A method for optimizing the dispatch of regional distribution networks on both sides

Abstract

The present invention discloses a method for optimizing the two-side dispatching of a regional distribution network, including the following steps: 1) constructing an economic optimization model for the load side of a regional distribution network; 2) constructing an economic optimization model for the source side of a regional distribution network; 3) using a firefly algorithm to solve the load curve when the economic cost of the load side of the regional distribution network is the lowest; 4) using a non-dominated sorting genetic algorithm to solve the economic and low-carbon goals of the source side of the regional distribution network. The present invention uses an algorithm that imitates nature or designs artificial heuristic rules to solve problems, uses a firefly algorithm to find the optimal solution and simulates solutions in the search space; uses a non-dominated sorting genetic algorithm to simulate operations such as selection, crossover, and mutation in the evolutionary process to generate the next generation of solutions from the population, and uses non-dominated sorting and congestion distance calculation to maintain a set of non-dominated excellent solutions.

Description

Technical Field

[0001] The present invention relates to an energy distribution optimization method, in particular to an energy optimization scheduling method for the load side and energy side of a regional distribution network. Background Art

[0002] With the gradual implementation of the new power system ideas and measures based on new energy and the implementation of the pilot program for rooftop distributed photovoltaic development, the roofs of buildings such as schools, hospitals, industrial and commercial enterprises, industrial parks, factories, and residential communities are rationally planned and installed according to the principle of building as much as possible and connecting as much as possible. New energy, especially distributed photovoltaic development, will usher in a larger-scale explosive growth, and distributed photovoltaics will present a development pattern of "high-speed development of installed capacity and increase in the proportion of low-voltage grid connection". However, new energy power generation faces the problems of instability and intermittency, which leads to fluctuating energy supply and may affect the power network. Summary of the invention

[0003] The present invention proposes a regional distribution network double-side optimization scheduling method to solve the above problems.

[0004] The present invention is implemented by the following technical means: a method for optimizing the dispatching of two sides of a regional distribution network, comprising the following steps:

[0005] 1) Construct an economic optimization model for the load side of the regional distribution network;

[0006] 2) Construct an economic optimization model for the source side of the regional distribution network;

[0007] 3) Using the firefly algorithm to solve the load curve when the economic cost of the load-side economic optimization model of the regional distribution network is the minimum;

[0008] 4) The load obtained by the previous steps is input into the economic optimization model of the source side of the regional distribution network, and the non-dominated sorting genetic algorithm is used to solve the power generation of each part of the energy on the source side of the regional distribution network to achieve the optimal solution for economy and low carbon.

[0009] The load-side economic optimization model of the regional distribution network includes the objective function and constraints:

[0010] The objective function is specifically: minf1 = P EV +P ESS , P EV The total cost of charging an electric vehicle, P ESS The total cost of charging the energy storage device;

[0011] The constraints include constraints on power balance, constraints on electric vehicle charging power, and constraints on energy storage device charging power;

[0012] The power balance constraint condition is as follows: L(t) = L base (t)+L EV (t)+L ESS (t), L(t) is the total electrical load during period t; L base (t) is the basic electrical load; L EV (t) is the energy storage load of electric vehicles; L ESS (t) is the energy storage load;

[0013] The constraints of the electric vehicle charging power are as follows: EV，min <P EV (t)<P EV，max , P EV，min The lower limit of the charging power of electric vehicles; P EV，max Capping the power of charging electric vehicles;

[0014] The constraints of the energy storage device charging power are as follows: ESS，min <P ESS (t)<P ESS，max , P EV，min is the lower limit of energy storage charging power; P EV，max The upper limit of energy storage charging power.

[0015] The regional distribution network source-side economic optimization model includes an objective function and constraints.

[0016] The objective function includes constraints on economic costs and carbon emissions. The constraints on economic costs are as follows: minf2 = C PV +C DG +C EV +C ESS +C N +C TL , C PV is the power generation cost of the photovoltaic power generation system, C DG is the total cost of the diesel generator during operation, C EV is the electric vehicle battery discharge model, C ESS is the energy storage system discharge model, C N is the cost of purchasing electricity from the main grid, C TL The main network loss cost is the economic cost, and the minimum value of the above sum is taken;

[0017] The carbon emission constraints include the carbon emission constraints of diesel generators and the carbon emission constraints of main grid power generation.

[0018] The carbon emissions of diesel generator sets are expressed as:

[0019] In the formula, α i, β i , γ i are the carbon emission coefficients, α i =24.144kg / (kW·h), β i =1.728kg / (kW·h), γ i =0.0017kg / (kW·h); P i (t) is the output power of the i-th diesel generator set at time t; N is the total number of diesel generator sets; T is the scheduling period;

[0020] The carbon emissions of the main network are expressed as:

[0021] F N =λ N P N (t), λ N is the emission coefficient of main grid power generation; P N (t) Purchase electricity for the main grid;

[0022] The constraints include constraints on power balance, constraints on photovoltaic output, constraints on electric vehicle output and SOC status, constraints on energy storage device output and SOC status, and constraints on diesel generator power;

[0023] The constraints on power balance are as follows: P i (t) is the output power of the i-th diesel generator set at time t; P EV (t) is the charge and discharge power of the electric vehicle during period t, where discharge is positive and charge is negative; P ESS (t) is the charge and discharge power of the energy storage during period t, where discharge is positive and charge is negative; P(t) is the load value during period t; P loss (t) is the power loss during period t; P PV (t) is the photovoltaic output power during period t;

[0024] The constraints on photovoltaic output are as follows: PV (t)≤P PV，f (t), P PV，f (t) is the predicted value of photovoltaic output at time t;

[0025] The output constraint conditions of the electric vehicle are as follows: EV，min <P EV (t)<P EV，max , P EV，min is the lower limit of the charging and discharging power of electric vehicles; P EV，max is the upper limit of the charging and discharging power of electric vehicles, P EV，max =λ(t)N EV P EV，c(t)η EV，c , λ(t) is the stopping probability of the electric vehicle at time t; N EV is the number of dispatchable electric vehicles; P EV，c (t) is the charging power of the electric vehicle in the optimized period t; P EV，r (t) is the discharge power of the electric vehicle in the optimized period t; η EV，c Charging efficiency for electric vehicles; η EV，r Discharge efficiency of electric vehicles;

[0026] The electric vehicle battery SOC state constraints are as follows:

[0027] α EV , β EV are the lower and upper limit coefficients of the battery capacity of electric vehicles respectively; q EVmax The maximum capacity of electric vehicle batteries;

[0028] The output constraint of the energy storage system is: ESS，min <P ESS (t) <P ESS，max , P EV，min is the lower limit of energy storage charging and discharging power; P EV，max is the upper limit of energy storage charging and discharging power, P ESS，max =P ESS，c (t)η ESS，c , P ESS，c (t) is the charging power of the energy storage in the optimized period t; P ESS，r (t) is the discharge power of the energy storage in the optimized period t; η ESS，c is the energy storage charging efficiency; η ESS，r is the energy storage discharge efficiency;

[0029] The energy storage system SOC state constraints are as follows:

[0030] α ESS , β ESS are the lower and upper limit coefficients of energy storage capacity respectively; q ESSmax is the maximum capacity of energy storage;

[0031] The constraints of the diesel generator power are as follows: P PG is the actual output power of the diesel generator; They are the upper and lower limits of diesel generator output power respectively.

[0032] The method of using the firefly algorithm to solve the load curve with the minimum economic cost of the load-side economic optimization model of the regional distribution network specifically includes the following calculation steps:

[0033] 3-1. Based on the total charging cost of electric vehicles, the total charging cost of energy storage and the time-of-use electricity price model, a regional distribution network optimization dispatching model with the minimum economic cost on the load side as the objective function is established;

[0034] 3-2. Input the data collected by the cloud master station into the model;

[0035] 3-3. Constrain the data according to the constraints;

[0036] 3-4. Use the electricity price for 24 hours as the firefly individual, and randomly initialize the initial firefly population with a population size of N;

[0037] 3-5. Calculate the objective function of each firefly individual, that is, the absolute brightness of the firefly individual, and calculate the Euclidean distance r between firefly i and firefly j ij , the calculation formula is as follows:

[0038] In the formula, x i 、x j are the spatial positions of fireflies i and j respectively; i = 1, 2, 3, ..., n; j = 1, 2, 3, ..., n; n is the number of fireflies; d is the spatial dimension;

[0039] 3-6. Update the position of each firefly according to the brightness and Euclidean distance of the fireflies. When firefly i is attracted by another firefly j with higher brightness, the movement of firefly i to firefly j can be expressed as follows:

[0040] β0 represents the maximum attraction value, which is usually set to 1; γ is the light absorption coefficient; α is the step factor, which is between 0 and 1; rand is a random number that follows a uniform distribution on [0, 1];

[0041] 3-7. Repeat steps 3-5 to 3-6 until the number of iterations reaches the maximum value, thereby obtaining the optimized total load curve when the economic cost of the load side including the electric vehicle charging load and the energy storage charging load is the lowest.

[0042] The total charging cost of the electric vehicle P EV for:

[0043] T is the scheduling period; C c (t) is the charging electricity price in the optimized period t; C r (t) is the discharge electricity price in the optimized period t; P EV(t) is the charging power of the electric vehicle in the optimization period t; η EVc Charging efficiency for electric vehicles; η EVr is the discharge efficiency of electric vehicles; T o (t) is the optimization period;

[0044] The total energy storage charging cost F ESS for:

[0045] T is the scheduling period; Cc(t) is the charging price in the optimal period t; C r (t) is the discharge electricity price in the optimized period t; P ESS (t) is the energy storage charging power in the optimized period t; η ESSc is the energy storage charging efficiency; η ESSr is the energy storage discharge efficiency; T o (t) is the optimization period.

[0046] The time-of-use electricity price model in step 3-1 is:

[0047] C v is the off-peak electricity price; C p is the peak electricity price; C v <C p ; (t1, t2) is the off-peak electricity price period. Charging behavior during this period is calculated and charged according to the off-peak electricity price.

[0048] The non-dominated sorting genetic algorithm is used to solve the power generation of each part of the energy on the source side of the regional distribution network, which specifically includes the following calculation steps:

[0049] 4-1. According to the power generation cost of photovoltaic power generation system, the discharge cost of electric vehicle battery, the discharge cost of energy storage, the total operating cost of diesel generator, and the total cost of main grid power purchase, a regional network optimization scheduling model with the lowest economic cost and minimum carbon emission on the source side as the objective function is constructed;

[0050] 4-2. Input the data collected by the cloud master station into the model, substitute the load solved by the firefly algorithm into the model, and use the constraint conditions to establish the relevant supply and demand balance constraints for the model;

[0051] 4-3. Use photovoltaic power generation, diesel generator output power, and main grid power as variables to initialize the population, randomly generate an initial parent population P0 with a population size of N, and generate a child population Q0 through genetic operators, whose population size is also N;

[0052] 4-4. The parent population P n and the offspring population Q n Merge to form a synthetic population R of size 2N n, perform fast non-dominated sorting, and set R n All 2N individuals in are reclassified according to the non-dominated sequence number (level) to obtain levels F1, F2, F3, ...;

[0053] 4-5. Calculate the individual local crowding distance of each non-dominated layer and sort them. The crowding degree calculation formula is as follows:

[0054]

[0055] In the formula, I[i+1].m and I[i-1].m are the function values ​​of the next and previous solutions of solution i on function m respectively; and are the maximum and minimum values ​​of the mth function respectively.

[0056] 4-6. Select N individuals as the new parent population P according to the sorting results n+1 ;

[0057] 4-7. Generate a new offspring population Q through genetic operators n+1 ;

[0058] 4-8. Repeat steps 4-4 to 4-7 until the maximum number of iterations set by the algorithm is reached, and finally the power generation power of each part of the energy on the source side after optimization is obtained.

[0059] The power generation cost of the photovoltaic power generation system is:

[0060] T is the scheduling period; maintenance costs to operate the PV generator; Cost of managing power for photovoltaic generators; is the unit operation and maintenance coefficient of the photovoltaic generator; P is the photovoltaic generator power generation management cost coefficient; PV (t) PV output power in period t;

[0061] The electric vehicle battery discharge cost is:

[0062] C c (t) is the charging electricity price at time t; C r (t) is the discharge electricity price at time t; P EV，c (t) is the charging power of the electric vehicle; P EV，r (t) is the discharge power of the electric vehicle; C EV，d C is the battery loss cost of EV charging and discharging; EV，m The maintenance costs required to charge and discharge electric vehicles;

[0063] The energy storage discharge cost is:

[0064]

[0065] is the charging electricity price at time t; C r (t) is the discharge electricity price at time t; P ESS，c (t) is the charging power of the energy storage at time t; P ESS，r (t) is the discharge power of the energy storage at time t; C ESS，d C is the battery loss cost of energy storage charging and discharging; ESS，m The maintenance costs required to charge and discharge the energy storage;

[0066] The total cost of the diesel generator during operation is:

[0067] C DG1 is the diesel engine operating cost; C DG2 is the maintenance cost of the diesel engine; P DG is the actual power of the diesel generator; a, b, c are the operating cost coefficients of the diesel generator respectively; is the maintenance factor of the diesel engine;

[0068] The total cost of the main grid electricity purchase is the electricity purchase cost and the network loss cost, where the electricity purchase cost is:

[0069] C cn (t) is the unit price of electricity purchased from the main grid during period t; P N (t) is the exchange power with the main network in time period t:

[0070] The network loss cost is:

[0071] C lpss (t) is the network loss price coefficient in time period t; P N (t) is the power loss in electric energy transmission during time period t.

[0072] The present invention solves problems by imitating nature or designing artificial heuristic rules, using the firefly algorithm to find the optimal solution and simulate the solution in the search space; using the non-dominated sorting genetic algorithm to simulate the selection, crossover, mutation and other operations in the evolutionary process to generate the next generation of solutions from the population, and using non-dominated sorting and crowding distance calculation to maintain a set of non-dominated excellent solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 It is a flow chart of the method for implementing the present invention;

[0074] Figure 2 This is a diagram of the dual-layer optimization scheduling model for photovoltaic storage and charging in the region;

[0075] Figure 3 This is the flow chart of the firefly algorithm;

[0076] Figure 4 This is the flow chart of the non-dominated sorting genetic algorithm;

[0077] Figure 5 It is a comparison diagram of non-dominated sorting and congestion degree;

[0078] Figure 6 Solve the structure diagram for the two-tier optimization scheduling algorithm. DETAILED DESCRIPTION

[0079] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0080] The present invention relates to a method for optimizing the two-side dispatching of a regional distribution network. Specifically, it is a method for constructing an intelligent regional distribution network based on time-of-use electricity prices under a V2G (Vehicle-to-Grid) mode, forming energy interactions between electric vehicles and energy storage on the load side and new energy and main grid electricity purchase on the source side. The method firstly models the load and source sides of the region respectively, and a cloud terminal inputs data into an upper-layer load-side economic optimization model, and uses a firefly algorithm (FA) to solve a net total load curve when the economic cost is the smallest; then the net total load curve obtained by the upper-layer optimization and the cloud terminal data are input into a lower-layer source-side optimization dispatching model, and finally a non-dominated sorting genetic algorithm (NSGA-Ⅱ) is used to solve the economic and low-carbon goals of the source side.

[0081] The method specifically includes the following steps:

[0082] 1) Construct an economic optimization model for the load side of the regional distribution network;

[0083] 2) Construct an economic optimization model for the source side of the regional distribution network;

[0084] 3) Use the firefly algorithm to solve the load curve when the economic cost of the load side of the regional distribution network is the minimum;

[0085] 4) Use the non-dominated sorting genetic algorithm to solve the economic and low-carbon objectives of the source side of the regional distribution network.

[0086] Specifically, the load-side economic optimization model of the regional distribution network described in step 1 includes the objective function and constraints:

[0087] The objective function is specifically: minf1 = F EV +F ESS , F EV The total cost of charging an electric vehicle, F ESS The total cost of charging the energy storage device;

[0088] The constraints include constraints on power balance, constraints on electric vehicle charging power, and constraints on energy storage device charging power;

[0089] The power balance constraint condition is as follows: L(t) = L base (t)+L EV (t)+L ESS (t), L(t) is the total electrical load during period t; L base (t) is the basic electrical load; L EV (t) is the energy storage load of electric vehicles; L ESS (t) is the energy storage load;

[0090] The constraints of the electric vehicle charging power are as follows: EV，min <P EV (t)<P EV，max , P EV，min The lower limit of the charging power of electric vehicles; P EV，max Capping the power of charging electric vehicles;

[0091] The constraints of the energy storage device charging power are as follows: ESS，min <P ESS (t)<P ESS，max , P EV，min is the lower limit of energy storage charging power; P EV，max The upper limit of energy storage charging power.

[0092] In step 2, the regional distribution network source-side economic optimization model includes an objective function and constraints.

[0093] The objective function includes constraints on economic costs and carbon emissions. The constraints on economic costs are as follows: minf2 = C PV +C DG +C EV +C ESS +C N +C TL , C PV is the power generation cost of the photovoltaic power generation system, C DG is the total cost of the diesel generator during operation, C EV is the electric vehicle battery discharge model, C ESS is the energy storage system discharge model, C N is the cost of purchasing electricity from the main grid, C TL The main network loss cost, the economic cost is the minimum of the above sum;

[0094] The carbon emission constraints include the carbon emission constraints of diesel generators and the carbon emission constraints of main grid power generation.

[0095] The carbon emissions of diesel generator sets are expressed as:

[0096] In the formula, α i , β i , γ i (α i =24.144kg / (kW·h), β i =1.728kg / (kW·h), γ i =0.0017kg / (kW·h))

[0097] are the carbon emission coefficients; P i (t) is the output power of the i-th diesel generator set at time t; N is the total number of diesel generator sets; T is the scheduling period;

[0098] The carbon emissions of the main network are expressed as:

[0099] F N =λ N P N (t), λ N is the emission coefficient of main grid power generation; P N (t) Purchase electricity for the main grid;

[0100] The constraints include constraints on power balance, constraints on photovoltaic output, constraints on electric vehicle output and SOC status, constraints on energy storage device output and SOC status, and constraints on diesel generator power;

[0101] The constraints on power balance are as follows: P i (t) is the output power of the i-th diesel generator set at time t; P EV (t) is the charge and discharge power of the electric vehicle during period t, where discharge is positive and charge is negative; P ESS (t) is the charge and discharge power of the energy storage during period t, where discharge is positive and charge is negative; P(t) is the load value during period t; P loss (t) is the power loss during period t; P PV (t) is the photovoltaic output power during period t;

[0102] The constraints on photovoltaic output are as follows: PV (t)≤P PV，f (t), P PV，f (t) is the predicted value of photovoltaic output at time t;

[0103] The output constraint conditions of the electric vehicle are as follows: EV，min <P EV (t)<PEV，max , P EV，min is the lower limit of the charging and discharging power of electric vehicles; P EV，max is the upper limit of the charging and discharging power of electric vehicles, P EV，max =λ(t)N EV P EV，c (t)η EV，c , λ(t) is the stopping probability of the electric vehicle at time t; N EV is the number of dispatchable electric vehicles; P EV，c (t) is the charging power of the electric vehicle in the optimized period t; P EV，r (t) is the discharge power of the electric vehicle in the optimized period t; η EV，c Charging efficiency for electric vehicles; η EV，r Discharge efficiency of electric vehicles;

[0104] The electric vehicle battery SOC state constraints are as follows:

[0105] α EV , β EV are the lower and upper limit coefficients of the battery capacity of electric vehicles respectively; q EVmax The maximum capacity of electric vehicle batteries;

[0106] The output constraint of the energy storage system is: ESS，min <P ESS (t)<P ESS，max , P EV，min is the lower limit of energy storage charging and discharging power; P EV，max is the upper limit of energy storage charging and discharging power, P ESS，max =P ESS，c (t)η ESS，c , P ESS，c (t) is the charging power of the energy storage in the optimized period t; P ESS，r (t) is the discharge power of the energy storage in the optimized period t; η ESS，c is the energy storage charging efficiency; η ESS，r is the energy storage discharge efficiency;

[0107] The energy storage system SOC state constraints are as follows:

[0108] α ESS , β ESS are the lower and upper limit coefficients of energy storage capacity respectively; q ESSmax is the maximum capacity of energy storage;

[0109] The constraints of the diesel generator power are as follows: PDG is the actual output power of the diesel generator; They are the upper and lower limits of diesel generator output power respectively.

[0110] In step 3, the firefly algorithm is used to solve the load curve when the economic cost of the load side of the regional distribution network is the minimum, which specifically includes the following steps:

[0111] 3-1. Based on the total charging cost of electric vehicles, the total charging cost of energy storage and the time-of-use electricity price model, a regional distribution network optimization dispatching model with the minimum economic cost on the load side as the objective function is established;

[0112] 3-2. Input the data collected by the cloud master station into the model;

[0113] 3-3. Constrain the data according to the constraints;

[0114] 3-4. Use the electricity price for 24 hours as the firefly individual, and randomly initialize the initial firefly population with a population size of N;

[0115] 3-5. Calculate the objective function of each firefly individual, that is, the absolute brightness of the firefly individual, and calculate the Euclidean distance r between firefly i and firefly j ij , the calculation formula is as follows:

[0116] In the formula, x i 、x j are the spatial positions of fireflies i and j respectively; i = 1, 2, 3, ..., n; j = 1, 2, 3, ..., n; n is the number of fireflies; d is the spatial dimension;

[0117] 3-6. Update the position of each firefly according to the brightness and Euclidean distance of the fireflies. When firefly i is attracted by another firefly j with higher brightness, the movement of firefly i to firefly j can be expressed as follows:

[0118] β0 represents the maximum attraction value, which is usually set to 1; γ is the light absorption coefficient; α is the step factor, which is between 0 and 1; rand is a random number that follows a uniform distribution on [0, 1];

[0119] 3-7. Repeat steps 3-3 to 3-6 until the number of iterations reaches the maximum value, thereby obtaining the optimized total load curve when the economic cost of the load side including the electric vehicle charging load and the energy storage charging load is the lowest.

[0120] The total charging cost of the electric vehicle P EV for:

[0121] T is the scheduling period; C c(t) is the charging electricity price in the optimized period t; C r (t) is the discharge electricity price in the optimized period t; P EV (t) is the charging power of the electric vehicle in the optimization period t; η EVc Charging efficiency for electric vehicles; η EVr is the discharge efficiency of electric vehicles; T o (t) is the optimization period;

[0122] The total energy storage charging cost F ESS for:

[0123] T is the scheduling period; C c (t) is the charging electricity price in the optimized period t; C r (t) is the discharge electricity price in the optimized period t; P ESS (t) is the energy storage charging power in the optimized period t; η ESSc is the energy storage charging efficiency; η ESSr is the energy storage discharge efficiency; T o (t) is the optimization period.

[0124] The time-of-use electricity price model in step 3-1 is:

[0125] C v is the off-peak electricity price; C p is the peak electricity price; C v <C p ; (t1, t2) is the off-peak electricity price period. Charging behavior during this period is calculated and charged according to the off-peak electricity price.

[0126] In step 4, the non-dominated sorting genetic algorithm is used to solve the economic and low-carbon objectives of the source side of the regional distribution network, which specifically includes the following steps:

[0127] 4-1. According to the power generation cost of photovoltaic power generation system, the discharge cost of electric vehicle battery, the discharge cost of energy storage, the total operating cost of diesel generator, and the total cost of main grid power purchase, a regional network optimization scheduling model with the lowest economic cost and minimum carbon emission on the source side as the objective function is constructed:

[0128] 4-2. Input the data collected by the cloud master station into the model, substitute the load solved by the firefly algorithm into the model, and use the constraint conditions to establish the relevant supply and demand balance constraints for the model;

[0129] 4-3. Use photovoltaic power generation, diesel generator output power, and main grid power as variables to initialize the population, randomly generate an initial parent population P0 with a population size of N, and generate a child population Q0 through genetic operators, whose population size is also N;

[0130] 4-4. The parent population P n and the offspring population Q n Merge to form a synthetic population R of size 2N n , perform fast non-dominated sorting, and set R n All 2N individuals in are reclassified according to the non-dominated sequence number (level) to obtain levels F1, F2, F3, ...;

[0131] 4-5. Calculate the individual local crowding distance of each non-dominated layer and sort them. The crowding degree calculation formula is as follows:

[0132]

[0133] In the formula, I[i+1].m and I[i-1].m are the function values ​​of the next and previous solutions of solution i on function m respectively; and are the maximum and minimum values ​​of the mth function respectively.

[0134] 4-6. Select N individuals as the new parent population P according to the sorting results n+1 ;

[0135] 4-7. Generate a new offspring population Q through genetic operators n+1 ;

[0136] 4-8. Repeat steps 4-4 to 4-7 until the maximum number of iterations set by the algorithm is reached, and finally the power generation power of each part of the energy on the source side after optimization is obtained.

[0137] The power generation cost of the photovoltaic power generation system is:

[0138] T is the scheduling period; maintenance costs to operate the PV generator; Cost of managing power for photovoltaic generators; is the unit operation and maintenance coefficient of the photovoltaic generator; P is the photovoltaic generator power generation management cost coefficient; PV (t) PV output power in period t;

[0139] The electric vehicle battery discharge cost is:

[0140]

[0141] The charging price at that moment; C r (t) is the discharge electricity price at time t; P EV，c (t) is the charging power of the electric vehicle; P EV，r (t) is the discharge power of the electric vehicle; C EV，dC is the battery loss cost of EV charging and discharging; EV，m The maintenance costs required to charge and discharge electric vehicles;

[0142] The energy storage discharge cost is:

[0143]

[0144] is the charging electricity price at time t; C r (t) is the discharge electricity price at time t; P ESS，c (t) is the charging power of the energy storage at time t; P ESS，r (t) is the discharge power of the energy storage at time t; C ESS，d C is the battery loss cost of energy storage charging and discharging; ESS，m The maintenance costs required to charge and discharge the energy storage;

[0145] The total cost of the diesel generator during operation is:

[0146] C DG1 is the diesel engine operating cost; C DG2 is the maintenance cost of the diesel engine; P tG is the actual power of the diesel generator; a, b, c are the operating cost coefficients of the diesel generator respectively; is the maintenance factor of the diesel engine;

[0147] The total cost of the main grid electricity purchase is the electricity purchase cost and the network loss cost, where the electricity purchase cost is:

[0148] c cn (t) is the unit price of electricity purchased from the main grid during period t; P N (t) is the exchange power with the main network in time period t;

[0149] The network loss cost is:

[0150] C loss (t) is the network loss price coefficient in time period t; P N (t) is the power loss in electric energy transmission during time period t.

Claims

1. A method for optimizing the dispatching of two sides of a regional distribution network, characterized in that: The steps include: 1) Construct an economic optimization model for the load side of the regional distribution network; 2) Construct an economic optimization model for the source side of the regional distribution network; 3) Using the firefly algorithm to solve the load curve when the economic cost of the load-side economic optimization model of the regional distribution network is the minimum; 4) Input the load curve obtained by the above steps into the economic optimization model of the source side of the regional distribution network, and use the non-dominated sorting genetic algorithm to solve the power generation of each part of the energy source side of the regional distribution network to achieve the optimal solution of economy and low carbon; The load-side economic optimization model of the regional distribution network described in step 1 includes the objective function and constraints: The objective function minf1 is specifically: minf1=P EV +P ESS , P EV is the total charging cost of electric vehicles, P ESS Total charging cost for the energy storage device; The constraints include constraints on power balance, constraints on the charging and discharging power of electric vehicles, and constraints on the charging and discharging power of energy storage devices; The power balance constraint condition is as follows: L(t) = L base (t)+L EV (t)+L ESS (t), L(t) is the total electrical load at time t; L base (t) is the basic electrical load at time t; L EV (t) is the energy storage load of the electric vehicle at time t; L ESS (t) is the energy storage load at time t; The constraints of the electric vehicle charging and discharging power are as follows: EV，min <P EV (t) <P EV，max , P EV，min is the lower limit of the charging and discharging power of electric vehicles; P EV，max P is the upper limit of the charging and discharging power of electric vehicles; EV (t) is the charge and discharge power of the electric vehicle at time t, where discharge is positive and charge is negative; The constraints of the charging and discharging power of the energy storage device are as follows: ESS,min <P ESS (t) <P ESS,max , P ESS,min is the lower limit of energy storage charging and discharging power; P ESS,max is the upper limit of energy storage charging and discharging power; P ESS (t) is the charge and discharge power at the energy storage time t, discharge is positive and charge is negative; The regional distribution network source-side economic optimization model described in step 2 includes objective functions and constraints. The objective function includes constraints on economic costs and carbon emissions. The constraints on economic costs are as follows: minf2 = C PV +C DG +C EV +C ESS +C N +C TL , C PV is the power generation cost of the photovoltaic power generation system, C DG is the total cost of the diesel generator during operation, C EV is the discharge cost of electric vehicle battery, C ESS is the energy storage system discharge model, C N is the cost of purchasing electricity from the main grid, C TL is the main network loss cost, and the economic cost minf2 is taken from the above C PV , C DG , C EV , C ESS , C N , C TL Minimum value of the sum; The carbon emission constraints include the carbon emission constraints of diesel generators and the carbon emission constraints of main grid power generation. The carbon emissions of diesel generator sets are expressed as: In the formula, α i , β i , γ i are the carbon emission coefficients, α i =24.144kg / (kW·h), β i =1.728kg / (kW·h), γ i =0.0017kg / (kW·h); P i (t) is the output power of the i-th diesel generator set at time t; N is the total number of diesel generator sets; T is the scheduling period; The carbon emissions of the main network are expressed as: F N =λ N P N (t), λ N is the emission coefficient of main grid power generation; P N (t) Purchase electricity for the main grid; The constraints include constraints on power balance, constraints on photovoltaic output, constraints on electric vehicle output and SOC status, constraints on energy storage device output and SOC status, and constraints on diesel generator power; The constraints on power balance are as follows: P i (t) is the output power of the i-th diesel generator set at time t; P(t) is the load value at time t; P loss (t) is the power loss at time t; P PV (t) is the photovoltaic output power at time t; The constraints on photovoltaic output are as follows: PV (t)≤P PV，f (t), P PV，f (t) is the predicted value of photovoltaic output at time t; The output constraint condition of the electric vehicle is as follows: EV，max =λ(t)N EV P EV，c (t)η EV,c , λ(t) is the stopping probability of the electric vehicle at time t; N EV is the number of dispatchable electric vehicles; P EV，c (t) is the charging power of the electric vehicle at time t; P EV，r (t) is the discharge power of the electric vehicle at time t; η EV,c Charging efficiency for electric vehicles; η EV,r Discharge efficiency of electric vehicles; The electric vehicle battery SOC state constraints are as follows: α EV , β EV are the lower and upper limit coefficients of the battery capacity of electric vehicles respectively; q EV,max is the maximum capacity of electric vehicle battery; Q EV (t) is the SOC state of the electric vehicle battery at time t; Q EV,min , Q EV,max They are the lower limit of SOC of electric vehicle battery and the upper limit of SOC of electric vehicle battery respectively; The output constraint of the energy storage system is: ESS,max =P ESS，c (t)η ESS,c , P ESS，c (t) is the charging power at energy storage time t; P ESS，r (t) is the discharge power at energy storage time t; η ESS，c is the energy storage charging efficiency; η ESS，r is the energy storage discharge efficiency; The energy storage system SOC state constraints are as follows: α ESS , β ESS are the lower and upper limit coefficients of energy storage capacity respectively; q ESS,max is the maximum capacity of energy storage; Q ESS (t) is the SOC state at the energy storage time t; Q ESS,min , Q ESS,max They are the lower limit and upper limit of energy storage SOC respectively: The constraints of the diesel generator power are as follows: PD G is the actual output power of the diesel generator; They are the upper and lower limits of diesel generator output power respectively.

2. A method for optimizing the two-side dispatching of a regional distribution network according to claim 1, characterized in that: Using the firefly algorithm to solve the load curve with the minimum economic cost of the load-side economic optimization model of the regional distribution network specifically includes the following steps: 3-1. Based on the total charging cost of electric vehicles, the total charging cost of energy storage and the time-of-use electricity price model, a regional distribution network optimization dispatching model with the minimum economic cost on the load side as the objective function is established; 3-2. Input the data collected by the cloud master station into the model; 3-3. Constrain the data according to the constraints; 3-4. Use the electricity price for 24 hours as the firefly individual, and randomly initialize the initial firefly population with a population size of N; 3-5. Calculate the objective function of each firefly individual, that is, the absolute brightness of the firefly individual, and calculate the Euclidean distance r between firefly i and firefly j ij , the calculation formula is as follows: In the formula, x i 、x j are the spatial positions of firefly i and firefly j respectively; x i，k 、x j，k are the positions of fireflies i and j on dimension k respectively; i = 1, 2, 3, ..., n; j = 1, 2, 3, ..., n; n is the number of fireflies; d is the spatial dimension; 3-6. Update the position of each firefly according to the brightness and Euclidean distance of the fireflies. When firefly i is attracted by another firefly j with higher brightness, the movement of firefly i to firefly j can be expressed as follows: β0 represents the maximum attraction value, which is usually set to 1; γ is the light absorption coefficient; α is the step factor, which is between 0 and 1; rand is a random number that follows a uniform distribution on [0, 1]; 3-7. Repeat steps 3-5 to 3-6 until the number of iterations reaches the maximum value, thereby obtaining the optimized total load curve when the economic cost of the load side including the electric vehicle charging load and the energy storage charging load is the lowest.

3. A method for optimizing the two-side dispatching of a regional distribution network according to claim 1, characterized in that: The total charging cost of the electric vehicle P EV for: T is the scheduling period; C c (t) is the charging electricity price at time t; C r (t) is the discharge electricity price at time t; P EV (t) is the charging and discharging power of the electric vehicle at time t; η EV，c Charging efficiency for electric vehicles; η EV,r is the discharge efficiency of electric vehicles; T o (t) is the optimization period; Total charging cost of energy storage F ESS for: P ESS (t) is the charge and discharge power at energy storage time t; η ESS，c Charging efficiency for energy storage; η ESS,r is the energy storage discharge efficiency.

4. A method for optimizing the two-side dispatching of a regional distribution network according to claim 1, characterized in that: Time-of-use electricity price model f in step 3-1 pv for: C v is the off-peak electricity price; C p is the peak electricity price; C v <C p ; (t1, t2) is the off-peak electricity price period. Charging behavior during this period is calculated and charged according to the off-peak electricity price.

5. A method for optimizing the dispatching of two sides of a regional distribution network according to claim 1, characterized in that: The non-dominated sorting genetic algorithm is used to solve the power generation of each part of the energy on the source side of the regional distribution network, which specifically includes the following calculation steps: 4-1. According to the power generation cost of photovoltaic power generation system, the discharge cost of electric vehicle battery, the discharge cost of energy storage, the total operating cost of diesel generator, and the total cost of main grid power purchase, a regional network optimization scheduling model with the lowest economic cost and minimum carbon emission on the source side as the objective function is constructed; 4-2. Input the data collected by the cloud master station into the model, substitute the power load curve solved by the firefly algorithm into the model, and use the constraint conditions to establish the relevant supply and demand balance constraints for the model; 4-3. Use photovoltaic power generation, diesel generator output power, and main grid power as variables to initialize the population, randomly generate an initial parent population P0 with a population size of N, and generate a child population Q0 through genetic operators, whose population size is also N; 4-4. The parent population P n and the offspring population Q n Merge to form a synthetic population R of size 2N n , perform fast non-dominated sorting, and set R n All 2N individuals in are reclassified according to the non-dominated sequence to obtain levels F1, F2, F3, ...; 4-5. Calculate the individual local crowding distance of each non-dominated layer and sort them. The calculation formula of crowding degree Id is as follows: In the formula, I[i+1].m and I[i-1].m are the function values ​​of the next and previous solutions of solution i on function m respectively; and are the maximum and minimum values ​​of the mth function respectively; 4-6. Select N individuals as the new parent population P according to the sorting results n+1 ; 4-7. Generate a new offspring population Q through genetic operators n+1 ; 4-8 Repeat steps 4-4 to 4-7 until the maximum number of iterations set by the algorithm is reached, and finally the power generation power of each part of the energy on the source side after optimization is obtained.

6. A method for optimizing the two-side dispatching of a regional distribution network according to claim 5, characterized in that: The power generation cost of photovoltaic power generation system C PV for: T is the scheduling period; maintenance costs to operate the PV generator; Cost of managing power for photovoltaic generators; is the unit operation and maintenance coefficient of the photovoltaic generator; P is the photovoltaic generator power generation management cost coefficient; PV (t) PV output power at time t; The electric vehicle battery discharge cost C EV for: P EV，c (t) is the charging power of the electric vehicle at time t; P EV，r (t) is the discharge power of the electric vehicle at time t; C EV，d The battery loss cost of EV charging and discharging; C EV，m The maintenance costs required to charge and discharge electric vehicles; The energy storage discharge cost C ESS for: C c (t) is the charging electricity price at time t; C r (t) is the discharge electricity price at time t; P ESS，c (t) is the charging power at energy storage time t; P ESS，r (t) is the discharge power at energy storage time t; C ESS,d C is the battery loss cost of energy storage charging and discharging; ESS,m The maintenance costs required to charge and discharge the energy storage; The total cost C generated by the diesel generator during operation DG for: C DG1 is the diesel engine operating cost; C DG2 is the maintenance cost of the diesel engine; P DG is the actual power of the diesel generator; a, b, c are the operating cost coefficients of the diesel generator respectively; is the maintenance factor of the diesel engine; The total cost of the main grid electricity purchase is the electricity purchase cost and network loss cost, where the electricity purchase cost C N for: C cn (t) is the unit price of electricity purchased from the main grid at time t; P N (t) is the exchange power with the main network at time t; The network loss cost C TL for: C loss (t) is the network loss price coefficient in time period t; P loss (t) is the power loss in electric energy transmission at time t.

Images