Multi-objective game algorithm for optimal scheduling of electric vehicle charging load

Abstract

The present invention discloses a multi-objective game algorithm for optimizing the scheduling of electric vehicle charging loads, and the steps are as follows: 1. Establish a mathematical model of electricity price and voltage drop of the public node where the electric vehicle charging station is located; 2. Take the electricity price and voltage drop as the objective function, adopt a weighted method based on two-person zero-sum game, convert the multi-objective optimization scheduling problem into a single-objective optimization scheduling problem, determine the relevant constraints and game form, and give an optimization scheduling plan; 3. Use a simulated annealing particle swarm optimization algorithm based on a particle swarm optimization method to balance the goals of the car charging station and the consumer, and find a balanced solution; 4. Set the maximum number of iterations, the particle swarm fitness threshold or the minimum value set by the particle swarm; output the optimal charging power allocation plan when the stop condition is met, and return to continue iterating if it is not met, and finally obtain the optimal solution. The present invention uses electricity prices to guide both parties and obtain the optimal strategy under various scene conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of electric vehicle management, and in particular to a multi-objective game algorithm for optimizing the scheduling of electric vehicle charging loads. Background Art

[0002] Although the electric vehicle industry is developing rapidly and the number of electric vehicles in use is increasing rapidly, electric vehicles are idle most of the time, and the charging time can be flexibly adjusted. When they are connected to a microgrid, they can also be used as a storage power source to sell electricity to the system.

[0003] At the same time, large-scale car charging behavior will have an impact on the normal operation of the power grid, causing power loss and voltage fluctuations.

[0004] Moreover, in the grid peak-shaving, energy and backup market scenarios, the effective and flexible scheduling of resources such as electric vehicles, temperature control loads and distributed energy storage can greatly alleviate the impact of new energy access on the power system.

[0005] Therefore, how to use charging users as the main body and use electricity prices to guide both parties to obtain the optimal strategy under various scenario conditions has become a technical problem that technical personnel in this field urgently need to solve. Summary of the invention

[0006] In view of the above-mentioned defects of the prior art, the present invention provides a multi-objective game algorithm for optimizing the scheduling of electric vehicle charging loads, the purpose of which is to take charging users as the main body and use electricity prices to guide both parties, so as to obtain the optimal strategy under various scenario conditions.

[0007] To achieve the above object, the present invention discloses a multi-objective game algorithm for optimizing the scheduling of electric vehicle charging loads, comprising the following steps:

[0008] Step 1: Establish a mathematical model of the electricity price and voltage drop ΔV at the public node where the electric vehicle charging station is located;

[0009] Step 2: Taking the electricity price and the voltage drop ΔV of the public node where the electric vehicle charging station is located as the objective function, a weighted method based on a two-person zero-sum game is adopted to convert the multi-objective optimization scheduling problem into a single-objective optimization scheduling problem, determine the relevant constraints and game forms, and provide an optimization scheduling plan;

[0010] Step 3: Use simulated annealing particle swarm optimization algorithm based on particle swarm optimization method to balance the goals of car charging stations and consumers and find a balanced solution;

[0011] Step 4, set the maximum number of iterations, the particle swarm fitness threshold or the minimum value set; output the optimal charging power allocation plan when the stop condition is met, and realize the optimal strategy combination between the voltage drop at the public connection point of the electric vehicle charging station and the charging electricity price at the public connection point. If it is not met, return to step 3 to continue iterating and finally obtain the Pareto optimal solution.

[0012] Preferably, step 1 specifically comprises: taking the minimum voltage drop of the public node where the electric vehicle charging station is located and the minimum charging electricity price of the public connection point as optimization objectives, determining the key variables affecting the grid voltage and the charging electricity price;

[0013] The key variables include the charging power of each charging pile, the electricity price, the voltage drop ΔV, the payment ability of the power supplier and the payment ability of the consumer;

[0014] Wherein, the voltage drop ΔV is the voltage drop amount of the common connection point;

[0015] Then the mathematical model of the voltage drop ΔV at the common connection point is:

[0016]

[0017] Among them, α is the proportional constant of voltage drop; P i is the charging power of the i-th charging pile; R is the resistance of the grid line; n is the total number of charging piles;

[0018] The mathematical model of the electricity price c is:

[0019] c=f prise (ΔV) = β·ΔV;

[0020] Wherein, β is the proportionality coefficient between the electricity price and the voltage drop ΔV;

[0021] The payment function of the electricity supplier is:

[0022]

[0023] The consumer's payment function is:

[0024]

[0025] The optimization objective function is:

[0026]

[0027] Among them, the charging power limit of the charging pile is: P min ≤P i ≤P max

[0028] The voltage stability constraint is: ΔV≤ΔV max

[0029] The revenue limit of the power supplier of the power grid is: c ≥ c min .

[0030] More preferably, step 2 is: balancing the two objective functions using a weighted method based on a two-person zero-sum game, the formula of which is:

[0031]

[0032] Among them, λ i is the weighting coefficient, indicating the importance attached to each goal;

[0033] The original multi-objective optimization problem has two objectives, namely U supply and U consumer , then the weighting coefficients are λ 1 and λ 2 , and λ 1 +λ 2 =1;

[0034] When optimizing objective i alone, the optimal solution to the original problem is denoted as The set of optimal solutions is denoted as X * ;

[0035] There are two players in the game, one of whom is 1 Select one strategy as its own strategy, and the other participant selects a set X from the optimal solution * Choose a solution as your strategy;

[0036] The goal of the game is to find an equilibrium strategy between the electricity supplier and the charging consumer so that the objective functions of both participants are stable at the end of the game.

[0037] More preferably, step 3 is specifically as follows: initializing a group of particles; making the solution of each particle represent a variable combination, each having a position representing charging power and a speed, and then searching in the solution space by the particle, the particle updates the position according to its own historical best solution and the best solution of all particles;

[0038] Among them, position represents the value of the current solution, and speed represents the moving speed of the particle in the solution space;

[0039] The goal of the particle is to find the game equilibrium between power suppliers and charging consumers by optimizing the allocation of charging power, thereby achieving the optimal compromise of the payment function;

[0040] The particle swarm optimization algorithm searches for the global optimal solution by updating the speed and position of particles. The speed update formula is:

[0041] V i (t+1)=ω·V i (t)+c 1 ·r 1 ·(P best -P i )+c 2 ·r 2 ·(P best -P i );

[0042] Among them, ω is the inertia weight, which controls the influence of the particle's historical velocity; c 1 and c 2 is the acceleration constant, controlling the particle's movement toward the personal best position and the global best position; r 1 and r 2 It is a random number, which increases the randomness of the search;

[0043] The position update formula is: P i (t+1)=P i (t)+V i (t+1);

[0044] At the same time, the simulated annealing mechanism is introduced, that is, the probability of accepting the solution is: if the new solution P i The fitness value of (t+1) is better than the old solution P i (t) is better, then directly accept this solution;

[0045] If the new solution P i The fitness value of (t+1) is not as good as the old solution P i (t), then the probability of acceptance using the Metropolis criterion is:

[0046]

[0047] E=λ 1 U supply +λ 2 U consumer

[0048] Among them, E is the current function value, that is, the energy of the solution; E new is the energy of the new solution; T is the temperature parameter, which is used to control the amplitude of the disturbance. As the iteration proceeds, the temperature gradually decreases and the disturbance decreases, thereby refining the search.

[0049] Beneficial effects of the present invention:

[0050] The present invention combines a two-person zero-sum game model with a simulated annealing particle swarm optimization algorithm (SAPSO) to solve multi-objective game problems. The game model can be used to clarify the conflicts and interactions between participants, while using the global search and global optimization capabilities of SAPSO to avoid local optimal solutions and find compromise solutions between multiple objectives. SAPSO can efficiently handle multiple conflicting objectives (such as voltage drop and charging electricity price), and approach the Pareto optimal solution through the optimization process, providing the optimal strategy combination for game participants, with strong flexibility and scalability, and flexibly adjust the electricity charges corresponding to different charging efficiencies according to the node voltage drop to balance the grid load and the income of the car charging station.

[0051] The concept, specific structure and technical effects of the present invention will be further described below in conjunction with the accompanying drawings to fully understand the purpose, characteristics and effects of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A flow chart showing an embodiment of the present invention is shown.

[0053] Figure 2 The execution flow of step 4 in one embodiment of the present invention is shown. DETAILED DESCRIPTION

[0054] Example

[0055] like Figure 1 and Figure 2 As shown in FIG. 1 , the multi-objective game algorithm for optimizing the scheduling of electric vehicle charging load includes the following steps:

[0056] Step 1: Establish a mathematical model of the electricity price and voltage drop ΔV at the public node where the electric vehicle charging station is located;

[0057] Step 2: Taking the electricity price and voltage drop ΔV of the public node where the electric vehicle charging station is located as the objective function, a weighted method based on a two-person zero-sum game is used to convert the multi-objective optimization scheduling problem into a single-objective optimization scheduling problem, determine the relevant constraints and game forms, and give an optimization scheduling plan;

[0058] Step 3: Use simulated annealing particle swarm optimization algorithm based on particle swarm optimization method to balance the goals of car charging stations and consumers and find a balanced solution;

[0059] Step 4, set the maximum number of iterations, the particle swarm fitness threshold or the minimum value set; output the optimal charging power allocation plan when the stop condition is met, and realize the optimal strategy combination between the voltage drop at the public connection point of the electric vehicle charging station and the charging electricity price at the public connection point. If it is not met, return to step 3 to continue iterating and finally obtain the Pareto optimal solution.

[0060] The master-slave game optimization scheduling strategy for the interaction between electric vehicle charging stations and electric vehicles takes charging users as the main body and uses electricity prices to guide both parties, so as to obtain the optimal strategy under various scenario conditions. The site selected by the user for charging is used as a variable, and a game model is established based on the minimum value of the voltage drop at the public connection point of the electric vehicle charging station and the charging electricity price at the public connection point.

[0061] Step 1 of the present invention analyzes the daily behavior characteristics of electric vehicles, collects and statistics the main influencing factors in the charging process of electric vehicles, and establishes a mathematical model of the electricity price and voltage drop ΔV of the electric vehicle charging station.

[0062] The construction of the two-person zero-sum game model and the design of the objective function of the present invention regard the power supplier and the charging consumer as two participants in the game. The power supplier affects the voltage drop and the electricity price by setting the electricity price and the charging power, while the consumer chooses the charging power according to the electricity price and voltage stability.

[0063] The global search and annealing mechanism of the simulated annealing particle swarm optimization algorithm (SAPSO) of the present invention, in order to avoid local optimum, simulated annealing introduces temperature control and random perturbations, accepts more bad solutions in the early stage of the algorithm to expand the search range, and gradually focuses on the global optimal solution as the temperature decreases. The annealing mechanism increases the exploration ability of the algorithm and enhances the global search ability of particle swarm optimization.

[0064] In some embodiments, step 1 specifically includes: taking the minimum voltage drop of the public node where the electric vehicle charging station is located and the minimum charging electricity price of the public connection point as the optimization objectives, determining the key variables that affect the grid voltage and the charging electricity price;

[0065] The key variables include the charging power of each charging pile, the electricity price, the voltage drop ΔV, the payment ability of the power supplier and the payment ability of the consumer;

[0066] Wherein, the voltage drop ΔV is the voltage drop amount of the common connection point;

[0067] Then the mathematical model of the voltage drop ΔV at the common connection point is:

[0068]

[0069] Among them, α is the proportional constant of voltage drop; P i is the charging power of the i-th charging pile; R is the resistance of the grid line; n is the total number of charging piles;

[0070] The mathematical model of the electricity price c is:

[0071] c=f prise(ΔV) = β·ΔV;

[0072] Wherein, β is the proportionality coefficient between the electricity price and the voltage drop ΔV;

[0073] The payment function of the electricity supplier is:

[0074]

[0075] The consumer's payment function is:

[0076]

[0077] The optimization objective function is:

[0078]

[0079] Among them, the charging power limit of the charging pile is: P min ≤P i ≤P max

[0080] The voltage stability constraint is: ΔV≤ΔV max

[0081] The revenue limit of the power supplier of the power grid is: c ≥ c min

[0082] In some embodiments, step 2 is: balancing the two objective functions using a weighted method based on a two-person zero-sum game, the formula of which is:

[0083]

[0084] Among them, λ i is the weighting coefficient, indicating the importance attached to each goal;

[0085] The original multi-objective optimization problem has two objectives, namely U supply and U consumer , then the weighting coefficients are λ 1 and λ 2 , and λ 1 +λ 2 =1;

[0086] When optimizing objective i alone, the optimal solution to the original problem is denoted as The set of optimal solutions is denoted as X * ;

[0087] There are two players in the game, one of whom is 1 Select one strategy as its own strategy, and the other participant selects a set X from the optimal solution * Choose a solution as your strategy;

[0088] The goal of the game is to find an equilibrium strategy between the electricity supplier and the charging consumer so that the objective functions of both participants are stable at the end of the game.

[0089] In some embodiments, step 3 is specifically as follows: initialize a group of particles; make the solution of each particle represent a variable combination, each having a position representing charging power and a speed, and then search the solution space by the particle, and the particle updates the position according to its own historical best solution and the best solution of all particles;

[0090] Among them, position represents the value of the current solution, and speed represents the moving speed of the particle in the solution space;

[0091] The goal of the particle is to find the game equilibrium between power suppliers and charging consumers by optimizing the allocation of charging power, thereby achieving the optimal compromise of the payment function;

[0092] The particle swarm optimization algorithm searches for the global optimal solution by updating the speed and position of particles. The speed update formula is:

[0093] V i (t+1)=ω·V i (t)+c 1 ·r 1 ·(P best -P i )+c 2 ·r 2 ·(P best -P i );

[0094] Among them, ω is the inertia weight, which controls the influence of the particle's historical velocity; c 1 and c 2 is the acceleration constant, controlling the particle's movement toward the personal best position and the global best position; r 1 and r 2 It is a random number, which increases the randomness of the search;

[0095] The position update formula is: P i (t+1)=P i (t)+V i (t+1);

[0096] At the same time, the simulated annealing mechanism is introduced, that is, the probability of accepting the solution is: if the new solution P i The fitness value of (t+1) is better than the old solution P i (t) is better, then directly accept this solution;

[0097] If the new solution P i The fitness value of (t+1) is not as good as the old solution Pi (t), then the probability of acceptance using the Metropolis criterion is:

[0098]

[0099] E=λ 1 U supply +λ 2 U consumer

[0100] Among them, E is the current function value, that is, the energy of the solution; E new is the energy of the new solution; T is the temperature parameter, which is used to control the amplitude of the disturbance. As the iteration proceeds, the temperature gradually decreases and the disturbance decreases, thereby refining the search.

[0101] The pseudo code is as follows:

[0102] Input: Weighting coefficient λ 1 and λ 2 , temperature parameter T, number of iterations K, initial state of the particle (initial velocity V 0 , initial position P 0 );

[0103] Output: Charging pile power P best , electricity price C best , voltage drop ΔV best ;

[0104]

[0105] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A multi-objective game algorithm for optimizing the scheduling of electric vehicle charging loads; characterized in that: The steps include: Step 1: Establish a mathematical model of the electricity price and voltage drop ΔV at the public node where the electric vehicle charging station is located; Step 2: Taking the electricity price and the voltage drop ΔV of the public node where the electric vehicle charging station is located as the objective function, a weighted method based on a two-person zero-sum game is adopted to convert the multi-objective optimization scheduling problem into a single-objective optimization scheduling problem, determine the relevant constraints and game forms, and provide an optimization scheduling plan; Step 3: Use simulated annealing particle swarm optimization algorithm based on particle swarm optimization method to balance the goals of car charging stations and consumers and find a balanced solution; Step 4, set the maximum number of iterations, the particle swarm fitness threshold or the minimum value set; output the optimal charging power allocation plan when the stop condition is met, and realize the optimal strategy combination between the voltage drop at the public connection point of the electric vehicle charging station and the charging electricity price at the public connection point. If it is not met, return to step 3 to continue iterating and finally obtain the Pareto optimal solution.

2. The multi-objective game algorithm for optimizing the scheduling of electric vehicle charging load according to claim 1 is characterized in that: Step 1 is specifically as follows: Step 1 is specifically as follows: taking the minimum voltage reduction of the public node where the electric vehicle charging station is located and the minimum charging electricity price of the public connection point as optimization objectives, determining the key variables that affect the grid voltage and the charging electricity price; The key variables include the charging power of each charging pile, the electricity price, the voltage drop ΔV, the payment ability of the power supplier and the payment ability of the consumer; Wherein, the voltage drop ΔV is the voltage drop amount of the common connection point; Then the mathematical model of the voltage drop ΔV at the common connection point is: Among them, α is the proportional constant of voltage drop; P i is the charging power of the i-th charging pile; R is the resistance of the grid line; n is the total number of charging piles; The mathematical model of the electricity price c is: c=f prise (ΔV)=β·ΔV; Wherein, β is the proportionality coefficient between the electricity price and the voltage drop ΔV; The payment function of the electricity supplier is: The consumer's payment function is: The optimization objective function is: Among them, the charging power limit of the charging pile is: P min ≤P i ≤P max The voltage stability constraint is: ΔV≤ΔV max The revenue limit of the power supplier of the power grid is: c ≥ c min 。 3. The multi-objective game algorithm for optimizing the scheduling of electric vehicle charging load according to claim 2 is characterized in that: Step 2 is to balance the two objective functions using a weighted method based on a two-person zero-sum game, and the formula is: Among them, λ i is the weighting coefficient, indicating the importance attached to each goal; The original multi-objective optimization problem has two objectives, namely U supply and U consumer , then the weighting coefficients are λ1 and λ2 respectively, and λ1+λ2=1; When optimizing objective i alone, the optimal solution of the original problem is denoted as The set of optimal solutions is denoted as X * ; There are two players in the game. One player selects a strategy from U1 as his own strategy, and the other player selects a strategy from the set X of optimal solutions. * Choose a solution as your strategy; The goal of the game is to find an equilibrium strategy between the electricity supplier and the charging consumer so that the objective functions of both participants are stable at the end of the game.

4. The multi-objective game algorithm for optimizing the scheduling of electric vehicle charging load according to claim 3 is characterized in that: Step 3 is as follows: initialize a group of particles; make the solution of each particle represent a variable combination, each having a position representing the charging power and a speed, and then search in the solution space through the particles, and the particles update their positions according to their own historical best solutions and the best solutions of all particles; Among them, position represents the value of the current solution, and speed represents the moving speed of the particle in the solution space; The goal of the particle is to find the game equilibrium between power suppliers and charging consumers by optimizing the allocation of charging power, thereby achieving the optimal compromise of the payment function; The particle swarm optimization algorithm searches for the global optimal solution by updating the speed and position of particles. The speed update formula is: V i (t+1)=ω·V i (t)+c1·r1·(P best -P i )+c2·r2·(P best -P i ); Among them, ω is the inertia weight, which controls the particle's influence on the historical velocity; c1 and c2 are acceleration constants, which control the particle's movement to the personal best position and the global best position; r1 and r2 are random numbers, which increase the randomness of the search; The position update formula is: P i (t+1)=P i (t)+V i (t+1); At the same time, the simulated annealing mechanism is introduced, that is, the probability of accepting the solution is: if the new solution P i The fitness value of (t+1) is better than the old solution P i (t) is better, then directly accept this solution; If the new solution P i The fitness value of (t+1) is not as good as the old solution P i (t), then the probability of acceptance using the Metropolis criterion is: E=λ1U supply +λ2U consumer Among them, E is the current function value, that is, the energy of the solution; E new is the energy of the new solution; T is the temperature parameter, which is used to control the amplitude of the disturbance. As the iteration proceeds, the temperature gradually decreases and the disturbance decreases, thereby refining the search.

Images