A method for economic optimization dispatch of distribution network considering stochastic correlation between wind and solar power

Abstract

The present invention discloses a distribution network economic optimization scheduling method that takes into account the random correlation between wind and solar power, constructs a joint probability distribution model for wind and solar power output forecast errors, transforms the chance-constrained programming problem into a mixed-integer linear programming optimization problem, and calculates the system's scheduling cost. The method takes minimizing the distribution network scheduling cost and enhancing the system's power supply reliability as optimization goals, establishes a joint probability distribution model for wind and solar power output forecast errors, samples the wind power and photovoltaic output forecast errors, transforms the rotating reserve capacity constraint in the form of chance-constrained programming into a mixed-integer linear programming form, and solves the problem to achieve system optimal economic scheduling. Based on the established distribution network economic optimization scheduling model, the present invention can achieve coordinated operation of new energy and small gas turbines, improve the wind and solar power output forecast error fitting accuracy and the utilization rate of wind and solar new energy, ensure the economy and reliability of distribution network scheduling, and provide a decision-making basis for distribution network scheduling.

Description

Technical Field

[0001] The present invention relates to the field of power system optimization and dispatching, and in particular to a distribution network economic optimization and dispatching method taking into account the random correlation between wind and solar power. Background Art

[0002] Compared to traditional power generation sources like thermal and hydropower, the output of renewable energy sources like wind turbines and photovoltaics is susceptible to a variety of natural factors, including environmental and weather conditions, exhibiting significant uncertainty, including randomness and intermittency. These factors increase the uncertainty of distribution networks, posing significant challenges to system stability, economic viability, and the reliability of power supply to end users. Therefore, accurately characterizing the uncertain output of renewable energy sources like wind and solar power is a crucial step in building a new distribution network.

[0003] In the past, most distribution network scheduling schemes that take into account the randomness of wind and solar power considered wind power and photovoltaic power stations separately. However, in fact, due to the influence of meteorological factors, the output of wind farms and photovoltaic power stations located in the same geographical area is not completely independent, but has a certain correlation in time and space. By accurately characterizing this correlation, the output prediction error can be effectively reduced, thereby helping the system make optimal scheduling decisions and improving the reliability and economy of distribution network operation. Summary of the Invention

[0004] The purpose of the present invention is to provide a distribution network economic optimization scheduling method that takes into account the random correlation of wind and solar power. This method does not assume the probability distribution of wind and solar power output errors, but instead estimates the joint probability distribution of wind and solar power historical output error data based on the multivariate non-parametric kernel density estimation method, and establishes a distribution network economic optimization scheduling model with the goal of minimizing operating costs, providing a decision-making plan for distribution network scheduling, and improving the power supply reliability and economy of the system.

[0005] To achieve the above functions, the present invention designs a distribution network economic optimization scheduling method that takes into account the stochastic correlation between wind and solar power. For a system with wind farms and photovoltaic power stations connected to the grid, the following steps S1 to S5 are performed to complete the analysis of system operating costs and the formulation of a distribution network scheduling plan:

[0006] Step S1: Based on the historical output forecast error data of wind farms and photovoltaic power stations, a joint probability distribution model of wind and solar output forecast error is established based on the non-parametric two-dimensional kernel density estimation method.

[0007] Step S2: Based on the joint probability distribution model of wind and solar power output prediction error According to the calculation method of marginal distribution, the marginal probability distribution model of wind power output error is established respectively and the marginal probability distribution model of photovoltaic output error

[0008] Step S3: Based on the wind power output error marginal probability distribution model Calculate the predicted power error value of each wind farm in each period, and use the photovoltaic output error marginal probability distribution model Calculate the predicted power error value of each photovoltaic power station in each time period; establish the penalty cost of the prediction error based on the predicted power error value of each wind farm in each time period and the predicted power error value of each photovoltaic power station in each time period;

[0009] Step S4: Based on the penalty cost of the prediction error, as well as the cost of the small gas turbine unit, the charging and discharging cost of the energy storage device, the compensation cost for the curtailment of wind and solar power generation, the loss cost caused by load loss, and the cost of purchasing electricity from the upper-level power grid, an objective function with the goal of minimizing the system operation cost is constructed. Constraints are set according to the system power balance constraints, power flow constraints, line current constraints, node voltage constraints, unit output constraints, power purchase constraints, energy storage device constraints, ramping constraints, and spare capacity constraints, and a distribution network economic optimization scheduling model is established.

[0010] Step S5: Based on the distribution network economic optimization dispatch model, the system operation cost is analyzed using the mixed integer linear optimization problem solving method, and a distribution network dispatch plan is formulated.

[0011] Beneficial effects: Compared with the prior art, the advantages of the present invention include:

[0012] The present invention designs a distribution network economic optimization scheduling method that takes into account the random correlation of wind and solar power, realizes the coordinated operation of new energy and small gas turbines, improves the fitting accuracy of wind and solar power output prediction errors and the utilization rate of wind and solar power new energy, ensures the economy and reliability of distribution network scheduling, and provides a decision-making basis for distribution network scheduling. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 This is a workflow diagram of a distribution network economic optimization scheduling method taking into account the random correlation of wind and solar power according to an embodiment of the present invention;

[0014] Figure 2 is a joint probability density diagram of wind and solar power output prediction errors provided according to an embodiment of the present invention;

[0015] Figure 3 is a topological diagram of the IEEE-33 node system structure tested according to an embodiment of the present invention;

[0016] Figure 4 is a daily dispatch curve diagram of wind and solar output compared with a traditional unit without considering the correlation between wind and solar output according to an embodiment of the present invention;

[0017] Figure 5is a graph showing the operating cost, power purchase, and power abandonment at different confidence levels compared with other common error models according to an embodiment of the present invention;

[0018] Figure 6 1 is a graph comparing operating costs and load loss amounts of other common error models under different electricity purchase limits according to an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention.

[0020] Reference Figure 1 The embodiment of the present invention provides a distribution network economic optimization scheduling method that takes into account the stochastic correlation between wind and solar power. For a system in which a wind farm and a photovoltaic power station are connected to the grid, the following steps S1 to S5 are performed to analyze the system operating costs and formulate a distribution network scheduling plan:

[0021] Step S1: Based on the historical output forecast error data of wind farms and photovoltaic power stations, a joint probability distribution model of wind and solar output forecast error is established based on the non-parametric two-dimensional kernel density estimation method.

[0022] The method of establishing a joint probability distribution model of wind and photovoltaic output forecast errors using multivariate nonparametric kernel density estimation method takes into account the correlation between wind and photovoltaic output in space and time. The method uses the characteristics of multivariate joint probability distribution in mathematical statistics that can characterize the correlation between variables, and performs joint probability density estimation on the two variables of wind turbine and photovoltaic output forecast errors, which can improve the accuracy of error fitting.

[0023] Establishing a joint probability distribution model for wind and solar power output forecast errors based on nonparametric two-dimensional kernel density estimation method First, in order to improve the calculation accuracy, the historical output data of wind turbines and photovoltaic power plants are normalized based on the installed capacity of wind farms and photovoltaic power plants, respectively, so that the actual and predicted values are between [0, 1], and the corresponding error values are between [-1, 1]. Then, the kernel function K(·) and the optimal bandwidth matrix H are selected. The kernel function K(·) is calculated as follows: Where x=[x1,x2,…,x d ] T is a d-dimensional random variable, which here refers to the fan output error e W and photovoltaic output error e PV Two variables, ∑ is the covariance matrix of the historical output forecast error sample data of wind farms and photovoltaic power stations; the second is to calculate the optimal bandwidth moment H = diag (h1, ..., h i ,…,h d), the calculation formula is h i =σ i {4 / [(d+2)n s ]} 1 / (d+4) , i=1,2,...,d, where σ i is the standard deviation of the historical data of the i-th variable. The present invention only contains two random variables, wind and solar output prediction errors, that is, d = 2; n s is the number of historical data. According to the selected kernel function K(·) and bandwidth matrix H, the wind and solar power historical output error data can be used to obtain the joint probability distribution of wind and solar power output prediction error. The calculation formula is:

[0024]

[0025] Among them, X i is the i-th sample data; Figure 2 The joint probability density map of wind and solar power output forecast errors is shown.

[0026] Step S2: Based on the joint probability distribution model of wind and solar power output prediction error According to the calculation method of marginal distribution, the marginal probability distribution model of wind power output error is established respectively and the marginal probability distribution model of photovoltaic output error

[0027] Marginal probability distribution model of wind power output error and the marginal probability distribution model of photovoltaic output error As follows:

[0028]

[0029]

[0030] Where, e W and e PV They are wind turbine output error and photovoltaic output error respectively.

[0031] Step S3: Based on the wind power output error marginal probability distribution model Calculate the predicted power error value of each wind farm in each period, and use the photovoltaic output error marginal probability distribution model Calculate the predicted power error value of each photovoltaic power station in each time period; establish the penalty cost of the prediction error based on the predicted power error value of each wind farm in each time period and the predicted power error value of each photovoltaic power station in each time period;

[0032] Due to the randomness and volatility of wind and solar power, the prediction of wind and solar power output will inevitably have a certain degree of deviation. When the wind and solar power output is overestimated, it is necessary to increase the output of conventional energy to make up for the shortfall of wind and solar power output, and the increase in conventional energy consumption will lead to increased carbon emissions; when the wind and solar power output is underestimated, some new energy power will not be able to be connected to the grid, resulting in a waste of resources. In order to take into account the economic and environmental benefits of the system after wind and solar power are connected, the negative impact caused by the prediction deviation must be fully considered when establishing a dispatching model for a wind and solar power distribution network. In order to quantify this negative impact, the penalty cost caused by the prediction error is calculated as Introduced into the objective function, the penalty cost of the prediction error established As follows:

[0033]

[0034]

[0035]

[0036] Where, T is the total number of time periods in the dispatch cycle; m is the number of the photovoltaic power station, M is the total number of photovoltaic power stations; k is the number of the wind farm, K is the total number of wind farms; C PVm and ΔP PVm,t are the prediction error penalty coefficient of the mth photovoltaic power station and the predicted power error value in time period t; C Wk and ΔP Wk,t are the prediction error penalty coefficient of the k-th wind farm and the predicted power error value in period t, respectively.

[0037] Step S4: Based on the penalty cost of the prediction error, as well as the cost of the small gas turbine unit, the charging and discharging cost of the energy storage device, the compensation cost for the curtailment of wind and solar power generation, the loss cost caused by load loss, and the cost of purchasing electricity from the upper-level power grid, an objective function with the goal of minimizing the system operation cost is constructed. Constraints are set according to the system power balance constraints, power flow constraints, line current constraints, node voltage constraints, unit output constraints, power purchase constraints, energy storage device constraints, ramping constraints, and spare capacity constraints, and a distribution network economic optimization scheduling model is established.

[0038] Establishing the economic optimization dispatch model of the distribution network requires setting two parts: the objective function and the constraints. The daily dispatch of the distribution network mainly considers the operating cost of the system. The objective function is to minimize the cost. After introducing the penalty cost, the objective function constructed in step S4 is as follows:

[0039]

[0040] Where, F C is the system operating cost, C Grepresents the cost of small gas turbine unit, C ESS represents the charging and discharging cost of the energy storage device, represents the compensation cost for abandoned wind and solar power, C load-L Indicates the loss cost caused by load loss, C grid-P represents the cost of purchasing electricity from the upper grid, represents the penalty cost;

[0041] The cost of small gas turbine unit is C G The calculation is as follows:

[0042]

[0043] Where n is the number of small gas turbines; N is the total number of small gas turbines; g(P n,t ) represents the fuel cost of small gas turbine n; S n,t is a 0-1 variable representing the state of the small gas turbine, with 1 for startup and 0 for shutdown; n and κ n represent the spinning reserve cost and startup cost respectively; P n,t and R n,t represent the output power and spinning reserve capacity of the small gas turbine n during period t, respectively.

[0044] C ESS The cost of charging and discharging the energy storage device is expressed as follows:

[0045]

[0046] Where r is the energy storage device number, and R is the total number of energy storage devices; and are the discharge power and discharge cost of the energy storage device r during time t; and The corresponding charging power and charging cost.

[0047] The compensation cost for curtailing wind and solar power is expressed as follows:

[0048]

[0049] in, and are the abandoned power and abandoned cost of PV power station m during time t; and are the wind power curtailment and wind cost of wind farm k during time t, respectively.

[0050] C loss-L It represents the loss cost caused by load loss, and the formula is as follows:

[0051]

[0052] Among them, λ and Respectively represent the unit loss load value and loss load amount;

[0053] C grid-P It represents the cost of purchasing electricity from the upper-level power grid, and the formula is as follows:

[0054]

[0055] Among them, C p,t and They represent the electricity purchase price coefficient and the electricity power purchased from the superior power grid respectively.

[0056] The constraints set for the objective function in step S4 are as follows:

[0057] The system power balance constraint is as follows:

[0058]

[0059] Where, P n,t is the output power of the nth small gas turbine; P PVm,t and P Wk,t are the predicted power of the mth photovoltaic power station and the kth wind farm at time t; ΔP Wk,t and ΔP PVm,t are the power prediction errors of the kth wind turbine power station and the mth photovoltaic power station at time t; P ESC,t and P ESF,t are the charging and discharging power of the energy storage device at time t respectively; P is the power purchased from the upper grid at time t; L,t is the load power at time t;

[0060] The power flow constraint is as follows:

[0061]

[0062] Where U i,t 、P i,t and Q i,t are the voltage amplitude, injected active power and reactive power of node i at time t; U j,t is the voltage amplitude of node j at time t; θ i,t and θ j,t are the voltage phase angles between nodes i and j at time t, θ ij,t =θ i,t -θ j,t is the voltage phase angle difference between node i and node j at time t; G ijWith B ij are the conductance and susceptance of the lines at nodes i and j in the distribution network, respectively; i, j∈S, where S is the set of all nodes in the distribution network;

[0063] The line current constraint is as follows:

[0064] I l,t ≤I l,max

[0065] Where, I l,t is the current amplitude of the lth line at time t; I l,max is the maximum current amplitude allowed to flow through the lth line at time t; l∈L, where L is the set of all lines in the distribution network;

[0066] The node voltage constraint is as follows:

[0067] U i,min ≤U i,t ≤U i.max

[0068] Where U i,min with U i,max is the minimum and maximum value of the voltage amplitude of node i at time t;

[0069] The unit output constraint is as follows:

[0070]

[0071] Where, P PVm max and P Wk max The maximum output power of photovoltaic and wind turbine respectively; P n min and P n max They represent the lower and upper limits of the output of small gas turbine n respectively; S n,t It is a 0-1 variable representing the status of the small gas turbine, 1 for startup and 0 for shutdown;

[0072] The power purchase constraint is as follows:

[0073]

[0074] in, The maximum limit of electricity purchased by the system from the upper power grid at time t;

[0075] The energy storage device constraints are as follows. Energy storage devices deployed in distribution networks can smooth power fluctuations, achieve peak load shifting, and improve power quality:

[0076]

[0077]

[0078] Cmin ≤C r,t ≤C max

[0079] in, and are the upper limits of the charging power and discharging power of the energy storage device r at time t, C r,t+1 with C r,t Represents the remaining energy of the energy storage device r in period t+1 and period t respectively; η C and η F are the charging efficiency and discharging efficiency of the energy storage device respectively; Δt is the time variation of the energy storage device during the charging and discharging process; μ ch and μ dch Indicates the operating state coefficient of the energy storage device. When the energy storage device is charging, μ ch is 1 and μ dch is 0, when it discharges μ ch is 0 and μ dch is 1; C min with C max Respectively represent the minimum and maximum values of the remaining energy of the energy storage device;

[0080] The ramp constraint is as follows. The small gas turbine unit bears the main power generation task of the system and should meet the output change constraint when adjusting the output:

[0081]

[0082] in, and are the upward and downward ramp rates of the small gas turbine n, respectively, and Δt1 is the ramp time;

[0083] The spare capacity constraints are as follows:

[0084] To achieve a better balance between system reliability and economic efficiency, the system reserve capacity is set in the form of a chance constraint. The system reserve capacity constraint, which contains the random variables of wind power and photovoltaic output forecast errors, holds true under a certain confidence level. The level of confidence reflects the dispatcher's trade-off between distribution network operation risk and economic efficiency. The reserve constraint includes the positive and negative spinning reserve constraints of the generator sets and the total spinning reserve constraint, as shown in the following formula:

[0085]

[0086] in, are respectively the positive and negative spinning reserve capacities of small gas turbine n during time t; R + 、R - are the positive and negative spinning reserve requirements respectively; β1 and β2 are the confidence levels set to meet the positive and negative spinning reserve requirements, Pnmin,t and P nmax,t They represent the lower and upper limits of the output of the small gas turbine n at time t respectively.

[0087] Step S5: Based on the distribution network economic optimization dispatch model, the system operation cost is analyzed using the mixed integer linear optimization problem solving method, and a distribution network dispatch plan is formulated.

[0088] In step S5, the reserve capacity constraint condition is converted into a mixed integer constraint condition using a sampling-based opportunity constraint deterministic conversion method, and the established distribution network economic optimization scheduling model is solved using a Cplex optimizer to analyze the system operation cost;

[0089] Using Latin hypercube sampling method, random variable ΔP Wk,t , ΔP PVm,t According to the marginal error probability distribution, N sample times sampling, and summing the sample values ε t =ΔP PVm,t +ΔP Wk,t Sort in ascending order, that is:

[0090]

[0091] Among them, s sort (·) is the ascending order function;

[0092] The spare capacity constraint is converted into a mixed integer constraint as follows:

[0093]

[0094] Among them, f floor is the floor function; c ceil is the ceiling function.

[0095] use Figure 3 The improved IEEE-33 node system shown was tested. Figure 3 The numbers 0-32 represent the numbers of the 33 nodes in the IEEE-33 node system; the distribution network economic optimization scheduling model established by using the Cplex optimizer is compared with the traditional unit and wind-solar output daily scheduling curve without considering the correlation between wind and solar output. Figure 4 As shown in the figure, when the correlation between wind and solar output is considered, the output of the generator sets is different from that when there is no correlation. The output of wind farms and photovoltaic power stations increases in most time periods, while the output of the traditional generator sets decreases. Therefore, the scheduling model proposed in this invention that considers the wind and solar combined prediction error is conducive to improving the wind and solar absorption capacity and renewable energy utilization rate of the distribution network, and reducing the system's need to purchase electricity from the upper power grid. Figure 5The present invention compares the operating costs, power purchases and power abandonment of other common error models at different confidence levels. It can be seen that at the same confidence level, compared with the output prediction error without considering the correlation between wind and solar output, the normal distribution error model and the one-dimensional kernel density error model, the two-dimensional kernel density marginal error model taking into account the correlation between wind and solar output has the lowest system daily average operating cost, power purchases and power abandonment. Figure 6 The present invention compares the operating costs and load loss amounts of other common error models under different power purchase limits. It can be seen that under the same power purchase limit, compared with the output prediction error model that does not consider the correlation between wind and solar output, the normal distribution error model and the one-dimensional kernel density error model that ignores the correlation between wind and solar output, the two-dimensional kernel density edge error model that takes into account the correlation between wind and solar output has the lowest system daily average operating cost and daily average load loss amount.

[0096] The embodiments of the present invention are described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Various changes can be made within the scope of knowledge possessed by ordinary technicians in this field without departing from the spirit of the present invention.

Claims

1. A distribution network economic optimization scheduling method taking into account the random correlation of wind and solar power, characterized in that: For a system with wind farms and photovoltaic power stations connected to the grid, perform the following steps S1 to S5 to complete the analysis of system operating costs and the formulation of a distribution network scheduling plan: Step S1: Based on the historical output forecast error data of wind farms and photovoltaic power stations, a joint probability distribution model of wind and solar output forecast error is established based on the non-parametric two-dimensional kernel density estimation method. Select the kernel function K(·) and the optimal bandwidth matrix H, and the calculation formula is as follows: Where x=[x1,x2,…,x d ] T is a d-dimensional random variable; ∑ is the covariance matrix of the sample data of historical output forecast errors of wind farms and photovoltaic power stations; h i =σ i {4 / [(d+2)n s ]} 1 / (d+4) ,i=1,2,...,d Where h i represents the bandwidth of the i-th random variable, σ i is the standard deviation of the historical data of the i-th variable; Where x is the fan output error e W and photovoltaic output error e PV Two variables; the i-th sample data is recorded as X i =[X 1i ,X 2i ,…,X di ] T , i=1,2,…,n s , n s is the total number of samples; Step S2: Based on the joint probability distribution model of wind and solar power output prediction error According to the calculation method of marginal distribution, the marginal probability distribution model of wind power output error is established respectively and the marginal probability distribution model of photovoltaic output error As follows: Where, e W and e PV They are wind turbine output error and photovoltaic output error respectively; Step S3: Based on the wind power output error marginal probability distribution model Calculate the predicted power error value of each wind farm in each period, and use the photovoltaic output error marginal probability distribution model Calculate the predicted power error value of each photovoltaic power station in each time period; establish the penalty cost of the prediction error based on the predicted power error value of each wind farm in each time period and the predicted power error value of each photovoltaic power station in each time period; Step S4: Based on the penalty cost of the prediction error, as well as the cost of the small gas turbine unit, the charging and discharging cost of the energy storage device, the compensation cost for the curtailment of wind and solar power generation, the loss cost caused by load loss, and the cost of purchasing electricity from the upper-level power grid, an objective function with the goal of minimizing the system operation cost is constructed. Constraints are set according to the system power balance constraints, power flow constraints, line current constraints, node voltage constraints, unit output constraints, power purchase constraints, energy storage device constraints, ramping constraints, and spare capacity constraints, and a distribution network economic optimization scheduling model is established. Step S5: Based on the distribution network economic optimization dispatch model, the system operation cost is analyzed using the mixed integer linear optimization problem solving method, and a distribution network dispatch plan is formulated.

2. The method for economic optimization and dispatching of distribution network taking into account the stochastic correlation of wind and solar power according to claim 1 is characterized in that: The penalty cost for the prediction error established in step S3 As follows: Where, T is the total number of time periods in the dispatch cycle; m is the number of the photovoltaic power station, M is the total number of photovoltaic power stations; k is the number of the wind farm, K is the total number of wind farms; C PVm and ΔP PVm,t are the prediction error penalty coefficient of the mth photovoltaic power station and the predicted power error value in time period t; C Wk and ΔP Wk,t are the prediction error penalty coefficient of the k-th wind farm and the predicted power error value in period t, respectively.

3. The method for economic optimization and dispatching of distribution network taking into account the stochastic correlation of wind and solar power according to claim 1, characterized in that: The objective function constructed in step S4 is as follows: Where, F C is the system operating cost, C G represents the cost of small gas turbine unit, C ESS represents the charging and discharging cost of the energy storage device, represents the compensation cost for abandoned wind and solar power, C load-L Indicates the loss cost caused by load loss, C grid-P represents the cost of purchasing electricity from the upper grid, represents the penalty cost.

4. The method for economic optimization and dispatching of distribution network taking into account the stochastic correlation of wind and solar power according to claim 3 is characterized in that: The constraints set for the objective function in step S4 are as follows: The system power balance constraint is as follows: Where, P n,t is the output power of the nth small gas turbine; P PVm,t and P Wk,t are the predicted power of the mth photovoltaic power station and the kth wind farm at time t; ΔP Wk,t and ΔP PVm,t are the power prediction errors of the kth wind turbine power station and the mth photovoltaic power station at time t; P ESC,t and P ESF,t are the charging and discharging power of the energy storage device at time t respectively; P is the power purchased from the upper grid at time t; L,t is the load power at time t; The power flow constraint is as follows: Where U i,t 、P i,t and Q i,t are the voltage amplitude, injected active power and reactive power of node i at time t; U j,t is the voltage amplitude of node j at time t; θ i,t and θ j,t are the voltage phase angles between nodes i and j at time t, θ ij,t =θ i,t -θ j,t is the voltage phase angle difference between node i and node j at time t; G ij With B ij are the conductance and susceptance of the lines at nodes i and j in the distribution network, respectively; i, j∈S, where S is the set of all nodes in the distribution network; The line current constraint is as follows: I l,t ≤I l,max Where, I l,t is the current amplitude of the lth line at time t; I l,max is the maximum current amplitude allowed to flow through the lth line at time t; l∈L, where L is the set of all lines in the distribution network; The node voltage constraint is as follows: IN i,min ≤U i,t ≤U i.max Where U i,min with U i,max is the minimum and maximum value of the voltage amplitude of node i at time t; The unit output constraint is as follows: Where, P PVm max and P Wk max The maximum output power of photovoltaic and wind turbine respectively; P n min and P n max They represent the lower and upper limits of the output of small gas turbine n respectively; S n,t It is a 0-1 variable representing the status of the small gas turbine, 1 for startup and 0 for shutdown; The power purchase constraint is as follows: in, The maximum limit of electricity purchased by the system from the upper power grid at time t; The energy storage device constraints are as follows: C min ≤C r,t ≤C max in, and are the upper limits of the charging power and discharging power of the energy storage device r at time t, C r,t+1 with C r,t Represents the remaining energy of the energy storage device r in period t+1 and period t respectively; η C and η F are the charging efficiency and discharging efficiency of the energy storage device respectively; Δt is the time variation of the energy storage device during the charging and discharging process; μ ch and μ dch Indicates the operating state coefficient of the energy storage device. When the energy storage device is charging, μ ch is 1 and μ dch is 0, when it discharges μ ch is 0 and μ dch is 1; C min with C max Respectively represent the minimum and maximum values of the remaining energy of the energy storage device; The climbing constraint is as follows: in, and are the upward and downward ramp rates of the small gas turbine n, respectively, and Δt1 is the ramp time; The spare capacity constraint is as follows: in, are respectively the positive and negative spinning reserve capacities of small gas turbine n during time t; R + 、R - are the positive and negative spinning reserve requirements respectively; β1 and β2 are the confidence levels set to meet the positive and negative spinning reserve requirements, P nmin,t and P nmax,t They represent the lower and upper limits of the output of the small gas turbine n at time t, is the abandoned power of photovoltaic power station m at time t, is the abandoned wind power of wind farm k at time t.

5. The method for economic optimization and dispatching of distribution network taking into account the stochastic correlation of wind and solar power according to claim 1, characterized in that: In step S5, the reserve capacity constraint condition is converted into a mixed integer constraint condition using a sampling-based opportunity constraint deterministic conversion method, and the established distribution network economic optimization scheduling model is solved using a Cplex optimizer to analyze the system operation cost; Using Latin hypercube sampling method, random variable ΔP Wk,t , ΔP PVm,t According to the marginal error probability distribution, N sample times sampling, and summing the sample values ε t =ΔP PVm,t +ΔP Wk,t Sort in ascending order, that is: Among them, s sort (·) is the ascending order function; The spare capacity constraint is converted into a mixed integer constraint as follows: Among them, f floor is the floor function; c ceil is the ceiling function.

Images