A microgrid economic dispatch optimization method based on improved multi-objective artificial hummingbird algorithm

Abstract

The present invention proposes a microgrid economic dispatch optimization method based on an improved multi-objective artificial hummingbird algorithm, which relates to the field of power systems. The multi-objective artificial hummingbird algorithm is improved using chaotic initialization and optimized inertia coefficients to optimize the dispatch of wind power, photovoltaic power, energy storage, and diesel generators. The constraints of the microgrid system are determined, and the two objectives of minimizing the economic cost of microgrid operation and maximizing the wind and solar power consumption ratio are considered. A microgrid economic dispatch model containing wind power, photovoltaic power, energy storage, and diesel generators is constructed. Because the artificial hummingbird algorithm has few control parameters and strong exploration and development capabilities, the introduction of chaotic initialization and linear inertia weight coefficients improves the multi-objective artificial hummingbird algorithm, enhancing its global optimization capability and greatly improving the convergence speed of the global optimal solution, which is of great significance to the economic dispatch of microgrids.

Description

Technical Field

[0001] The technical solution of the present invention belongs to the field of power system dispatching, and specifically relates to a microgrid economic dispatch optimization method based on an improved multi-objective artificial hummingbird algorithm. Background Art

[0002] Microgrids are a crucial component of smart grids, encompassing various power equipment, including distributed power sources (DGs), loads, and energy storage devices. Microgrids can fully leverage the economic, environmental, and flexibility of DGs, providing users with high-quality, safe, and reliable power. Incorporating renewable DGs into microgrid planning is essential. However, natural energy sources like wind and solar are intermittent and vary depending on geographic constraints, making forecasting their power generation complex. This, in turn, complicates renewable energy generation planning in microgrids.

[0003] Microgrids offer exceptionally flexible operation, allowing them to be integrated into the main grid or disconnected from it to operate as isolated islands, ensuring power supply to critical loads in the event of a grid failure. Green energy is becoming a mainstream energy source advocated by the nation's energy supply. However, with the incorporation of numerous wind farms and photovoltaic power stations into microgrids, the intermittent fluctuations in their output power introduce significant uncertainty into the system's economic dispatch.

[0004] At present, there have been many studies on the economic and environmental protection of microgrids at home and abroad, analyzing the impact of micro-power output characteristics, access capacity and location on reliability indicators. However, in microgrid systems with a large number of new energy distributed power sources, the investment and use costs of wind power and photovoltaic units are relatively high. Existing microgrid economic dispatch often overuses thermal power units while focusing on economic costs and ignores the proportion of wind and solar power consumption, resulting in a large number of environmentally unfriendly wind and solar power abandonment phenomena; or without considering economic efficiency, all wind power and photovoltaic power are preferentially absorbed, resulting in high economic costs. Therefore, the present invention constructs a microgrid economic dispatch optimization model containing wind farms and photovoltaic power stations that coordinates economy and environmental protection. Based on the improved multi-objective artificial hummingbird algorithm, the dispatch model is optimized to find a dispatch solution that takes into account both economy and environmental protection. Summary of the Invention

[0005] The present invention provides a microgrid economic dispatch optimization method based on an improved multi-objective artificial hummingbird algorithm. The method takes wind power, photovoltaic power, diesel generators and energy storage devices into consideration, comprehensively considers the requirements of the economy, environmental protection and reliability of the microgrid, and realizes the stable operation of the microgrid power system under the premise of considering the optimal wind and solar power consumption ratio and operating economic cost.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0007] A microgrid economic dispatch optimization method based on an improved multi-objective artificial hummingbird algorithm, the method comprising:

[0008] Step 1: Obtaining basic data on the operation and load of generator sets in a microgrid system containing renewable energy, wherein the microgrid system includes wind turbines, photovoltaic cells, diesel generators, and battery energy storage devices;

[0009] Step 2: Based on the operating data of the generator set and the basic data of the load, an economic dispatch model of the microgrid system including wind power, photovoltaic power, diesel generators and energy storage is constructed;

[0010] Step 3: Establish multiple objective function models of the microgrid system and determine the power constraints of the microgrid system;

[0011] Step 4: Combine the improved multi-objective artificial hummingbird algorithm (MOAHA) with the wind and solar output control strategy and the economic dispatch optimization model to obtain the optimal solution for the economic dispatch of each distributed power source; apply the improved multi-objective artificial hummingbird algorithm to the optimal dispatch of the microgrid system to obtain the optimal solution and realize the economic optimal dispatch of the microgrid system.

[0012] Furthermore, the step 4 includes:

[0013] Step 4.1: Replace the random generation in MOAHA with Sine chaos initialization and place n hummingbirds on n food sources. The specific mapping method is shown in Equation (2). The population is initialized by Sine chaos mapping so that the hummingbird population is evenly distributed within the upper and lower limits of the search space, thereby improving the search performance of the artificial hummingbird algorithm:

[0014] x1=Low+r·(Up-Low) (1)

[0015]

[0016] Where Up and Low are the upper and lower bounds of the d-dimensional problem, r is a random vector in [0, 1], xi represents the location of the i-th food source of the solution to the given problem, and a is a parameter with a value of (0, 4);

[0017] Step 4.2: Food source access table initialization:

[0018]

[0019] Where, for i=j, VT ij =null indicates that the hummingbird feeds on its specific food source; i≠j, VT ij = 0 means that the jth food source has just been visited by the i-th hummingbird in the current iteration;

[0020] Step 4.3: Using the improved linear foraging behavior inertia coefficient, replace the original 50% probability in MOAHA to guide foraging behavior or territorial foraging behavior;

[0021]

[0022]

[0023] Where W is the inertia coefficient of linear foraging behavior; W max and W min The maximum and minimum inertia coefficients are [0, 1] respectively; it is the number of iterations; Maxit is the maximum number of iterations; v i (t+1) is the candidate food source location of the i-th hummingbird at time t+1; x i (t) is the food source location of the i-th hummingbird at time t, x i,tar (t) is the location of the target food source that the i-th hummingbird intends to visit; D is the flight index coefficient, which is 0 or 1; a is the guide factor, which follows the standard normal distribution N(0, 1); b is the territory factor, which follows the standard normal distribution N(0, 1);

[0024] Step 4.4: A hummingbird uses three methods to find its target food source.

[0025] Under the random probability of rand < W, the hummingbird chooses to perform the following guided foraging behavior:

[0026] v i (t+1)=x i,tar (t)+a·D·(x i (t)-x i,tar (t)) (7)

[0027] The position of the i-th food source is updated as follows:

[0028]

[0029] Where f(·) is the fitness value of the function. If the fitness of the candidate food source is lower than that of the current food source, the hummingbird will abandon the current food source and stay at the candidate food source generated by formula (7) to feed; update the visit table;

[0030] In the case of random probability rand ≥ W, the hummingbird chooses to perform the following territorial foraging behavior:

[0031] v i (t+1)=x i,tar (t)+b·D·x i (t) (9)

[0032] The location of the i-th food source is updated in the same way as in (8), and the visit table is updated;

[0033] Step 4.5: Determine the number of iterations. If the number of iterations exceeds the predetermined value of the migration coefficient 2n, the hummingbird will migrate to a new food source randomly generated in the entire search space. The hummingbird's position is updated as follows:

[0034] x i =Low+r·(Up-Low) (10)

[0035] Update visit list;

[0036] Step 4.6: Calculate the fitness function and obtain the Pareto solution;

[0037] Step 4.7: Perform non-dominated sorting to obtain the Pareto frontier;

[0038] Step 4.8: Obtain the optimal result.

[0039] Furthermore, the population dimensions initialized in step 4.1 are 96 parameters obtained based on the 24-hour output of four distributed power sources in the economic dispatch of the microgrid system; wherein the 1st to 24th parameters represent the 24-hour output of the photovoltaic cell; wherein the 25th to 48th parameters represent the 24-hour output of the wind turbine; wherein the 49th to 72nd parameters represent the 24-hour output of the battery energy storage device; wherein the 73rd to 96th parameters represent the 24-hour output of the diesel generator.

[0040] Furthermore, the upper and lower limits of the search space in the initialization of step 4.1 are the upper and lower limits of the charge and discharge power obtained according to the operating capacity of each distributed power source, which serve as boundary constraints for each dimension.

[0041] Furthermore, the constraints are as follows:

[0042] Distributed power generation output constraints:

[0043] P i min≤P i ≤P i max (11)

[0044] Where, P i min, P i max are the lower and upper limits of the distributed power output, P i Provide power for distributed power generation;

[0045] Power balance constraints:

[0046]

[0047] Where, P Load is the microgrid load; PBS is the battery charging and discharging power, when P BS >0, indicating that the battery is discharged. BS When <0, it means the battery is charging;

[0048] Battery constraints:

[0049] P BS min≤P BS ≤P BS max (13)

[0050] E BS min≤E Bs ≤E BS max (14)

[0051] Where, P BS min, P BS max are the minimum and maximum charge and discharge power of the battery respectively; E BS min, E BS max are the minimum and maximum capacities of the battery respectively; E BS is the capacity of the battery.

[0052] Furthermore, there are two fitness functions in step 4.6:

[0053] The total operating economic cost fitness function is:

[0054]

[0055] Where T is the number of hours of the microgrid dispatch cycle; C PV (t), C WT (t), C BT (t), C DE (t), are the power generation costs of photovoltaic, wind power, energy storage, and diesel generators respectively; C EMI (x) is the emission cost;

[0056] Power generation cost:

[0057] C i (t) = CF i，t +IV i，t +OM i，t (16)

[0058] Among them, CF i，t is the fuel cost of distributed power source i at time t; IV i，t is the depreciation cost of distributed power source i converted to unit time; OM i，t is the maintenance cost of distributed generation i at time t;

[0059] Emission costs:

[0060] C EMI (t)=α*σ*P i，t (17)

[0061] Among them, C EMI (t) is the fuel cost of diesel generator i at time t, α is the emission price, σ is the emission coefficient Pr, P i，t is the rated power of the i-th distributed power supply;

[0062] The fitness function of the total wind and solar power consumption ratio is:

[0063] f(2)=2-(PV pro +WT pro ) (18)

[0064] Among them PV pro is the photovoltaic absorption ratio, WT pro is the wind power consumption ratio.

[0065] Furthermore, the optimal result in step 4.8 is obtained by normalizing the compromise solution of each data, and finally the solution with the minimum total fitness is selected as the optimal compromise solution;

[0066]

[0067] Where Cost.obj is the optimal compromise solution; Cost(i).f(1) is the fitness value of the objective function f(1) at the i-th hummingbird position; Cost.f(1)max is the maximum fitness value of the objective function f(1); and the same applies to f(2).

[0068] According to the obtained optimal result, the optimal economic dispatch data of the microgrid is obtained.

[0069] The beneficial effects of the present invention are embodied in:

[0070] The microgrid economic environment scheduling model provided by the present invention does not only consider the single objective function of the total operating cost of the microgrid system, but comprehensively considers the renewable energy consumption ratio and total operating cost while meeting the basic electricity demand of users.

[0071] Thanks to the application of a multi-objective algorithm, the present invention can not only consider the two objectives in this example, but can also be expanded to consider more objective functions, such as minimizing environmental costs and minimizing carbon emissions, when more microgrid data is obtained.

[0072] The present invention improves the multi-objective artificial hummingbird algorithm. The artificial hummingbird algorithm has few control parameters and is easy to calculate. By combining the advantages of chaotic initialization and linear inertia weight coefficient, the basic multi-objective artificial hummingbird algorithm is improved, and the global optimization ability and convergence speed of the multi-objective artificial hummingbird algorithm are enhanced, thereby finding the optimal scheduling solution more quickly. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 This is a flowchart of the microgrid economic dispatch model optimized by the improved multi-objective artificial hummingbird algorithm of the present invention;

[0074] Figure 2 Flowchart of the improved multi-objective artificial hummingbird algorithm of the present invention;

[0075] Figure 3 This is a 24-hour load demand diagram of a microgrid in a certain location in an example of the present invention;

[0076] Figure 4 is a Pareto frontier graph in an example of the present invention;

[0077] Figure 5 This is the 24h economic dispatch optimization diagram of the microgrid in the example of the present invention. DETAILED DESCRIPTION

[0078] The technical solutions in the examples of the present invention will be described clearly and completely below with reference to the accompanying drawings in the present invention.

[0079] like Figure 1 As shown, an example of the present invention introduces a microgrid economic dispatch optimization method based on a multi-objective artificial hummingbird algorithm, comprising the following steps:

[0080] 1) Obtain the installed capacity and operating data of the generator sets of the microgrid system, the tiered electricity price cost coefficient, the emission penalty coefficient, and the user demand load data.

[0081] 2) Establish an optimal dispatch model and objective function for the microgrid system;

[0082] 3) Determine the power constraints of each generator set in the microgrid system;

[0083] 4) Based on the aforementioned optimization scheduling model, objective function, and constraints, a novel bionic optimization algorithm, the Artificial Hummingbird algorithm, was introduced to optimize the microgrid economic scheduling model. The Artificial Hummingbird algorithm has excellent exploration and development capabilities, determining high-quality solutions with fewer control parameters, and can perform global optimization within the algorithm. Chaotic initialization and improved optimization were performed on the multi-objective Artificial Hummingbird algorithm, enhancing its ability to overcome local optimization issues.

[0084] The improved multi-objective artificial hummingbird algorithm specifically includes the following steps:

[0085] Step 1: Randomly generate an initial population through Sine chaos optimization and place n hummingbirds on n food sources. Sine chaos initialization replaces the random generation in MOAHA. The specific mapping method is shown in Equation (2). Randomly generated initial populations can easily cause the algorithm to fall into local optimality too early. Initializing the population through Sine chaos mapping makes the mayfly population evenly distributed within the upper and lower limits of the search space, improving the search performance of the artificial hummingbird algorithm:

[0086] x1=Low+r·(Up-Low) (1)

[0087]

[0088] Where Up and Low are the upper and lower bounds of the d-dimensional problem, r is a random vector in [0, 1], and x i represents the location of the i-th food source as a solution to the given problem, and a is a parameter with a value of (0, 4).

[0089] Step 2: Initialize the food source access table:

[0090]

[0091] Where, for i=j, VT ij =null indicates that the hummingbird feeds on its specific food source; i≠j, VT ij = 0 means that the jth food source has just been visited by the i-th hummingbird in the current iteration.

[0092] Step 3: Using an improved linear foraging behavior selection coefficient, we replace the original 50% probability in MOAHA to guide foraging behavior or territorial foraging behavior. This helps to focus on guiding foraging behavior to explore other spaces in the early iterations, and focus on territorial foraging behavior to find the local optimal solution in the later iterations.

[0093]

[0094]

[0095] Where W is the inertia coefficient of linear foraging behavior; W max and W min The maximum and minimum inertia coefficients are [0, 1] respectively; it is the number of iterations; Maxit is the maximum number of iterations; v i (t+1) is the candidate food source location of the i-th hummingbird at time t+1; x i (t) is the food source location of the i-th hummingbird at time t, x i,tar(t) is the location of the target food source that the i-th hummingbird intends to visit; D is the flight index coefficient, which takes the value of 0 or 1; a is the guiding factor, which follows the standard normal distribution N(0, 1); b is the territory factor, which follows the standard normal distribution N(0, 1).

[0096] Step 4: The hummingbird searches for its target food source using three methods. Under modified random probability, the hummingbird chooses to perform the following guided foraging behaviors:

[0097] Under the random probability of rand < W, the hummingbird chooses to perform the following guided foraging behavior:

[0098] v i (t+1)=x i,tar (t)+a·D·(x i (t)-x i,tar (t)) (7)

[0099] The position of the i-th food source is updated as follows:

[0100]

[0101] Where f(·) is the fitness value of the function. If the fitness of the candidate food source is lower than that of the current food source, the hummingbird will abandon the current food source and stay at the candidate food source generated by formula (7) to feed; update the visit table;

[0102] In the case of random probability rand ≥ W, the hummingbird chooses to perform the following territorial foraging behavior:

[0103] v i (t+1)=x i,tar (t)+b·D·x i (t) (9)

[0104] The location of the i-th food source is updated in the same way as in (8), and the visit table is updated;

[0105] Step 5: Determine the number of iterations. If the number of iterations exceeds the predetermined value of the migration coefficient 2n, the hummingbird will migrate to a new food source randomly generated in the entire search space. The hummingbird's position is updated as follows:

[0106] x i =Low+r·(Up-Low) (10)

[0107] Update visit list.

[0108] Step 6: Calculate the fitness function and obtain the Pareto solution.

[0109] Step 7: Perform non-dominated sorting to obtain the Pareto frontier.

[0110] Step 8: Based on the obtained optimal result, obtain the optimal economic dispatch data of the microgrid.

[0111] The above steps are reflected in Figure 2 In the algorithm flow chart.

[0112] A further solution is to initialize the population dimension i in Step 1 with 96 parameters derived from the 24-hour output of the four distributed generation sources in the microgrid system's economic dispatch. Parameters 1-24 represent the 24-hour output of photovoltaic cells; parameters 25-48 represent the 24-hour output of wind turbines; parameters 49-72 represent the 24-hour output of battery energy storage devices; and parameters 73-96 represent the 24-hour output of diesel generators.

[0113] A further solution is that the upper and lower limits of the search space in the initialization described in Step 1 are the upper and lower limits of the charging and discharging power obtained according to the operating capacity of each distributed power source, which serve as boundary constraints for each dimension.

[0114] A further solution is that the total operating economic cost fitness function is:

[0115]

[0116] Where T is the number of hours of the microgrid dispatch cycle; C PV (t), C WT (t), C BT (t), C DE (t), are the power generation costs of photovoltaic, wind power, energy storage, and diesel generators respectively; C EMI (x) is the emission cost;

[0117] Power generation cost:

[0118] C i (t) = CF i，t +IV i，t +OM i，t (12)

[0119] Among them, CF i，t is the fuel cost of distributed power source i at time t; IV i，t is the depreciation cost of distributed power source i converted to unit time; OM i，t is the maintenance cost of distributed generation i at time t;

[0120] Emission costs:

[0121] C EMI (t)=α*σ*P i，t (13)

[0122] Among them, CEMI (t) is the fuel cost of diesel generator i at time t, α is the emission price, σ is the emission coefficient Pr, P i，t is the rated power of the i-th distributed power supply;

[0123] A further solution is that the fitness function of the total wind and solar power consumption ratio is:

[0124] f(2)=2-(PV pro +WT pro ) (14)

[0125] Among them PV pro is the photovoltaic absorption ratio, WT pro is the wind power consumption ratio.

[0126] A further solution is that the model constraints are as follows:

[0127] Distributed power generation output constraints:

[0128] P i min≤P i ≤P i max (15)

[0129] Where, P i min, P i max are the lower and upper limits of the distributed power output, P i is the output power of distributed power source i.

[0130] Power balance constraints:

[0131]

[0132] Where, P Load is the microgrid load; P BS is the battery charging and discharging power, when P BS >0, indicating that the battery is discharged. BS When <0, it means the battery is charging.

[0133] Battery constraints:

[0134] P BS min≤P BS ≤P BS max (17)

[0135] E BS min≤E BS ≤E BS max (18)

[0136] Where, P BS min, P BSmax are the minimum and maximum charge and discharge power of the battery respectively; E BS min, E BS max are the minimum and maximum capacities of the battery respectively; P BS is the charging and discharging power of the battery; E BS is the capacity of the battery.

[0137] A further solution is that the optimal result is obtained by normalizing the compromise solution of each data, and finally the solution with the minimum total fitness is selected as the optimal compromise solution.

[0138]

[0139] Where Cost.obj is the optimal compromise solution; Cost(i).f(1) is the fitness value of the objective function f(1) at the i-th hummingbird position; Cost.f(1)max is the maximum fitness value of the objective function f(1); and the same applies to f(2).

[0140] According to the obtained optimal result, the optimal economic dispatch data of the microgrid is obtained.

[0141] In the present invention, matlab2018b programming is used to verify the above model, and the parameter values in the algorithm are as follows: the size of the artificial hummingbird population is nPop=100, the archive size is ArchiveSize=100, and the solution space dimension Dim=96.

[0142] Figure 3 This is a typical 24-hour load demand diagram of a microgrid in a certain place in the example of the present invention. It can be seen from the figure that the peak-to-valley difference of the load is large, and optimized scheduling is needed after wind power and photovoltaic power are connected.

[0143] Figure 4 It is a Pareto frontier diagram of the optimal solution obtained after iteration in the example of the present invention.

[0144] Figure 5 This is a 24-hour optimization scheduling diagram for a microgrid system using the improved multi-objective artificial hummingbird algorithm of the present invention. The diagram shows the optimized scheduling output of each device in the microgrid system every hour within 24 hours of the day. Renewable energy sources such as wind energy and solar energy are prioritized for absorption. When the battery energy storage system is insufficient to supply the load, the diesel generator is put into operation to supplement the electrical load after the absorption of wind energy and solar energy. Since wind energy and solar energy are uncertain and vary greatly every hour, in order to maintain power balance, if the microgrid generates too much electricity, it will be supplied to the battery for charging, and if it generates too little electricity, it will be discharged by the battery. The optimization scheduling method of the present invention can achieve a larger wind and solar absorption ratio, reduce the economic cost of microgrid operation, and ensure the stability and reliability of the microgrid system.

[0145] This paper investigates the 24-hour optimal energy dispatch problem for microgrid systems. While satisfying system constraints, it considers the economic, environmental, and reliability of the microgrid. A multi-objective optimal dispatch model is established that balances wind and solar power consumption with operating costs. The improved MOAHA algorithm is used to solve the multi-objective economic dispatch model for microgrids that considers wind and solar power consumption. Simulation results demonstrate that this model offers valuable insights and guidance for optimizing microgrid dispatch.

[0146] Finally, it should be noted that the above description only describes specific embodiments of the present invention in detail. However, the present invention is not limited to the specific embodiments described above. Equivalent modifications and substitutions made by those skilled in the art are also within the scope of the present invention. Therefore, equivalent changes and modifications made without departing from the spirit and scope of the present invention are encompassed within the scope of the present invention.

Claims

1. A microgrid economic dispatch optimization method based on an improved multi-objective artificial hummingbird algorithm, characterized by: The method includes: Step 1: Obtain the operating data of the generating units and the basic data of the load in the microgrid system with renewable energy. Among them, the microgrid system includes a wind turbine PV, a photovoltaic cell WT, a diesel generator DE, and a battery energy storage device BT; Step 2: Based on the operating data of the generating units and the basic data of the load, construct an economic dispatch model for the microgrid system with wind power, photovoltaic, diesel generators, and energy storage; Step 3: Determine the power constraint conditions of the microgrid system; Step 4: Improve the multi-objective artificial hummingbird algorithm MOAHA and combine it with the wind-solar output control strategy and the economic dispatch model, and then obtain the optimal solution for the economic dispatch of each distributed power source, so as to obtain the optimal solution and achieve the economic optimal dispatch of the microgrid system; The said Step 4 includes: Step 4.1: Instead of randomly generating in MOAHA, initialize through Sine chaos. Place n hummingbirds on n food sources. The specific mapping method is shown in Equation (2). Initialize the population through Sine chaos mapping so that the hummingbird population is evenly distributed within the upper and lower limits of the search space: x1 = Low + r·(Up - Low) (1) where Up and Low are the upper and lower bounds of the d-dimensional problem, r is a random vector in [0,1], and x i It represents the location of the i-th food source that is the solution to the given problem, and a is a parameter with a value of (0,4); Step 4.2: Initialize the food source access table: Where, for i=j, VT ij =null indicates that the hummingbird feeds on its specific food source; i≠j,VT ij = 0 means that the jth food source has just been visited by the i-th hummingbird in the current iteration; Step 4.3: Through the improved linear foraging behavior inertia coefficient, conduct guided foraging behavior or territorial foraging behavior; Where W is the inertia coefficient of linear foraging behavior; W max and W min The maximum and minimum inertia coefficients are [0,1] respectively; it is the number of iterations; Maxit is the maximum number of iterations; v i (t+1) is the candidate food source location of the i-th hummingbird at time t+1; x i (t) is the food source location of the i-th hummingbird at time t, x i,tar (t) is the location of the target food source that the i-th hummingbird intends to visit; D is the flight index coefficient, which is 0 or 1; a is the guide factor, which follows the standard normal distribution N(0,1); b is the territory factor, which follows the standard normal distribution N(0,1); Step 4.4: The hummingbird finds its target food source through three methods When the random probability rand < W, the hummingbird chooses to perform the following guided foraging behavior: v i (t+1)=x i,tar (t)+a·D·(x i (t)-x i,tar (t)) (7) The position of the i-th food source is updated as follows: Where f(·) is the function fitness value. If the fitness of the candidate food source is lower than the current food source, the hummingbird will abandon the current food source and stay at the candidate food source generated by Equation (7) to feed; update the visit table; When the random probability rand ≥ W, the hummingbird chooses to perform the following territorial foraging behavior: v i (t+1)=x i,tar (t)+b·D·x i (t) (9) The position update of the i-th food source is the same as Equation (8), and update the visit table; Step 4.5: Judge the number of iterations. If the number of iterations exceeds the predetermined value of the migration coefficient 2n, the hummingbird will migrate to a new food source randomly generated in the entire search space; the hummingbird position is updated as follows: x i =Low+r·(Up-Low) (10) Update the visit table; Step 4.6: Calculate the total operating economic cost fitness function and the total wind-solar consumption ratio fitness function to obtain the pareto solution; Step 4.7: Perform non-dominated sorting to obtain the pareto front; Step 4.8: Obtain the optimal result.

2. An optimized method for microgrid economic dispatch based on an improved multi-objective artificial hummingbird algorithm according to claim 1, characterized in that: In the said Step 4.1, the initialized population dimension is 96 parameters obtained from the 24-hour output of four distributed power sources in the microgrid system economic dispatch; among them, the 1st - 24th parameters represent the 24-hour output of the photovoltaic cell; among them, the 25th - 48th parameters represent the 24-hour output of the wind turbine; among them, the 49th - 72nd parameters represent the 24-hour output of the battery energy storage device; among them, the 73rd - 96th parameters represent the 24-hour output of the diesel generator.

3. The microgrid economic dispatch optimization method based on the improved multi-objective artificial hummingbird algorithm according to claim 2 is characterized by: In the initialization of step 4.1, the upper and lower limits of the search space are the upper and lower limits of the charge and discharge power obtained according to the operating capacity of each distributed power source, which serve as the boundary constraints of each dimension.

4. The microgrid economic dispatch optimization method based on the improved multi-objective artificial hummingbird algorithm according to claim 3 is characterized in that: The constraints are as follows: Distributed power generation output constraints: P i min≤P i ≤P i max (11) Where, P i min, P i max are the lower and upper limits of the distributed power output, P i Output for the i-th distributed generation; Power balance constraints: Where, P Load is the microgrid load; P BS is the battery charging and discharging power, when P BS >0, indicating that the battery is discharged. BS When <0, it means the battery is charging; Battery constraints: P BS min≤P BS ≤P BS max (13) AND BS min≤E BS ≤E BS max (14) Where, P BS min, P BS max are the minimum and maximum charge and discharge power of the battery respectively; E BS min, E BS max are the minimum and maximum capacities of the battery respectively; E BS is the capacity of the battery.

5. The microgrid economic dispatch optimization method based on the improved multi-objective artificial hummingbird algorithm according to claim 3 is characterized by: There are two fitness functions in step 4.6: The total operating economic cost fitness function is: Where T is the number of hours of the microgrid dispatch cycle; C PV (t),C WT (t), C BT (t),C DE (t), respectively, are the power generation costs of photovoltaic, wind power, energy storage, and diesel generators; C EMI (x) is the emission cost; Power generation cost: C i (t)=CF i，t +IV i，t +OM i，t (16) Among them, CF i，t is the fuel cost of distributed power source i at time t; IV i，t is the depreciation cost of distributed power source i converted to unit time; OM i，t is the maintenance cost of distributed generation i at time t; Emission costs: C EMI (t)=α*σ*P i，t (17) Among them, C EMI (t) is the fuel cost of diesel generator i at time t, α is the emission price, σ is the emission coefficient Pr, P i，t is the rated power of the i-th distributed power supply; The fitness function of the total wind and solar power consumption ratio is: f(2)=2-(PV pro +WT pro ) (18) Among them PV pro is the photovoltaic absorption ratio, WT pro is the wind power consumption ratio.

6. The microgrid economic dispatch optimization method based on the improved multi-objective artificial hummingbird algorithm according to claim 5 is characterized by: The optimal result in step 4.8 is obtained by normalizing the compromise solution of each data, and finally the solution with the minimum total fitness is selected as the optimal compromise solution; Where Cost.obj is the optimal compromise solution; Cost(i).f(1) is the fitness value of the objective function f(1) at the i-th hummingbird position; Cost.f(1)max is the maximum fitness value of the objective function f(1); Cost(i).f(2) is the fitness value of the objective function f(2) at the i-th hummingbird position; Cost.f(2)max is the maximum fitness value of the objective function f(2).

Images