period t in which distributed energy source i supplies power to adjustable load j, where x ij ∈[0,1].

[0053] The objective is to minimize the total cost of the virtual power plant over the entire dispatch cycle. The total cost includes the generation cost of distributed energy resources, the cost of interaction with the grid, and the adjustment cost of adjustable loads, where T is the set of dispatch periods, and t∈T.

[0054]

[0055] Step 3. Set the constraints as follows:

[0056] (1) Power balance constraint

[0057] At each time period t, the power generation of distributed energy, the power interaction with the grid, and the adjustment amount of adjustable load should satisfy the power balance relationship, that is, the total power generation equals the total load demand.

[0058]

[0059] (2) Distributed energy generation power constraints

[0060] The power generation of distributed energy source i in each time period t must be within its allowable minimum and maximum power generation range.

[0061]

[0062] (3) Power constraints of grid interaction

[0063] The power exchange between the virtual power plant and the grid at each time period t must be within the specified minimum and maximum power exchange range.

[0064]

[0065] (4) Adjustable load adjustment amount constraint

[0066] The adjustment amount of the adjustable load j in each time period t must be within its allowable minimum and maximum adjustment range.

[0067]

[0068] Power balance constraints ensure that the power supply and demand of the virtual power plant are balanced in each time period; the power generation constraints of distributed energy, the power interaction constraints of the grid, and the adjustment constraints of adjustable loads limit the operating range of each device, thus ensuring the feasibility of the dispatching scheme.

[0069] Step 4. Combine the ant colony algorithm and simulated annealing algorithm to solve the objective function and output the optimal solution, as follows:

[0070] Step 4.1. Initialize the simulated annealing algorithm parameters:

[0071] (1) Parameters: Initial temperature T0, exponential cooling strategy T k+1 =rT k T k T represents the current temperature in the simulated annealing algorithm. k+1 This indicates that the simulated annealing algorithm is in T k Based on the iterative temperature, cooling rate r, and maximum number of iterations N max Termination temperature T end ;

[0072] (2) Initial temperature: To ensure that the simulated annealing algorithm can accept large changes in the objective function value in the initial stage and extensively explore the solution space, the initial temperature is T0 = 500.

[0073] (3) Cooling rate: The cooling rate determines how quickly the temperature decreases, affecting the speed at which the algorithm transitions from global search to local search. To obtain the optimal solution, the temperature decrease rate should be reduced, i.e., the cooling rate r = 0.95;

[0074] (4) Termination temperature: Set the termination temperature T end When the temperature drops below this value, the algorithm stops iterating. To obtain the optimal solution, the solution accuracy needs to be increased, i.e., the termination temperature T needs to be set. end =1.

[0075] Step 4.2. Generate an initial solution using the ant colony algorithm, and optimize path selection using a pheromone update mechanism:

[0076] Step 4.2.1. Set the basic parameters required for the initial solution: the range of values ​​for the pheromone evaporation coefficient ρ, the pheromone importance factor α, and the heuristic information importance factor β, the number of ants m, and the pheromone matrix τ matching the problem size. IJ .

[0077] Randomly generate decision variable x ij (t) and a set of parameter values ​​of the ant colony algorithm are used as the initial solution of the simulated annealing algorithm, including the pheromone evaporation coefficient ρ (ρ∈0.1-0.5), the pheromone importance factor α (α∈1-5), the range of values ​​of the heuristic information importance factor β (β∈2-5), and the number of ants m (m∈100-1000).

[0078] Pheromone matrix initialization: Constructing a pheromone matrix τ that matches the problem size. IJ The virtual power plant has I distributed energy sources and J adjustable loads, and the pheromone matrix has a dimension of (I+J)×(I+J).

[0079] Initialize all elements to τ IJ =0.1.

[0080] Ant placement: Each ant is randomly placed in a different initial load aggregation state, that is, the power supply of each distributed energy source to each adjustable load is randomly determined, but basic constraints such as power balance must be met.

[0081] Step 4.2.2. Ant Colony Algorithm Search Phase:

[0082] (1) Set the main parameters: τ represents the set of possible states that ant Q (Q∈m) can choose for its next move; ij: Indicates pheromone concentration; Heuristic information, indicating the degree of cost reduction resulting from transitioning from the current scheduling state to the next state; d ij τ: The cost of supplying power to distributed energy source i when supplying power to adjustable load j; is : indicates the pheromone that represents the ant's next state selection; η is : represents the heuristic information for the ant's next state selection; s: represents the set of possible states for ant Q to choose next. One of the specific optional states in the system.

[0083] (2) Path Construction: Ant Q starts from the initial position and selects the next scheduling state according to the path selection probability formula. The ant continuously selects, which means matching the distributed energy and adjustable load in the virtual power plant until a complete scheduling scheme is constructed. Finally, it determines which distributed energy supplies which adjustable loads, thus determining the aggregation scheme. The path selection probability formula is:

[0084]

[0085] Among them, heuristic information Defined as: (when ΔC) ij <0, i.e., cost reduction), where ∈ is a very small positive number to avoid the denominator being zero; d ij The rate of cost reduction is inversely proportional to the magnitude of the cost decrease; the greater the cost reduction, the higher the d. ij The smaller the value, the higher the probability that the ant will choose that path.

[0086] In calculating d ij Beforehand, it is necessary to ensure that the scheduling scheme after the state transition meets the set constraints. If the state transition causes a constraint violation, then d ij =0, to prevent ants from choosing infeasible paths.

[0087] The current state is that distributed energy source i is not supplying power to adjustable load j. The next state is when distributed energy source i supplies power to adjustable load j. The cost change ΔC gen It consists of the following parts:

[0088] ΔC gen =x ij (t)C gen ,i ·P gen ,i (t) (7)

[0089] If distributed energy source i is added to supply power to adjustable load j, the power generation P needs to be increased. gen,i (t), leading to an increase in power generation costs and a change in grid interaction costs ΔC. grid for:

[0090]

[0091] If distributed energy source i supplies power to adjustable load j, it can reduce the amount of electricity purchased from the grid or increase the amount of electricity sold, thereby reducing grid interaction costs. The adjustment cost change of the adjustable load is ΔC. adj for:

[0092] ΔC adj =C adj,j ·ΔL j (t) (9)

[0093] If the adjustable load j participates in the polymerization, the adjustment amount ΔL can be reduced. j (t), thereby reducing adjustment costs, and the total cost change ΔC ij for:

[0094] ΔC ij =ΔC gen +ΔC grid +ΔC adj (10)

[0095] By integrating and optimizing formulas (6) to (10), and to simplify the calculation, d can be... ij Normalization to the [0,1] interval enhances the discriminative power of heuristic information, avoids path selection bias caused by differences in units, and obtains d. ij The calculation formula is:

[0096]

[0097] Where, ΔC is This represents the cost changes that occur when energy is allocated and transmitted from the current distributed energy node i to an selectable adjustable load or other distributed energy state s.

[0098] (3) Pheromone Update: After all ants have completed path construction, they calculate the cost (i.e., objective function value) for each path and update the pheromone based on the objective function value. The pheromone update affects the path selection of subsequent ants, and thus affects the decision of the devices participating in the aggregation. As the number of iterations increases, the pheromone concentration gradually converges to a stable state, guiding the ants to construct a better scheduling scheme and more accurately determine the devices participating in the aggregation.

[0099] For path (i,j), the pheromone update formula is:

[0100]

[0101] in, This represents the pheromone increment left by ant Q on path (i,j). The lower the cost, the greater the pheromone increment. The larger; L Q Let denot Q be the length (cost) of the path traveled by ant Q.

[0102] Step 4.3. Combining simulated annealing, a new solution is generated by perturbing the neighborhood of the solution generated by the ant colony algorithm. The new solution is accepted or rejected according to the Metropolis criterion, gradually converging to the optimal solution:

[0103] Step 4.3.1. Generate new solution: In the neighborhood of the current ant colony algorithm parameters, generate new ant colony algorithm parameters through certain perturbation rules, that is: add a small random quantity Δρ to the current ρ value ((Δρ∈[-0.05,0.05])); add a small random quantity Δα to the current α value ((Δα∈[-0.5,0.5])); add a small random quantity Δβ to the current β value ((Δβ∈[-0.5,0.5])); add a small random quantity Δm to the current m value ((Δm∈[-50,50])).

[0104] Step 4.3.2. Calculate the change in the objective function value: Substitute the current ant colony algorithm parameters and the newly generated ant colony algorithm parameters into the ant colony algorithm respectively, and run the ant colony algorithm to obtain the corresponding objective function value f. n and f n-1 Calculate the change in the objective function value Δf = f n -f n-1 .

[0105] Step 4.3.3. Determine whether to accept the new solution: If Δf ≤ 0, it means the new solution is better or the same as the current solution, and the new solution is accepted directly, i.e., the newly generated ant colony algorithm parameters are accepted; if Δf > 0, then according to the Metropolis criterion, the new solution is accepted with probability. Accept the newly generated ant colony algorithm parameters, generate a random number k in the interval [0,1]. If k < p, accept the newly generated ant colony algorithm parameters; otherwise, retain the current ant colony algorithm parameters.

[0106] Step 4.3.4. Check termination conditions: Determine the current temperature T k Has the termination temperature T been reached? end If yes, stop the simulated annealing algorithm and continue to step 4.3.5; otherwise, return to step 4.3.1.

[0107] Step 4.3.5. Cooling operation: Update the temperature according to the cooling rate r, i.e., T k+1 =rT k As the temperature decreases, the probability of the algorithm accepting a poor solution gradually decreases, eventually leading to the globally optimal solution.

[0108] Step 4.4. Based on the obtained global optimal solution, the final relationship between distributed energy resources and adjustable loads with respect to x is generated.ij The matrix of (t) represents the optimal scheduling scheme, guiding the aggregation of virtual power plant resources.

[0109] In this embodiment of the invention, optimization calculations are performed according to the method described above to obtain the final optimized aggregation scheme matrix of the virtual power plant as follows:

[0110]

[0111] In the formula: x 13 =0.1 indicates that the first distributed energy source (rooftop solar photovoltaic) will supply 10% of its power generation to the third load (commercial lighting system).

[0112] like Figure 2 As shown, the optimized aggregation scheme obtained through this method has a lower power supply cost, enabling the virtual power plant to rationally select and utilize resources under various operating environments, ensuring its stable and efficient operation, and solving the problem of resource optimization and allocation in complex and ever-changing environments.

[0113] The method provided by this invention has been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are merely for the purpose of helping to understand the core ideas of this invention. It should be noted that those skilled in the art can make various improvements and modifications to this invention without departing from its principles, and these improvements and modifications also fall within the protection scope of the claims of this invention.

Claims

1. A method for distributed resource optimization and aggregation in power distribution networks based on ant colony simulated annealing algorithm, characterized in that, The method includes the following steps: Step 1. Obtain the parameters of distributed energy resources in the virtual power plant; Step 2. Construct an objective function with the goal of minimizing the total cost of the virtual power plant; the objective function is as follows: （1） in, This represents the total cost of the virtual power plant over the entire dispatch cycle. Represents the set of scheduling periods. , Indicates distributed energy. , Indicates adjustable load. , Let be the decision variable, representing distributed energy. During the period For adjustable load The ratio of power supply Distributed energy The power generation cost coefficient, Distributed energy During the period Power generation capacity, Indicates time period The price of electricity purchased from the grid, Indicates the virtual power plant during the time period Interaction power with the power grid, Indicates time period The price of electricity sold to the grid. Indicates adjustable load The adjustment cost coefficient, Indicates adjustable load During the period The adjustment amount; Step 3. Set constraints; Step 4. Combine the ant colony algorithm and the simulated annealing algorithm to solve the objective function and output the optimal solution, which is the optimal aggregation method for distributed resources in the distribution network.

2. The method for distributed resource optimization and aggregation in distribution networks based on ant colony simulated annealing algorithm according to claim 1, characterized in that: The parameters of distributed energy in the virtual power plant obtained in step 1 include: the power generation cost coefficient, power generation capacity, minimum and maximum power generation capacity of each distributed energy in each time period; the interaction power between the virtual power plant and the grid in each time period; the minimum and maximum interaction power with the grid; the electricity price purchased from the grid; the electricity price sold to the grid; and the adjustment cost coefficient, original load demand, adjustment amount, minimum and maximum adjustment amount of each adjustable load in each time period.

3. The method for distributed resource optimization and aggregation in distribution networks based on ant colony simulated annealing algorithm according to claim 1, characterized in that: The total cost of the virtual power plant mentioned in step 2 includes the generation cost of distributed energy, the interaction cost with the power grid, and the adjustment cost of adjustable load.

4. The method for distributed resource optimization and aggregation in distribution networks based on ant colony simulated annealing algorithm according to claim 1, characterized in that: The constraints described in step 3 are as follows: (1) The total power generation in each time period is equal to the total load demand; (2) The power generation capacity of distributed energy sources in each time period is within the range of their allowable minimum and maximum power generation capacity; (3) The interaction power between the virtual power plant and the power grid in each time period is within the specified minimum and maximum interaction power range; (4) The adjustment amount of the adjustable load in each time period is within the range of its allowable minimum and maximum adjustment amount.

5. The method for distributed resource optimization and aggregation in distribution networks based on ant colony simulated annealing algorithm according to claim 1, characterized in that: Step 4, which combines ant colony optimization and simulated annealing to solve the objective function and output the optimal solution, is as follows: First, an initial solution is generated using ant colony optimization. Based on the objective function, the path selection is optimized through a pheromone update mechanism. Second, a new solution is generated by perturbing the neighborhood of the solution generated by ant colony optimization. The new solution is accepted or rejected according to the Metropolis criterion, and the solution is gradually converged to the optimal solution.

6. The method for distributed resource optimization and aggregation in distribution networks based on ant colony simulated annealing algorithm according to claim 5, characterized in that: The method of generating an initial solution using the ant colony algorithm and optimizing path selection based on the objective function through a pheromone update mechanism is as follows: First, set the basic parameters required for the ant colony algorithm, construct the pheromone matrix of distributed energy and adjustable load, and randomly place each ant in a different initial load aggregation state. Next, each ant starts from its initial position and selects the next scheduling state according to the path selection probability formula until the path is constructed. Finally, the objective function value corresponding to each path is obtained, and the pheromone is updated according to the objective function value until the pheromone converges.

7. The method for distributed resource optimization and aggregation in distribution networks based on ant colony simulated annealing algorithm according to claim 6, characterized in that: The formula for updating pheromones is as follows: （12） in, Indicates time period pheromone concentration, Indicates the pheromone evaporation coefficient. Indicates the number of ants. Ants In the path The lower the cost of the pheromone increase left on the surface, the better. The larger; , For ants The length of the path traveled.

8. The method for distributed resource optimization and aggregation in distribution networks based on ant colony simulated annealing algorithm according to claim 5, characterized in that: The simulated annealing algorithm described above generates a new solution by perturbing the neighborhood of the solution generated by the ant colony algorithm, thereby obtaining the change in the objective function value; If the change is less than or equal to 0, the new solution is accepted directly. If the change is greater than 0, the new solution is accepted or rejected according to the Metropolis criterion. The iteration continues until the current temperature reaches the threshold, at which point a cooling operation is performed to obtain the optimal solution.

9. The method for distributed resource optimization and aggregation in distribution networks based on ant colony simulated annealing algorithm according to claim 8, characterized in that: The specific meaning of accepting or rejecting a new solution based on the Metropolis criterion is: using probability... Accept the new solution and generate a random number. ,judge Is it less than If so, accept the new solution; otherwise, reject the new solution.

Images