Intelligent community electric vehicle charging pile optimal dispatching system matched with virtual power plant

Abstract

The invention discloses an intelligent community electric vehicle charging pile optimization scheduling system matched with a virtual power plant. The system belongs to the technical field of energy management and intelligent scheduling, and comprises a wind and light energy power generation system, an energy storage system, a local platform, a centralized control platform, a power market, a community power system, a community electric vehicle charging pile, a household controllable load, a monitoring center and a prediction module. According to the system, information is gathered through a community local platform and a centralized control platform, a coordinated and consistent energy production and supply system is formed by integrating distributed energy resources such as a wind and light energy power generation system, an energy storage system and a household controllable load, and the optimal economic benefit of the system is given by applying a WSA algorithm; the highest charging price of the electric vehicle in a low-carbon scene is taken as a target by a community owner; through the scheduling of the virtual power plant, the dependence of the community on the traditional power grid can be reduced, the resource configuration can be optimized, the consumption of renewable energy sources can be promoted, and the carbon footprint of the intelligent community can be reduced.

Description

Technical Field

[0001] The invention belongs to the technical field of energy management and intelligent dispatching, and relates to an intelligent community electric vehicle charging pile optimization dispatching system cooperating with a virtual power plant. Background Art

[0002] In the existing technology, traditional community electric vehicle charging piles have many disadvantages. When the temperature is high in summer, the use of high-power electrical appliances such as air conditioners and refrigerators will cause a surge in the load on the power grid. In addition, since traditional charging piles cannot adjust the charging power according to the load conditions of the power grid, it is not conducive to the stable operation of the power grid. In the absence of the coordination of virtual power plants, community charging piles may be idle in large numbers during non-peak hours, resulting in a waste of resources. In addition, the popularization of this technology is not perfect, and the voltage may be deliberately reduced. Many communities are limited by factors such as distribution capacity, resulting in few access points for electric vehicle charging equipment and difficult access. Summary of the invention

[0003] In view of the above problems, the purpose of the present invention is to propose an optimization and dispatching system for electric vehicle charging piles in smart communities that cooperates with virtual power plants; by aggregating decentralized charging facilities, the grid resource allocation is optimized, the charging efficiency and equipment utilization rate are improved, and by aggregating charging piles, energy storage and other resources, load regulation during peak power consumption is achieved, reducing the load pressure on the grid, and V2G (electric vehicles supplying electricity to the grid) allows electric vehicles to store excess electricity during low power consumption periods, thereby optimizing resource allocation.

[0004] The technical solution of the present invention is: the intelligent community electric vehicle charging pile optimization dispatching system in cooperation with a virtual power plant described in the present invention is characterized by comprising an interconnected wind and solar power generation system, an energy storage system, a local platform, a centralized control platform, a power market, a community power system, a community electric vehicle charging pile, a household controllable load, a trading center, an optimization platform, a prediction module and a monitoring center;

[0005] Furthermore, the wind-solar power generation system includes photovoltaic power generation and wind power generation;

[0006] The power generation units of the wind and solar power generation system realize unified management and dispatch of energy production and output through the local platform and centralized control platform;

[0007] The function of the energy storage system is to adjust the energy supply and demand relationship and ensure a stable supply of energy; it includes energy storage batteries and supercapacitors;

[0008] The local platform and the centralized control platform are responsible for the operation and management of the entire virtual power plant. The local platform collects and analyzes the data of the internal units of the smart community, and the centralized control platform collects and analyzes the data of the power market and the community power system. The two work together to coordinate and control the operation and management of the entire system; at the same time, the centralized control platform can formulate power dispatch plans according to market demand and floating power market prices;

[0009] The trading center formulates reasonable trading strategies and price strategies based on the optimization results of the optimization platform and is responsible for trading and managing electricity;

[0010] The monitoring center detects the historical carbon emission costs and charging and discharging costs of the community electric vehicle charging piles and transmits the data to the prediction module;

[0011] The optimization platform uses an intelligent scheduling algorithm to optimize the charging costs of car owners and optimize resource allocation under the low-carbon goal.

[0012] Furthermore, during transactions with the electricity market, the trading center classifies the distributed energy within the community and constructs an operation control framework, and makes dynamic adjustments based on market demand and price signals through intelligent algorithms. After the transaction is completed, the virtual power plant conducts economic settlement based on actual energy output and market prices.

[0013] Furthermore, the prediction steps of the prediction module are as follows:

[0014] Step (1): Input historical carbon emission data and charging price and combine them as historical data columns, perform data preprocessing, and perform Hilbert transformation on the data. The transformation formula is as follows:

[0015]

[0016] In the formula, u k is the modal component, x is the data sequence number, k is the number of modal components, δ Dirac function, w k is the frequency;

[0017] Step (2): extract each modal component and a residual component of the data series;

[0018] Step (3): Normalize each component and divide it into training data set and test data set;

[0019] Step (4): Bring the training data set as input x into the KELM prediction model for training, and determine the output of the jth hidden layer node, where the output is the electricity price cost and carbon emissions generated by the predicted load;

[0020] Step (5): Determine the output H(x) of the hidden layer of the KELM prediction model. The calculation formula is as follows:

[0021] H(x)=[h1(x),h2(x),…,h L (x)]

[0022] Step (6): When the signal passes through the hidden layer and enters the output layer, the kernel function is introduced to replace h(x). The output of the KELM prediction model is The expression formula is as follows:

[0023]

[0024] Where, Ω=HH T , h(x)=H is the output matrix of the hidden layer, H + is the generalized inverse matrix of matrix H, K(x i ,x j ) is the Gaussian kernel function, expressed as:

[0025]

[0026] In the formula, ψ is the kernel function, and its value directly affects the generalization ability of KELM. In addition, the penalty coefficient C will affect the prediction accuracy of the KELM model.

[0027] Step (7): Solve the minimum output error and find the minimum training error by solving the equation using the mean absolute percentage. The objective function to be solved is as follows:

[0028]

[0029] Where: ξ represents the training error, Num is the total number of samples: y * 、x * The measured and predicted values ​​for the load.

[0030] Furthermore, the optimization platform is provided with the following control method:

[0031] Step (1): Improve initialization, initialize the benefits corresponding to each optimized particle population; each improved particle population has a set of corresponding values ​​of the objective function, and the initialization process is improved by using Logistic chaotic mapping. Logistic chaotic mapping has the function of making the system in a chaotic state, and the mapping distribution uniformity reaches the extreme. The specific formula is as follows:

[0032] X n+1 =αX n (1-X n )

[0033] Where: X n∈[0,1]; α∈[0,3] is the Logistic parameter;

[0034] Step (2): Calculate the fitness of each population, i.e., the objective function. The objective function is the optimal benefit value of the system. The calculation formula is as follows. The system operation goal is to minimize its own operating costs within the operation decision cycle while satisfying the system conditions. That is, the residents of the community minimize the cost of buying electricity from the power system and the system's own power generation cost:

[0035]

[0036] Where: represents the output power of the fast start-stop unit in period t; f g,i represents the total power generation cost in T operating periods; and For the price of electricity purchase and sale, and It is a decision-making process for purchasing and selling electricity, and the following constraints are met:

[0037]

[0038] In the formula, C k is the carbon emissions that need to be allocated during the charging pile cycle T, C ev is the carbon emissions generated during the charging and discharging process, Δt c and η c are the charging power, charging time and charging loss coefficient of the charging pile in period t, Δt 1 and η t are the discharge efficiency, discharge duration and discharge loss coefficient of the charging pile in the t period, LC is the maximum equivalent full charge and discharge cycles of the charging pile, E r is the rated efficiency of the energy storage device. The objective function needs to satisfy the condition of minimizing the charging cost. k The smaller the better;

[0039] Step (3): Enter the development phase. The formula is as follows:

[0040] W i new =-β·r2·cos(πi / n w2 )·(W best -W fi )+W fi

[0041]

[0042] In the formula, β = 0.75 + ei / nw2, β is the reflection intensity coefficient, r2 is a random value between 0 and 1, and nw2 is the number of particles simulating reflected electromagnetic waves, W fi is the position matrix after W is rearranged in order of fitness value from small to large, W best Indicates the lowest price for charging electric vehicles;

[0043] Step (4): Enter the local survey phase and emit electromagnetic waves:

[0044]

[0045] In the formula, C k is the waveform size control coefficient, that is, the carbon emission, m is a column vector whose elements obey the normal distribution and are arranged in order, W best is the current optimal position, indicating the lowest price for electric vehicle charging, W i For W press and W best The proximity of the rearranged position matrix, f max is the maximum fitness value in the group, W i new It is to reduce the possibility of falling into the local optimum and improve the search efficiency. i It is an improved greedy mechanism, which makes the group position no worse than the current group position when the group position fluctuates outward;

[0046] Step (5): Reflect electromagnetic waves, the expression is:

[0047] W i new =-β·r2·cos(πi / n w2 )·(W best -W fi )+W fi

[0048]

[0049] In the formula, β = 0.75 + ei / nw2, β is the reflection intensity coefficient, r is a random value between 0 and 1, and n w2 is the number of particles simulating reflected electromagnetic waves, W fi is the position of the electric vehicle charging price after W is rearranged in the order of fitness value from small to large;

[0050] Step (6): Receive electromagnetic waves, the expression is:

[0051]

[0052] W best *=[W best 1;W best 2;W best 3…W bestt]*([I1;I2;I3…It])

[0053] I1=I2=I3…It=1 / t

[0054] λ=(2t / T-0.7) / (0.78+|2t / T-0.7|)+1

[0055] Where δ is the receiving coefficient, δ = 0.6 + (1.2-0.5) sin (tπ / 2T), η is a random number that obeys the normal distribution, n w3 To simulate the number of particles receiving electromagnetic waves, is the historical optimal position obtained by convolution, that is, the best individual obtained by the algorithm up to the tth iteration, λ is the correction factor; r3, r4, r5, r6 are random numbers between 0 and 1;

[0056] Step (7): Introduce Levy flight to optimize the algorithm, the expression is:

[0057] W i =lb+Levy·(ub-lb)ifW i >uborW i <lb

[0058]

[0059] Where Levy(D) represents the Levy flight distribution strategy, s is a constant, which is 0.01, u and v are random numbers between 0 and 1; lb and ub are the lower and upper boundaries of the search range respectively;

[0060] Step (8): Determine whether the maximum number of iterations has been reached. If yes, proceed to step (9). If not, return to step (4).

[0061] Step (9): Output the optimal strategy suitable for the system, that is, the optimal charging price for electric vehicles under low carbon emissions.

[0062] The beneficial effects of the present invention are: 1. The KELM sub-prediction module adopted by the system can accurately predict the electricity price cost and carbon emissions generated by electric vehicle charging piles, providing a scientific basis for optimizing scheduling; at the same time, the control method in the optimization platform adopts an improved particle swarm optimization algorithm, which effectively improves the search efficiency and reduces the possibility of falling into local optimality by introducing strategies such as Logistic chaos mapping and Levy flight; 2. The system can be dynamically adjusted according to market demand and price signals to achieve optimal allocation of resources; 3. During the peak period of electricity demand, the virtual power plant can reduce the overall power consumption by adjusting the working mode of the charging pile, realize demand-side management, and reduce the burden on the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is a schematic diagram of the structure of the present invention;

[0064] Figure 2 is a flow chart of the prediction module of the present invention;

[0065] Figure 3 It is the algorithm flow chart of the present invention;

[0066] Figure 4 This is a comparison diagram before and after the optimized carbon emission of the present invention;

[0067] Figure 5 It is a comparison chart of the satisfaction of community car owners with charging prices before and after the optimization of the present invention. DETAILED DESCRIPTION

[0068] The specific technical scheme of the present invention is further described in detail below with reference to specific examples.

[0069] As shown in the figure, the intelligent community electric vehicle charging pile optimization and dispatching system in cooperation with a virtual power plant described in the present invention includes a wind and solar power generation system, an energy storage system, a local platform, a centralized control platform, a power market, a community power system, a community electric vehicle charging pile, a household controllable load, a trading center, an optimization platform, and a prediction module;

[0070] The wind and solar power generation system is a core component of the system, which includes photovoltaic power generation and wind power generation. The power generation units of the wind and solar power generation system realize unified management and scheduling of energy production and output through the local platform and the centralized control platform; the energy storage system is the part of the system used to store and manage energy. Its main function is to adjust the energy supply and demand relationship and ensure a stable supply of energy. The energy storage system is mainly composed of energy storage batteries and supercapacitors; the local platform and the centralized control platform are the central nervous system of the virtual power plant, responsible for the operation and management of the entire virtual power plant. The local platform mainly collects and analyzes the data of the internal units of the smart community, and the centralized control platform is responsible for the power market and the community power system. The data is collected and analyzed, and the two work together to coordinate and control the operation and management of the entire system, realizing the unified management and scheduling of energy production, storage and use; at the same time, the centralized control platform can formulate reasonable power dispatch plans according to market demand and floating electricity market prices; the trading center formulates reasonable trading strategies and price strategies based on market conditions, customer needs and other factors, and the optimization results of the optimization platform, and is responsible for trading and managing electricity; the monitoring center detects the historical carbon emission costs and charging and discharging costs of community electric vehicle charging piles and transmits the data to the prediction module; the optimization platform uses intelligent scheduling algorithms to optimize resource allocation under the low-carbon goal.

[0071] In the intelligent community electric vehicle charging pile optimization and dispatching system that cooperates with a virtual power plant, the trading center classifies the distributed energy within the community and constructs an operation control framework during transactions with the power market, and makes dynamic adjustments based on market demand and price signals through intelligent algorithms.

[0072] like Figure 2-3 As shown, the prediction steps of the KELM sub-prediction module are as follows:

[0073] Step (1): Input historical carbon emission data and charging price and combine them as historical data columns, perform data preprocessing, and perform Hilbert transformation on the data. The transformation formula is as follows:

[0074]

[0075] In the formula, u k is the modal component, x is the data sequence number, k is the number of modal components, δ Dirac function, w k is the frequency;

[0076] Step (2): extract each modal component and a residual component of the data series;

[0077] Step (3): Normalize each component and divide it into training data set and test data set;

[0078] Step (4): Bring the training data set as input x into the KELM prediction model for training, and determine the output of the jth hidden layer node, where the output is the electricity price cost and carbon emissions generated by the predicted load;

[0079] Step (5): Determine the output H(x) of the hidden layer of the KELM prediction model. The calculation formula is as follows:

[0080] H(x)=[h1(x),h2(x),…,h L (x)]

[0081] Step (6): When the signal passes through the hidden layer and enters the output layer, the kernel function is introduced to replace h(x). The output of the KELM prediction model is The expression formula is as follows:

[0082]

[0083] Where, Ω=HH T , h(x)=H is the output matrix of the hidden layer, H + is the generalized inverse matrix of matrix H, K(x i ,x j ) is the Gaussian kernel function, expressed as:

[0084]

[0085] In the formula, ψ is the kernel function, and its value directly affects the generalization ability of KELM. In addition, the penalty coefficient C will affect the prediction accuracy of the KELM model.

[0086] Step (7): Solve the minimum output error and find the minimum training error by solving the equation using the mean absolute percentage. The objective function to be solved is as follows:

[0087]

[0088] Where: ξ represents the training error, Num is the total number of samples: y * 、x * are the measured and predicted values ​​of the load.

[0089] Furthermore, the optimization platform is provided with the following control method:

[0090] Step (1): Improve initialization, initialize the benefits corresponding to each optimized particle population; each improved particle population has a set of corresponding values ​​of the objective function, and the initialization process is improved by using Logistic chaotic mapping. Logistic chaotic mapping has the function of making the system in a chaotic state, and the mapping distribution uniformity reaches the extreme. The specific formula is as follows:

[0091] X n+1 =αX n (1-X n )

[0092] Where: X n ∈[0,1]; α∈[0,3] is the Logistic parameter;

[0093] Step (2): Calculate the fitness of each population, i.e., the objective function. The objective function is the optimal benefit value of the system. The calculation formula is as follows. The system operation goal is to minimize its own operating costs within the operation decision cycle while satisfying the system conditions. That is, the residents of the community minimize the cost of buying electricity from the power system and the system's own power generation cost:

[0094]

[0095] Where: represents the output power of the fast start-stop unit in period t; f g,i represents the total power generation cost in T operating periods; and For the price of electricity purchase and sale, and It is a decision-making process for purchasing and selling electricity, and the following constraints are met:

[0096]

[0097] In the formula, C k is the carbon emissions that need to be allocated during the charging pile cycle T, C ev is the carbon emissions generated during the charging and discharging process, Δt c and η c are the charging power, charging time and charging loss coefficient of the charging pile in period t, Δt 1 and η t are the discharge efficiency, discharge duration and discharge loss coefficient of the charging pile in the t period, LC is the maximum equivalent full charge and discharge cycles of the charging pile, E r is the rated efficiency of the energy storage device. The objective function needs to satisfy the condition of minimizing the charging cost. k The smaller the better;

[0098] Step (3): Enter the development phase. The formula is as follows:

[0099] W i new =-β·r2·cos(πi / n w2 )·(W best -W fi )+W fi

[0100]

[0101] Where β = 0.75 + ei / nw2, β is the reflection intensity coefficient, r2 is a random value between 0 and 1, and n w2 is the number of particles simulating reflected electromagnetic waves, W fi is the position matrix after W is rearranged in order of fitness value from small to large, W best Indicates the lowest price for charging electric vehicles;

[0102] Step (4): Enter the local survey phase and emit electromagnetic waves:

[0103]

[0104] In the formula, C k Here, the formula is the waveform size control coefficient, that is, the carbon emission, which needs to be as small as possible. m is a column vector whose elements obey the normal distribution and are arranged in order. W best is the current optimal position, indicating the lowest price for electric vehicle charging, W i For W press and W best The proximity of the rearranged position matrix, f max is the maximum fitness value in the group; W inew The role of W is to reduce the possibility of falling into local optimality and improve search efficiency; i It is an improved greedy mechanism, which makes the group position no worse than the current group position when the group position fluctuates outward;

[0105] Step (5): Reflect electromagnetic waves, the expression is:

[0106] W i new =-β·r2·cos(πi / n w2 )·(W best -W fi )+W fi

[0107]

[0108] In the formula, β = 0.75 + ei / nw2, β is the reflection intensity coefficient, r is a random value between 0 and 1, and n w2 is the number of particles simulating reflected electromagnetic waves; W fi is the position of the electric vehicle charging price after W is rearranged in the order of fitness value from small to large;

[0109] Step (6): Receive electromagnetic waves, the expression is:

[0110]

[0111] W best *=[W best 1;W best 2;W best 3…W best t]*([I1;I2;I3…It])

[0112] I1=I2=I3…It=1 / t

[0113] λ=(2t / T-0.7) / (0.78+|2t / T-0.7|)+1

[0114] Where δ is the receiving coefficient, δ = 0.6 + (1.2-0.5) sin (tπ / 2T), η is a random number that obeys the normal distribution, n w3 To simulate the number of particles receiving electromagnetic waves; is the historical optimal position obtained by convolution, that is, the best individual obtained by the algorithm up to the tth iteration (indicating the lowest charging price for electric vehicles); λ is the correction factor; r3, r4, r5, r6 are random numbers between 0 and 1;

[0115] This formula simulates that radar usually receives electromagnetic waves normally, but sometimes it is interfered with, that is, it simulates that the charging price may be affected by market fluctuations and charging demand, so it needs to be corrected and processed; at the same time, there is a certain probability that it will deflect in the W direction to reduce the possibility of falling into the local optimum; this formula further updates the position of electric vehicle charging prices;

[0116] Step (7): Introduce Levy flight to optimize the algorithm, the expression is:

[0117] W i =lb+Levy·(ub-lb)ifW i >uborW i <lb

[0118]

[0119] Where Levy(D) represents the Levy flight distribution strategy, s is a constant, which is 0.01, u and v are random numbers between 0 and 1; lb and ub are the lower and upper boundaries of the search range respectively;

[0120] Step (8): Determine whether the maximum number of iterations has been reached. If yes, proceed to step (9); if not, return to step (4);

[0121] Step (9): Output the optimal strategy suitable for the system, that is, the optimal charging price for electric vehicles under low carbon emissions.

[0122] like Figure 4-5 As shown, compared with the smart community electric vehicle pile without the introduction of the present invention, the present invention effectively reduces carbon emissions and meets the needs of most owners in the community.

[0123] Example

[0124] In a certain smart community, with the development and popularization of electric vehicles, a large number of electric vehicles charging during peak electricity consumption periods have brought tremendous pressure to the power grid; the peak charging period overlaps with the peak period of residential electricity consumption, resulting in excessive load on the power grid; to improve this situation, virtual power plant technology is introduced and combined with charging piles to achieve orderly charging and optimize the power grid load; by applying this technology to smart communities, effective charging strategies are formulated according to the results of power grid load forecasts, and then during peak periods, the pressure on the power grid is reduced by reducing the charging power of charging piles, and during low electricity consumption periods, the charging power is increased to promote the consumption of clean energy. This invention effectively reduces the load pressure on the power grid during peak hours, improves the stability and safety of power grid operation, improves the charging experience and satisfaction of car owners, and through reasonable energy management and optimization, it brings significant economic benefits to the community.

Claims

1. An intelligent community electric vehicle charging pile optimization scheduling system in cooperation with a virtual power plant, characterized in that: It includes interconnected wind and solar power generation systems, energy storage systems, local platforms, centralized control platforms, power markets, community power systems, community electric vehicle charging piles, household controllable loads, trading centers, optimization platforms, prediction modules and monitoring centers; The wind and solar power generation system includes photovoltaic power generation and wind power generation; The power generation units of the wind and solar power generation system realize unified management and dispatch of energy production and output through the local platform and centralized control platform; The function of the energy storage system is to adjust the energy supply and demand relationship and ensure a stable supply of energy; it includes energy storage batteries and supercapacitors; The local platform and the centralized control platform are responsible for the operation and management of the entire virtual power plant. The local platform collects and analyzes the data of the internal units of the smart community, and the centralized control platform collects and analyzes the data of the power market and the community power system. The two work together to coordinate and control the operation and management of the entire system; at the same time, the centralized control platform can formulate power dispatch plans according to market demand and floating power market prices; The trading center formulates reasonable trading strategies and price strategies based on the optimization results of the optimization platform and is responsible for trading and managing electricity; The monitoring center detects the historical carbon emission costs and charging and discharging costs of the community electric vehicle charging piles and transmits the data to the prediction module; The optimization platform uses intelligent scheduling algorithms to optimize the charging costs of car owners and optimize resource allocation under the low-carbon goal; During the transaction process with the electricity market, the trading center classifies the distributed energy within the community and builds an operation control framework, and makes dynamic adjustments based on market demand and price signals through intelligent algorithms. After the transaction is completed, the virtual power plant conducts economic settlement based on actual energy output and market prices.

2. According to claim 1, an intelligent community electric vehicle charging pile optimization scheduling system in cooperation with a virtual power plant is characterized in that: The prediction steps of the prediction module are as follows: Step (1): Input historical carbon emission data and charging price and combine them as historical data columns, perform data preprocessing, and perform Hilbert transformation on the data. The transformation formula is as follows: In the formula, u k is the modal component, x is the data sequence number, k is the number of modal components, δ Dirac function, w k is the frequency; Step (2): extract each modal component and a residual component of the data series; Step (3): Normalize each component and divide it into training data set and test data set; Step (4): Bring the training data set as input x into the KELM prediction model for training, and determine the output of the jth hidden layer node, where the output is the electricity price cost and carbon emissions generated by the predicted load; Step (5): Determine the output H(x) of the hidden layer of the KELM prediction model. The calculation formula is as follows: H(x)=[h1(x),h2(x),…,h L (x)] Step (6): When the signal passes through the hidden layer and enters the output layer, the kernel function is introduced to replace h(x). The output of the KELM prediction model is The expression formula is as follows: Where, Ω=HH T , h(x)=H is the output matrix of the hidden layer, H + is the generalized inverse matrix of matrix H, K(x i ,x j ) is the Gaussian kernel function, expressed as: In the formula, ψ is the kernel function, and its value directly affects the generalization ability of KELM. In addition, the penalty coefficient C will affect the prediction accuracy of the KELM model. Step (7): Solve the minimum output error and find the minimum training error by solving the equation using the mean absolute percentage. The objective function to be solved is as follows: Where: ξ represents the training error, Num is the total number of samples: y * 、x * The measured and predicted values ​​for the load.

3. According to the claim, an optimization and dispatching system for electric vehicle charging piles in a smart community in cooperation with a virtual power plant, characterized in that: The optimization platform is provided with the following control method: Step (1): Improve initialization and initialize the benefits corresponding to each optimized particle population; Step (2): Calculate the fitness of each population, i.e., the objective function; Step (3): Enter the development phase; Step (4): Enter the local survey stage; Step (5): reflecting electromagnetic waves; Step (6): receiving electromagnetic waves; Step (7): Introduce Levy flight to optimize the algorithm; Step (8): Determine whether the maximum number of iterations has been reached. If yes, proceed to step (9). If not, return to step (4). Step (9): Output the optimal strategy suitable for the system, that is, the optimal charging price for electric vehicles under low carbon emissions.

4. According to claim 3, the intelligent community electric vehicle charging pile optimization scheduling system in cooperation with the virtual power plant is characterized in that: In step (1), each improved particle population has a set of corresponding values ​​of the objective function. The Logistic chaotic mapping is used to improve the initialization process. The Logistic chaotic mapping has the function of making the system in a chaotic state and achieving the ultimate uniformity of the mapping distribution. The specific formula is as follows: X n+1 =αX n (1-X n ) Where: X n ∈[0,1]; α∈[0,3] is the Logistic parameter.

5. According to claim 3, the intelligent community electric vehicle charging pile optimization scheduling system in cooperation with the virtual power plant is characterized in that: In step (2), the objective function is the optimal benefit value of the system, which is calculated as follows: the system operation goal is to minimize its own operating costs within the operation decision cycle while satisfying the system conditions, that is, the community residents minimize the cost of buying electricity from the power system and the system's own power generation cost: Where: represents the output power of the fast start-stop unit in period t; f g,i represents the total power generation cost in T operating periods; and For the price of electricity purchase and sale, and It is a decision-making process for purchasing and selling electricity, and the following constraints are met: In the formula, C k is the carbon emissions that need to be allocated during the charging pile cycle T, C ev is the carbon emissions generated during the charging and discharging process, Δt c and η c are the charging power, charging time and charging loss coefficient of the charging pile in period t, Δt 1 and η t are the discharge efficiency, discharge duration and discharge loss coefficient of the charging pile in the t period, LC is the maximum equivalent full charge and discharge cycles of the charging pile, E r is the rated efficiency of the energy storage device. The objective function needs to satisfy the condition of minimizing the charging cost. k The smaller the better.

6. According to claim 3, the intelligent community electric vehicle charging pile optimization scheduling system in cooperation with the virtual power plant is characterized in that: In step (3), the formula is as follows: IN i new =-β r2 cos(πi / n w2 )·(IN best -IN fi )+W fi In the formula, β = 0.75 + ei / nw2, β is the reflection intensity coefficient, r2 is a random value between 0 and 1, and n w2 is the number of particles simulating reflected electromagnetic waves, W fi is the position matrix after W is rearranged in order of fitness value from small to large, W best Indicates the lowest price for charging an electric vehicle.

7. According to claim 3, the intelligent community electric vehicle charging pile optimization scheduling system in cooperation with the virtual power plant is characterized in that: In step (4), it emits electromagnetic waves: In the formula, C k is the waveform size control coefficient, that is, the carbon emission, m is a column vector whose elements obey the normal distribution and are arranged in order, W best is the current optimal position, indicating the lowest price for electric vehicle charging, W i For W press and W best The proximity of the rearranged position matrix, f max is the maximum fitness value in the group, W i new It is to reduce the possibility of falling into the local optimum and improve the search efficiency. i It is an improved greedy mechanism, which makes the group position no worse than the current group position when the group position fluctuates outward.

8. The intelligent community electric vehicle charging pile optimization scheduling system in cooperation with a virtual power plant according to claim 3 is characterized in that: In step (5), the expression is: IN i new =-β r2 cos(πi / n w2 )·(IN best -IN fi )+W fi In the formula, β = 0.75 + ei / nw2, β is the reflection intensity coefficient, r is a random value between 0 and 1, and n w2 is the number of particles simulating reflected electromagnetic waves, W fi It is the position of the electric vehicle charging price after W is rearranged in the order of fitness value from small to large.

9. The intelligent community electric vehicle charging pile optimization scheduling system in cooperation with a virtual power plant according to claim 3 is characterized in that: In step (5), the expression is: W best *=[W best 1;W best 2;W best 3…Wbestt]*([I1;I2;I3…It]) I1=I2=I3…It=1 / t λ=(2t / T-0.7) / (0.78+|2t / T-0.7|)+1 Where δ is the receiving coefficient, δ = 0.6 + (1.2-0.5) sin (tπ / 2T), η is a random number that obeys the normal distribution, n w3 To simulate the number of particles receiving electromagnetic waves, is the historical optimal position obtained by convolution, that is, it represents the best individual obtained by the algorithm up to the tth iteration, λ is the correction factor; r3, r4, r5, r6 are random numbers ranging from 0 to 1.

10. The intelligent community electric vehicle charging pile optimization scheduling system in cooperation with a virtual power plant according to claim 3 is characterized in that: In step (5), the expression is: W i =lb+Levy·(ub-lb)ifW i >uborW i <lb Where Levy(D) represents the Levy flight distribution strategy, s is a constant, which is 0.01, u and v are random numbers between 0 and 1; lb and ub are the lower and upper boundaries of the search range, respectively.

Images