AC / DC hybrid microgrid energy scheduling method based on improved particle swarm optimization
By improving the particle swarm algorithm, the optimal regulation of the energy supply and demand relationship of the microgrid is solved, and the problems of high operating costs of the microgrid and unstable new energy power generation are significantly reduced, the dependence of energy storage devices is improved, and the consumption of new energy electricity is improved. rate and customer satisfaction.
Patent Information
- Application Number
- CN202411510162.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-28
- Publication Date
- 2025-01-10
AI Technical Summary
In the prior art, the instability of microgrids in new energy power generation leads to high operating costs, and the optimization traversal ability of intelligent algorithms is low, which makes it easy to fall into precocious maturity, which reduces the optimal solution accuracy, resulting in excessive dependence and operation of energy storage devices. High cost.
The improved particle swarm algorithm is used to optimize the energy supply and demand relationship between the components of the microgrid. By constructing mathematical models and optimizing the objective function, the particle swarm algorithm is used to calculate the global optimal solution, reducing the dependence of energy storage devices, and improving new Local consumption rate and customer satisfaction of energy and electricity.
It significantly reduces the total operating cost of the microgrid, improves the local consumption rate and customer satisfaction of new energy electricity, and reduces the over-dependence and operating costs of energy storage devices.
Smart Images
Figure CN119298075A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of power grid operation, and in particular to an AC / DC hybrid microgrid energy scheduling method based on an improved particle swarm algorithm. Background Art
[0002] Microgrid is a new type of power system, which generally includes fixed loads, flexible loads, energy storage devices, renewable energy power generation devices and traditional power generation devices. When in use, the renewable energy power generation devices generate electricity and transmit electricity to fixed loads and flexible loads. Due to the instability of renewable energy power generation, the normal and stable operation of the microgrid is heavily dependent on energy storage devices and traditional generators, which will significantly increase the operating costs of the microgrid.
[0003] In the prior art, intelligent algorithms such as ant colonies and genetic algorithms are usually used for optimization. The supply and demand relationship of each component unit in the microgrid is regulated by the obtained optimal solution. However, the optimization traversal of the above algorithms is low, and it is easy to fall into premature optimization when performing target optimization, which reduces the accuracy of the optimal solution. This makes it impossible to improve the supply and demand relationship of each component unit under the regulated microgrid, and its operating cost is still high. Summary of the invention
[0004] In view of the shortcomings of the prior art, the present application provides an AC / DC hybrid microgrid energy scheduling method based on an improved particle swarm algorithm. The embodiment of the present application is a method of optimally regulating the energy supply and demand relationship between the various components of the microgrid through an improved particle swarm algorithm, which significantly reduces the microgrid's excessive reliance on energy storage devices, significantly reduces the total cost of the microgrid during operation, and improves the local consumption rate of new energy electricity and customer satisfaction.
[0005] The above application objectives of this application are achieved through the following technical solutions: An AC / DC hybrid microgrid energy scheduling method based on an improved particle swarm algorithm comprises the following steps: Obtaining a mathematical model of the power of each component unit of the microgrid, wherein the component units include a new energy power generation unit, an energy storage unit, a diesel power generation unit, and a flexible load unit; According to the obtained mathematical model of the power of each component unit of the microgrid, a microgrid operation index function is constructed. The microgrid operation index includes the economic cost C ECO , pollution control cost C POLLUTION , local consumption rate of new energy power generation units K PV And the load-side user satisfaction K ML ; Construct the optimization objective function: ; The particle swarm algorithm is used to calculate the global optimal solution of the objective function; The various components of the microgrid are regulated according to the calculated global optimal solution.
[0006] Optionally, divide a day evenly into several scheduling periods; Use the particle swarm algorithm to calculate the global optimal solution of the objective function in the current cycle; According to the global optimal solution in the current cycle, each component unit is regulated in the current cycle.
[0007] Optionally, the calculating the global optimal solution of the objective function by using a particle swarm algorithm includes: Initialize the particle swarm; The particle swarm is brought into the objective function for calculation to obtain the optimal fitness of the individual particles and the global optimal solution of the objective function.
[0008] Optionally, the initialization setting of the particle swarm includes: According to the objective function, a d-dimensional search space is established, and a particle swarm containing N particles is randomly constructed. Each particle is a d-dimensional vector, and each particle corresponds to an output flight speed. The i-th particle and its corresponding flight speed are as follows: .
[0009] Optionally, obtaining the optimal fitness of individual particles and the global optimal solution of the objective function includes: Each particle Substitute the objective function for calculation, and the obtained function value is recorded as the optimal fitness of the particle individual ; Take the minimum objective function value among all particle swarms as the global optimal .
[0010] Optionally, the calculating the global optimal solution of the objective function by using the particle swarm algorithm further includes: Update the particle swarm and generate new particles; Screen the newly generated particles according to the fitness probability; The screened new particles are brought into the objective function calculation to obtain the optimal fitness of the new particles and the global optimal solution of the objective function; Compare and update the global optimal solution of the objective function; Determine whether the current number of iterations meets the maximum iteration condition. If so, end the iteration and output the global optimal solution. Otherwise, return to repeat the iteration update.
[0011] Optionally, updating the particle swarm includes: Dynamically update the inertia weight of particle velocity update in the particle swarm algorithm, and dynamically update the weight for: Among them, w max is the maximum inertia weight, w min is the minimum inertia weight, N w Current iteration number, N w is the maximum number of iterations, and the current inertia weight is calculated in a linear decreasing manner; is the fitness ranking of the particle in the example group, is the size of the particle group.
[0012] Optionally, updating the particle swarm further includes: Dynamically adjust the acceleration factor of particle speed update in the particle swarm algorithm. and They are: Among them, d 1s and d 1e , d 2s and d 2e are the starting and ending values of the acceleration factors d1 and d2, respectively. η1 and η2 are random numbers in the interval [0,1]. η1 and β2 are acceleration influence factors, β1+β2=1. and is the acceleration factor weight coefficient; Calculate the updated particle flight speed and the position of the particle at the k+1th iteration.
[0013] Optionally, generating new particles includes: During the k-th iteration, the particle position whose average fitness is less than the k-th iteration is Using chaotic systems for new particle calculations: in, is a control parameter whose value range is μ∈(2,4].
[0014] Optionally, comparing and updating the global optimal solution of the objective function includes: The particle position obtained by the kth update Substitute the objective function to calculate the fitness; The local optimal solution obtained in the kth iteration The global optimal solution obtained by k-1 times Make comparisons; Update the global optimal solution based on the comparison results.
[0015] In summary, this application has the following beneficial technical effects: The embodiment of the present application is a method of optimally regulating the energy supply and demand relationship between the various components of the microgrid through an improved particle swarm algorithm, which significantly reduces the microgrid's excessive reliance on energy storage devices, significantly reduces the total cost of the microgrid during operation, and improves the local consumption rate of new energy electricity and customer satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 It is a flow chart of an improved particle swarm algorithm according to an embodiment of the present application. DETAILED DESCRIPTION
[0017] The present application is described in further detail below.
[0018] In order to more clearly understand the technical solutions presented in the embodiments of the present application, a brief introduction to the existing microgrid and its scheduling method is first given.
[0019] In recent years, in order to cope with increasingly severe environmental problems such as the greenhouse effect and acid rain, and to promote the development of high-quality energy and electricity, a large number of renewable energy sources have been connected to the power grid, and a new power system with source-grid-load-storage interaction and multi-energy complementarity has emerged. As a special new power system, microgrids can achieve flexible complementarity of multiple energy flows, effectively smooth out power fluctuations and increase the local absorption rate of renewable energy such as photovoltaics. With strong autonomy and independence, they have attracted much attention.
[0020] With the development of power electronics technology and new energy technology, flexible and adjustable loads such as super charging devices composed of power electronic devices and new energy storage devices have the characteristics of fast and flexible adjustment and peak shaving and valley filling. After being connected to the power system, the energy management system can coordinate the allocation of multi-terminal energy flows of sources, grids, loads and storage, optimize the energy supply structure, and improve the stability and flexibility of microgrids.
[0021] The AC / DC hybrid microgrid is mainly composed of wind power generation, photovoltaic power generation, energy storage and diesel generators. There is energy coupling between multiple energy flows, and the energy industry structure of the village-level AC / DC microgrid is highly decentralized and cyclically volatile. The power supply cost does not match the economic benefits, and the existing strategies are not very applicable to village-level microgrids. At present, in terms of energy distribution management of AC / DC microgrids, microgrids are prone to rely heavily on new energy storage devices such as lithium batteries and supercapacitors to smooth grid fluctuations, which often causes overcharging and over-discharging of energy storage units, seriously affecting the energy storage efficiency and operating life of energy storage devices such as lithium batteries and supercapacitors; and excessive use of diesel generators will result in high fuel costs and pollution control costs. The existing energy management strategy still focuses on energy storage devices, and does not fully utilize the supporting and stabilizing role of adjustable loads on microgrids. At present, mature intelligent algorithms such as ant colonies and genetic algorithms are often used for optimization when pre-allocating energy for microgrids. However, the above algorithms have low optimization traversal and are prone to premature optimization when performing target optimization, which reduces the accuracy of the optimal solution.
[0022] Generally speaking, the AC / DC microgrid energy management strategy in the prior art has the following defects and deficiencies: Over-reliance on energy storage devices such as supercapacitors and lithium batteries to smooth out peak loads in the power grid can easily lead to excessive charging and discharging of energy storage, reducing the service life and performance of energy storage equipment.
[0023] Unreasonable energy allocation can easily cause diesel generators to over-generate electricity, resulting in high fuel costs and pollution control costs, and increase operating costs.
[0024] The load distribution density of village-level AC / DC microgrids is low and has strong periodic fluctuations, and the local consumption rate of renewable energy is reduced.
[0025] Flexible and adjustable loads such as super charging piles do not participate deeply in the energy regulation and dispatch of microgrids, which reduces the energy autonomy of microgrids.
[0026] The existing commonly used optimization algorithms have low population ergodicity and randomness, are prone to fall into local optimal solutions, have low optimization accuracy, and reduced convergence.
[0027] In summary, according to the power characteristics of village-level AC / DC hybrid microgrids, it is urgent to propose a new energy management strategy to give full play to the scheduling potential of flexible and adjustable loads such as super charging piles, limit the fluctuation of energy storage battery management system, and improve the economic efficiency of power generation and the local consumption rate of renewable energy.
[0028] In response to the above problems, an embodiment of the present application provides an AC / DC hybrid microgrid energy scheduling method based on an improved particle swarm algorithm, which reduces the existing microgrid's excessive reliance on energy storage devices for grid peak shaving and valley filling, reduces the total operating cost of the microgrid, and improves the local absorption rate and customer satisfaction of the microgrid.
[0029] An AC / DC hybrid microgrid energy scheduling method based on an improved particle swarm algorithm comprises the following steps: S101, obtaining a mathematical model of the power of each component unit of the microgrid, wherein the component units include a new energy power generation unit, an energy storage unit, a diesel power generation unit, and a flexible load unit; S102, constructing a microgrid operation index function according to the acquired mathematical model of the power of each component unit of the microgrid, wherein the microgrid operation index includes economic cost CECO, pollution control cost CPOLLUTION, local consumption rate of new energy power generation unit KPV and load-end user satisfaction KML; S103, construct optimization objective function: ; S104, using a particle swarm algorithm to calculate a global optimal solution of the objective function; S105. Regulate each component unit of the microgrid according to the calculated global optimal solution.
[0030] The following is a further introduction based on specific usage scenarios.
[0031] In step S101, a mathematical model of the power of each component unit of the microgrid is obtained. The component units include a new energy power generation unit, an energy storage unit, a diesel power generation unit, and a flexible load unit. First, let's look at the new energy power generation unit. The embodiment of this application takes a photovoltaic power generation system as an example: The power and output voltage of the photovoltaic power generation system are and , the conductance and conductance increment are and , then the PU characteristic curve of the photovoltaic power generation system is: (1) (2) Formula (2) is the maximum power point of photovoltaic power generation. At this time, the recorded photovoltaic output power is .
[0032] The energy storage unit maintains the constant DC bus voltage by adjusting the power flow and its size, and absorbs the remaining energy in the microgrid. In order to avoid equipment loss caused by excessive charging and discharging, it is equipped with a battery management system (SOC) to monitor the charge state of the energy storage unit. The energy storage system plays an energy buffering role. The energy storage charging and discharging model is: (3) In formula (2), E is the nominal capacity of the energy storage system, is the output power of the energy storage system, Indicates the energy storage system discharge, Indicates that the energy storage system is charged. Indicates the efficiency of the energy storage system; Due to the intermittent nature of renewable energy generation and the volatility of user demand, in addition to using batteries as energy buffers in microgrids, diesel generators are also required as backup power sources. The output power of diesel generators can be manually controlled according to the real-time needs of the system, making it a controllable power source. Under the premise of ensuring stable output power and speed, the fuel consumption of diesel generators is calculated as follows: (4) In formula (4), F1 represents the fuel consumption of the diesel generator (L / h), PN(t) represents the rated power of the diesel generator (W), Pd(t) represents the output power of the diesel generator when it is running at time t (W), and a1 and a2 are the parameters of the corresponding fuel consumption-power curve; Since diesel generators use fossil fuels to generate electricity, the photovoltaic power generation device in the microgrid is a renewable clean energy, and its output pollutants can be ignored. Therefore, diesel generators are considered the only source of pollutant emissions. The pollutants they emit mainly include CO2, SO2, and unburned fuel. The fuel consumption F1 of the diesel generator directly determines the pollutant emission F2. The calculation formula of F2 is as follows: (5) Demand side user load Total output power of photovoltaic device Subtract and get the net load power of the system : (6) The microgrid involved in this patent includes a flexible load unit. The embodiment of this application takes a super charging pile as an example. The super charging pile and the energy storage unit jointly perform power regulation. When the user-side load exceeds the total energy supply of the microgrid, the super charging pile stops providing power to loads such as electric vehicles to balance the grid load; when the user-side load demand is low, the super charging pile can coordinate with the energy storage to store electricity. The specific principle of the super charging pile is as follows: (7) In formula (7) is the rated charging and discharging power of the super charging pile, It is the charging and discharging control parameter of the super charging device. The specific value selection is as follows: (8) Since there is no energy exchange between the microgrid and the large power grid in the embodiment of the present application, the output power of the energy storage unit is , diesel generator power 、Super charging pile charging and discharging power and net load power The following relationship exists: (9) Then, step 102 is executed to construct a microgrid operation index function according to the acquired mathematical model of the power of each component unit of the microgrid, wherein the microgrid operation index includes economic cost CECO, pollution control cost CPOLLUTION, local consumption rate of new energy power generation unit KPV and load-end user satisfaction KML; First, let’s look at the economic cost. The operating economic cost of the microgrid mainly includes: the operating and maintenance costs of the distributed photovoltaic power generation units, energy storage units, and super charging piles, as well as the diesel engine fuel usage costs. The specific expressions are as follows: (10) In formula (10), is the total economic cost of the microgrid, which is mainly composed of the microgrid operation and maintenance cost and microgrid operation fuel costs Composition, microgrid operation and maintenance costs It is composed of the maintenance costs required for each distributed power source, energy storage and super charging pile. The specific calculation method is as shown in formula (11): (11) In the above formula, is the unit maintenance cost of each equipment unit, T is 24 operation cycles, It is the running time in one running cycle.
[0033] According to formula (4), the fuel cost of the microgrid diesel generator can be obtained as follows: (12) in, The fuel price factor required to generate electricity for oil-fired generators; Let’s look at the pollution control cost. When the microgrid system is running, the environmental pollution caused by wind and solar power generation, which is represented by clean energy, is almost zero. Diesel generators provide a steady stream of electricity by consuming fuel in daily operation, which will produce more pollutants. These pollutants can only be discharged after they are treated to meet the standards. Therefore, the main consideration is the treatment cost of the pollutants such as CO2, SO2, and NOX. According to the pollutant emission calculation formula shown in formula (5), the pollution control cost is The calculation is as follows: (13) In the formula, KPC is the pollution control coefficient; For the local absorption rate of photovoltaic power generation devices, the actual power generation of photovoltaic power generation at time t is subtracted from the actual power consumption, and the difference and the proportion of actual power generation are used as indicators to evaluate the degree of photovoltaic absorption. The larger the value of the indicator, the lower the degree of local absorption of photovoltaic energy; the smaller the value of the indicator, the higher the degree of local absorption of photovoltaic energy: (14) In the above formula, is the actual power consumption of the i-th photovoltaic; The loads mentioned in the embodiments of the present application include fixed loads and flexible adjustable loads, and the load-side user satisfaction is specifically expressed as follows: (15) (16) in, is the necessary fixed load in the system at time t, It is the actual power demand of the flexible load. is the power demand of the i-th flexible load at time t, is the i-th flexible load switching parameter, When the load is cut off, it means that the load is connected, otherwise, the flexible adjustable load can be reasonably cut in and out according to the total energy demand of the microgrid distributed power supply and the load end at the current moment to achieve the supply and demand balance of the microgrid. When the flexible load is cut in and out, the user satisfaction should still be improved as much as possible. The embodiment of the present application designs a load satisfaction index as follows: (17) In combination with the establishment of the above model, in step 103, an optimization objective function is constructed: (18) In step S104, a particle swarm algorithm is used to calculate the global optimal solution of the objective function: First, the particle swarm is initialized. According to the optimization parameters of the objective function shown in formula (18), a d-dimensional search space is established. A particle swarm containing N particles is randomly constructed. Each particle is a d-dimensional vector, and each particle corresponds to an output flight speed. The i-th particle and its corresponding flight speed are as follows: (19) (20) Then the fitness evaluation is performed and each particle Substitute the objective function shown in formula (18) for calculation, and the obtained function value is recorded as the optimal fitness of the particle individual , and take the minimum objective function value among all particle swarms as the global optimal .
[0034] The update expression of the velocity of the i-th particle in the particle swarm algorithm is: (twenty one) In the above formula is the inertia weight, is the number of iterations, and is the acceleration factor, and is a random constant between 0 and 1.
[0035] In order to improve the global optimization ability of the particle swarm and achieve fast traversal, the embodiment of the present application will adjust the fixed weight in formula (21) Dynamic update, the specific formula is as follows: (twenty two) Where wmax is the maximum inertia weight, wmin is the minimum inertia weight, nw is the current number of iterations, Nw is the maximum number of iterations, and the current inertia weight is calculated in a linear decreasing manner; is the fitness ranking of the particle in the example group, is the size of the particle group.
[0036] In order to improve the dynamic iteration speed of the optimization algorithm, the acceleration factor and A dynamic adjustment factor is introduced to ensure the rapid convergence of the particle swarm. The improved acceleration factor is: (twenty three) Under this method, the acceleration factor and A larger value can be selected for initialization. d1s and d1e, d2s and d2e are the starting and ending values of the acceleration factors d1 and d2 respectively, η1 and η2 are random numbers in [0,1], β1 and β2 are acceleration influence factors, β1+β2=1, and is the acceleration factor weight coefficient, and the random inertia weight related; Substitute equations (22) and (23) into equation (21) to calculate the updated particle flight speed: .according to The position of the i-th particle at the k+1th iteration can be calculated as: (twenty four) For the latest particle generated In order to increase the diversity of particle samples and achieve global optimization, the embodiment of the present application uses chaotic motion to generate new individuals. In the process of k iterations, the particle positions that are less than the average fitness of the kth iteration are The Logistic chaotic system shown in formula (21) is used to calculate new particles: (25) In the above formula is the control parameter, and its value range is μ∈(2,4]; Due to the chaotic mapping method, the number of new particles generated by m particles will be greater than m. The roulette strategy is further introduced to select particles according to the proportional probability. The particles generated by self-iteration and the particles generated by chaotic mapping are reselected according to the fitness probability. The fitness of individual i is , the fitness expression is: (26) Then the probability of individual i being selected is: (27) The particle position obtained by the kth update Substitute the objective function to calculate the fitness and use the local optimal solution obtained in the kth iteration as The global optimal solution obtained by k-1 times Compare and update the global optimal solution; (28) Determine whether the current number of iterations meets the maximum iteration condition. If so, end the iteration and output the global optimal solution. Otherwise, return to repeat the iteration update.
[0037] Based on the global optimal solution obtained, the energy storage, super charging piles and flexible adjustable loads in the microgrid are controlled in real time to achieve economic and stable control of the system, thereby reducing the existing microgrid's excessive reliance on energy storage devices for peak shaving and valley filling, reducing the total operating cost of the microgrid and improving the microgrid's local absorption rate and customer satisfaction.
[0038] In general, the embodiment of the present application is a method of optimally regulating the energy supply and demand relationship between the various components of the microgrid through an improved particle swarm algorithm, which significantly reduces the microgrid's excessive reliance on energy storage devices, significantly reduces the total cost of the microgrid operation process, and improves the local consumption rate of new energy electricity and customer satisfaction.
[0039] The embodiments of this specific implementation method are all preferred embodiments of the present application, and are not intended to limit the protection scope of the present application. Therefore, all equivalent changes made based on the structure, shape, and principle of the present application should be included in the protection scope of the present application.
Claims
1. An AC / DC hybrid microgrid energy dispatching method based on an improved particle swarm algorithm is characterized by: include: Obtaining a mathematical model of the power of each component unit of the microgrid, wherein the component units include a new energy power generation unit, an energy storage unit, a diesel power generation unit, and a flexible load unit; According to the obtained mathematical model of the power of each component unit of the microgrid, a microgrid operation index function is constructed. The microgrid operation index includes the economic cost C ECO , pollution control cost C POLLUTION , local consumption rate of new energy power generation units K PV And the load-side user satisfaction K ML ; Construct the optimization objective function: ; The particle swarm algorithm is used to calculate the global optimal solution of the objective function; The various components of the microgrid are regulated according to the calculated global optimal solution.
2. A method according to claim 1, characterized in that Divide a day into several scheduling cycles evenly; Use the particle swarm algorithm to calculate the global optimal solution of the objective function in the current cycle; According to the global optimal solution in the current cycle, each component unit is regulated in the current cycle.
3. A method according to claim 2, characterized in that The method of calculating the global optimal solution of the objective function using the particle swarm algorithm includes: Initialize the particle swarm; The particle swarm is brought into the objective function for calculation to obtain the optimal fitness of the individual particles and the global optimal solution of the objective function.
4. A method according to claim 3, characterized in that The initialization setting of the particle swarm includes: According to the objective function, a d-dimensional search space is established, and a particle swarm containing N particles is randomly constructed. Each particle is a d-dimensional vector, and each particle corresponds to an output flight speed. The i-th particle and its corresponding flight speed are as follows: 。 5. A method according to claim 3, characterized in that: The method of obtaining the optimal fitness of individual particles and the global optimal solution of the objective function includes: Each particle Substitute the objective function for calculation, and the obtained function value is recorded as the optimal fitness of the particle individual ; Take the minimum objective function value among all particle swarms as the global optimal .
6. A method according to claim 3, characterized in that: The method of calculating the global optimal solution of the objective function using the particle swarm algorithm also includes: Update the particle swarm and generate new particles; Screen the newly generated particles according to the fitness probability; The screened new particles are brought into the objective function calculation to obtain the optimal fitness of the new particles and the global optimal solution of the objective function; Compare and update the global optimal solution of the objective function; Determine whether the current number of iterations meets the maximum iteration condition. If so, end the iteration and output the global optimal solution. Otherwise, return to repeat the iteration update.
7. A method according to claim 6, characterized in that The updating of the particle swarm comprises: Dynamically update the inertia weight of particle velocity update in the particle swarm algorithm, and dynamically update the weight for: Among them, w max is the maximum inertia weight, w min is the minimum inertia weight, N w Current iteration number, N w is the maximum number of iterations, and the current inertia weight is calculated in a linear decreasing manner; is the fitness ranking of the particle in the example group, is the size of the particle group.
8. A method according to claim 7, characterized in that The updating of the particle swarm further comprises: Dynamically adjust the acceleration factor of particle speed update in the particle swarm algorithm. and They are: Among them, d 1s and d 1e , d 2s and d 2e are the starting and ending values of the acceleration factors d1 and d2, respectively. η1 and η2 are random numbers in the interval [0,1]. η1 and β2 are acceleration influence factors, β1+β2=1. and is the acceleration factor weight coefficient; Calculate the updated particle flight speed and the position of the particle at the k+1th iteration.
9. A method according to claim 6, characterized in that: The generating of new particles comprises: During the k-th iteration, the particle position whose average fitness is less than the k-th iteration is Using chaotic systems for new particle calculations: in, is a control parameter whose value range is μ∈(2,4].
10. A method according to claim 6, characterized in that Comparison and update of the global optimal solution of the objective function include: The particle position obtained by the kth update Substitute the objective function to calculate the fitness; The local optimal solution obtained in the kth iteration The global optimal solution obtained by k-1 times Make comparisons; Update the global optimal solution based on the comparison results.