A multi-objective optimization scheduling method for active distribution network with distributed energy storage

Abstract

The application discloses a kind of active distribution network multi-objective optimization scheduling method containing distributed energy storage, belongs to power grid dispatching optimization technical field, its technical characteristics include: with peak clipping as target, time-of-use pricing is formulated as adjusting means, and demand response model is established based on demand price elasticity theory;With the minimum voltage deviation, the comprehensive operation cost as the target, the multi-objective optimization scheduling model of active distribution network is established;An improved multi-objective sparrow search algorithm is proposed as the algorithm tool for solving the model.The application combines power supply side optimization scheduling problem with demand side response problem, carries out comprehensive planning of resources, establishes "source-load-storage" multi-objective optimization scheduling model based on considering demand response, and solves the model using the improved multi-objective sparrow search algorithm, which is conducive to improving the economy and safety of distribution network operation.

Description

Technical Field

[0001] This invention relates to active distribution network optimization scheduling technology, specifically to a multi-objective optimization scheduling method for active distribution networks including distributed energy storage. Background Technology

[0002] Renewable energy generation technologies in distribution networks are receiving increasing attention. The intermittent and random nature of renewable energy output impacts the safe and stable operation of the system. Improving the distribution network's capacity to absorb renewable energy is a crucial aspect of smart grid development. Active distribution networks can autonomously regulate distributed power sources, energy storage devices, and controllable demand-side loads through control measures. They can also adaptively adjust their network structure, generation units, and loads based on the actual operating status of the distribution network to achieve safe and economical operation. However, coordinating various distributed energy sources, energy storage devices, and controllable demand-side loads within an active distribution network, and effectively integrating them into the current electricity market to improve voltage quality and operational efficiency, remains a pressing issue. Therefore, it is necessary to develop an optimized dispatching method for active distribution networks incorporating distributed energy storage.

[0003] Most existing distribution network optimization scheduling methods are designed for traditional distribution networks and do not consider the impact of user-side demand response on the economic efficiency of scheduling. Furthermore, the search capability and solution accuracy of scheduling model solving algorithms still need further improvement. To address this, this invention proposes an active distribution network multi-objective optimization scheduling method that incorporates distributed energy storage. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies by providing a multi-objective optimization scheduling method for active distribution networks with distributed energy storage.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] This invention discloses a multi-objective optimization scheduling method for an active distribution network including distributed energy storage, comprising the following steps:

[0007] S1. With peak shaving and valley filling as the goal and time-of-use pricing as the means of regulation, a demand-side time-of-use electricity price response model is established based on the theory of demand price elasticity.

[0008] S2. Using time-of-use pricing schemes, distributed generation, and day-ahead planned output of energy storage as decision variables, a multi-objective optimization scheduling model for active distribution networks that considers demand response was established.

[0009] S3. Based on the sparrow search algorithm, an improved multi-objective sparrow search algorithm is constructed from the perspectives of improving the diversity of the initial population and enhancing the global search capability of the population.

[0010] S4. First, use the improved multi-objective sparrow search algorithm to optimize the demand response and obtain the optimal load curve. Then, solve the multi-objective optimization scheduling scheme of the distribution network under the load curve.

[0011] The beneficial effects achieved by this invention are as follows: It considers the demand-side response of users in traditional optimal scheduling methods, strengthens the interaction between the power supply and consumption sides, and comprehensively plans resources, thus helping to improve the economic efficiency of power grid operation. It introduces Tent mapping and reverse learning strategies into the standard sparrow search algorithm, improving the quality of the initial solutions and increasing the probability of the algorithm finding the optimal solution. The improved Lévy flight strategy further enhances the algorithm's global search capability and convergence accuracy. Compared with traditional multi-objective optimal scheduling methods for distribution networks, this method is more economical and reliable. Attached Figure Description

[0012] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0013] Figure 1 This is a flowchart illustrating a multi-objective optimization scheduling method for an active distribution network that includes distributed energy storage.

[0014] Figure 2 This is the flowchart for the optimization algorithm. Detailed Implementation

[0015] To explain in detail the technical content, objectives, and effects of the present invention, the following description is provided in conjunction with the embodiments and accompanying drawings.

[0016] Please refer to Figure 1 and Figure 2 A multi-objective optimization scheduling method for active distribution networks with distributed energy storage includes the following steps:

[0017] S1. With peak shaving and valley filling as the goal and time-of-use pricing as the means of regulation, a demand-side time-of-use electricity price response model is established based on the theory of demand price elasticity.

[0018] Step S1 specifically includes the following steps:

[0019] The S1-1 theory of price elasticity of demand describes the behavioral patterns of users in demand response, specifically the ratio of the change in demand to the change in electricity price per unit time. The price elasticity coefficient represents the demand-side capacity to reduce load. The price elasticity coefficient is divided into self-elasticity coefficient and reciprocal elasticity coefficient, and its expression is:

[0020]

[0021] In the formula, the subscripts i and j are positive integers used to represent time periods, i = 1, 2, 3…24, j = 1, 2, 3…24, ε i,i ε is the self-elasticity coefficient; if the electricity price decreases during period i, the user increases their electricity consumption. i,j Let ε be the reciprocity coefficient. If the electricity price increases during time period j, the electricity consumption of users in time period j decreases and shifts to time period i, causing the electricity consumption of users in time period i to increase, i.e., ε i,i <0, ε i,j >0; ΔD i =D i -D i,0 , Δp i =p i -p i,0 , Δp j =p j -p j,0 ;D i,0 and D i Let ΔD be the original load and the load after response for time period i. i p represents the difference between the load after the response in time period i and the original load; i,0 and p i Let Δp be the initial electricity price and the changed electricity price for time period i. i Let p be the difference between the changed electricity price and the initial electricity price during time period i; j,0 and p j Let Δp be the initial electricity price and the changed electricity price for time period j. j Let be the difference between the electricity price after the change in time period j and the initial electricity price;

[0022] S1-2 uses an electricity price elasticity matrix to express the impact of electricity price changes on electricity consumption. The day is divided into 24 time periods, and the expression for the elasticity matrix is:

[0023]

[0024] S1-3 uses the uniform electricity price as a reference price and makes appropriate adjustments to the electricity price, reasonably increasing the price during peak electricity consumption periods and reasonably decreasing the price during off-peak periods. The expression for determining the electricity price for each period is as follows:

[0025]

[0026] In the formula, p is the 24-hour price matrix; p f The revised electricity price for peak load periods; p p For the changed electricity price during the grid parity period; p v The revised electricity price for off-peak hours; p ave The initial uniform electricity price is σ1 and σ2 are the peak and off-peak electricity price adjustment parameters, respectively. The peak electricity price is directly proportional to σ1, and the off-peak electricity price is inversely proportional to σ2.

[0027] S1-4, derived from the electricity consumption change matrix, provides the electricity demand for each time period after the price response. The user electricity consumption for each time period after the price response is as follows:

[0028]

[0029] S1-5 uses time-of-use pricing as the control method and minimizes the peak-valley load difference as the objective function to establish a price response model. The objective function is:

[0030] minF = (max(q) - min(q))

[0031] In the formula, minF is the minimum load peak-valley difference, q is the time series of user electricity consumption within the load dispatching cycle, which is generally the load distribution data within 24 hours, and max(q) and min(q) are the maximum and minimum load values ​​within the dispatching cycle;

[0032] The requirements response constraints are as follows:

[0033] (1) Electricity price upper and lower limits during peak and off-peak periods:

[0034]

[0035] In the formula, p max and p min These are the upper and lower limits of electricity prices, respectively.

[0036] (2) Considering the interests of both the power supply side and the power consumption side, the peak-valley electricity price ratio satisfies the following constraint:

[0037]

[0038] In the formula, ξ is the peak-valley electricity price ratio, ξ = p f / p v ξ min and ξ max The minimum and maximum peak-valley electricity price ratio;

[0039] Meanwhile, based on user electricity consumption and electricity prices before and after demand response, the expression for demand-side response cost can be derived as follows:

[0040] C tou = p0(t)·D0(t) - p(t)·D(t)

[0041] In the formula, C tou Let p0(t) be the demand-side response cost, p0(t) be the original electricity price at time t, D0(t) be the load demand at the original electricity price, p(t) be the electricity price at time t after the demand-side response, and D(t) be the load demand after the optimal demand-side response when the electricity price is p(t).

[0042] S2. Using time-of-use pricing schemes, distributed generation, and day-ahead planned output of energy storage as decision variables, a multi-objective optimization scheduling model for active distribution networks that considers demand response was established.

[0043] Step S2 specifically includes the following steps:

[0044] Based on electricity price response, S2-1 performs proactive distribution network "source-load-storage" coordinated optimization scheduling. The scheduling objective is to minimize the overall operating cost of the distribution network, which includes the cost of purchasing electricity from the upper-level grid, generation costs, operation and maintenance costs of energy storage, and demand response costs. The objective function for the economic optimization scheduling of the distribution network is:

[0045]

[0046] In the formula, F1 is the economic objective function for optimal dispatching of the distribution network, and T is the dispatching period, which is taken as T = 24 hours; C b (t) and P b (t) represents the electricity purchase price and electricity volume of the upper-level power grid during time period t; DG represents distributed generator units, N DG C represents the number of distributed generator sets. DG (t) and P DG,i (t) represents the output cost and active power of the i-th controllable DG in time period t; N ESS For the number of energy storage devices, C ESS (t) and P ESS,j (t) represents the operating cost and active power output of the j-th energy storage device during time period t; C tou (t) represents the demand-side response cost during time period t;

[0047] The voltage at node S2-2 is one of the important indicators for measuring system stability and power quality. The voltage at each node in the distribution network should be kept near its rated value and the fluctuation should be limited to a certain range. Therefore, the second objective of the optimization model is to minimize the voltage deviation, and the objective function is as follows:

[0048]

[0049] In the formula, F2 is the objective function for optimizing voltage deviation, and N bus U represents the number of system nodes. i,t Let be the voltage amplitude at node i at time t. This represents the average voltage of node i during the scheduling period.

[0050] In solving S2-3 for distribution network optimization, equality constraints, i.e., power flow constraints at distribution network nodes, must be satisfied. Inequality constraints must also be satisfied, i.e., all state variables and control variables must remain within defined ranges. Specific constraints are as follows:

[0051] System power balance constraints:

[0052]

[0053] In the formula, P G,i P DG,i and P L,i These represent the active power output of the generator, the active power output of the generator (DG), and the active power demand of the load at node i, respectively; Q G,i Q DG,i and Q L,i These are, respectively, the reactive power output of the generator at node i, the reactive power output of the generator (DG), and the reactive power demand of the load; U i Let U be the voltage at node i. j G represents the voltage at node j; ij and B ij θ represents the conductance and susceptance between nodes i and j, respectively; ij Let be the phase angle difference between nodes i and j.

[0054] Controllable DG output constraints:

[0055] P DG,i,min ≤P DG,i (t)≤P DG,i,max

[0056]

[0057] In the formula, P DG,i (t), P DG,i,min and P DG,i,max P represents the output, minimum output, and maximum output of the i-th generator during time period t, respectively; DG,i (t) and P DG,i (t-1) represent the power of the i-th controllable DG during time periods t and t-1, respectively; U DG,i and D DG,i These are the upper and lower limits of the ramp rate for the i-th DG, respectively;

[0058] Node voltage constraints:

[0059] U min ≤U i ≤U max

[0060] In the formula, U max and U min These are the upper and lower limits of the voltage amplitude at node i, respectively;

[0061] Transmission line power constraints:

[0062] P j ≤P j,max

[0063] In the formula, P j Let P be the transmission power of the j-th line. j,max This represents the maximum allowable transmission power for the j-th line.

[0064] Energy storage operation constraints:

[0065]

[0066] In the formula, P ESS,j (t) represents the charging and discharging power of the j-th energy storage at time t; These are the upper and lower limits of the charge and discharge power of the j-th energy storage device, respectively; SOC j (t) represents the state of charge of the j-th energy storage at time t. These represent the upper and lower limits of the j-th energy storage state of charge, respectively. Considering the impact of charging and discharging power on the energy storage lifespan, the state of charge at the initial and final moments of an energy storage system within a scheduling cycle is set to 0.4; the time interval Δt... The energy storage charging power during time period t. Let λ be the energy storage discharge power during time period t. c , λ d σ represents the energy storage charging and discharging efficiency, and σ represents the self-discharge current rate.

[0067] S3. Based on the sparrow search algorithm, an improved multi-objective sparrow search algorithm is constructed from the perspectives of improving the diversity of the initial population and enhancing the global search capability of the population.

[0068] Step S3 specifically includes the following steps:

[0069] Before the algorithm begins iteration, the initial sparrow population in S3-1 is generated using a Tent chaotic mapping and a reverse learning strategy. However, Tent chaotic iteration has unstable periodic points. To avoid this problem, a random variable is introduced for improvement. The improved expression is:

[0070]

[0071] In the formula, Z k Z is a random number between (0,1). k+1 The result after mapping is λ, which takes values ​​between (0,1), typically 0.49. rand(0,1) represents a random number within the range (0,1); N is the number of elements in the chaotic sequence. If X... i ∈[lb,ub], where ub and lb are the upper and lower bounds of the population search space, respectively. Mapping the generated chaotic sequence to the solution space, the expression for an individual sparrow is:

[0072] X i=lb+(ub-lb)·Z

[0073] In the formula, X i Let Z be the i-th sparrow individual, and Z be the result after the Tent chaotic mapping. Backward learning constructs its backward solution from the feasible solutions at the current position to increase the diversity of the population. It then selects the better solution from the current solution and the backward solution as the next generation of individuals. The formula for calculating the backward solution is:

[0074]

[0075] In the formula, For each sparrow, K is the inverse solution, where K is the dynamic coefficient taking values ​​in (0,1). The specific steps for initializing the population using Tent chaotic mapping and inverse learning strategy are as follows:

[0076] Step 1: Map the sequence of chaotic mappings to individual sparrows using the Tent chaotic mapping sequence and calculate the fitness of the population;

[0077] Step 2: Calculate the inverse solution population of the population after the Tent chaotic mapping, and calculate the fitness of the population;

[0078] Step 3: Merge the Tent chaotic map population and the reverse solution population, sort them according to their fitness values, and select the top N sparrow individuals as the initial population;

[0079] In the sparrow search algorithm S3-2, the discoverer plays a guiding role in the population. If the discoverer gets stuck in a local optimum, it can easily cause the entire population to stagnate. Therefore, a Lévy flight perturbation strategy is introduced to update the discoverer's position. Based on the basic Lévy flight strategy, the step size factor is changed to a dynamic value that changes with the number of iterations. The expression for the step size factor is:

[0080]

[0081] In the formula, γ(i) is the step size in the i-th iteration, t is the current iteration number, and T max Let ζ be the maximum number of iterations, exp(·) be the adjustment parameter, sinh(·) be the exponential function with base e, and sinh(·) be the hyperbolic sine function. The improved expression for updating the discoverer's position using the Levy flight strategy is:

[0082]

[0083]

[0084] In the formula, The current optimal solution; λ∈(1,3), generally taking the value λ=1.5; u and v follow a normal random distribution; ψ is the perturbation constant, г(·) is the gamma function; t represents the current iteration number; L represents a 1×D dimensional vector with all elements being 1; This represents the position of the i-th sparrow in the j-th dimension of the (t+1)-th generation. Let Ri represent the position of the i-th sparrow in the j-th dimension of the t-th generation; Q is a random number that follows a normal distribution of [0,1]; R2 and ST are the warning value and the safety value, respectively. When R2 < ST, it means that there are no natural enemies nearby, and the discoverer will conduct a wider search; when R2 ≥ ST, it means that the sparrow has discovered a natural enemy, and the entire population will adjust its search strategy and quickly move to a safe area.

[0085] The joiner updates its position by following the discoverer, and the formula is:

[0086]

[0087] In the formula, Let $t$ be the worst-case position of the sparrow in the $j$ dimension during the $t$-th iteration. Let A represent the optimal position of the discoverer in the j-th dimension at the (t+1)-th iteration; let A be a 1×D dimensional vector, where each element is randomly assigned the value 1 or -1. T Let A be the transpose of A, and A + =A T (AA T ) -1 N represents the number of sparrows in the population. When i > N / 2, it means that the less fit sparrows need to fly to other areas to find food. When i ≤ N / 2, it means that the current sparrows will choose to forage near the best location of the discoverer.

[0088] When an early warning animal senses danger, it will exhibit anti-predation behavior, and its location will be updated as follows:

[0089]

[0090] In the formula, Let f be the global optimal position at the t-th iteration; β is the step size control parameter following N(0,1); k∈[-1,1]; f i f represents the current fitness value of the sparrow. g f represents the current globally optimal fitness value. w This represents the current worst global fitness value; ε is a constant close to 0 to avoid a denominator of 0.

[0091] S3-3 uses an external archive outside the population to store and update the Pareto optimal solution set with the non-dominated solutions generated in each iteration of the algorithm. Since the external archive has limited capacity, and the population size in the external archive may increase during the update process, to maintain population diversity and prevent the population size from exceeding the limit, a method is used to remove similar individuals based on the crowding distance to maintain the balance of the Pareto solution set. The crowding distance is calculated using the following formula:

[0092]

[0093] In the formula, D(j) is the crowding distance of the j-th sparrow in the Pareto solution set, and m is the number of objective functions; F i (j+1) and F i (j-1) represents the j-th objective function value between the two adjacent individuals of the j-th sparrow individual, F i,max F i,min These are the maximum and minimum values ​​of the i-th objective function, respectively.

[0094] S4. First, use the improved multi-objective sparrow search algorithm to optimize the demand response and obtain the optimal load curve. Then, solve the multi-objective optimization scheduling scheme of the distribution network under the load curve.

[0095] Combination Figure 2 It can be seen that step S4 specifically includes the following steps:

[0096] S4-1 Inputs relevant data on the active distribution network, including distribution network structure parameters, distributed generation forecast data, load forecast data, demand response parameters, and related variable constraints. Sets the parameters for the multi-objective optimization algorithm, including population size, maximum number of iterations, discoverer ratio, number of early warning sparrows, and early warning value.

[0097] S4-2 Initialize the demand response population. Using time-of-use electricity price as the decision variable, within the constraints, the position of the demand response population Q2 is initialized using Tent chaotic mapping and a back-learning strategy.

[0098] S4-3 Iterative computation of the demand response population Q2. The objective function is the load peak-to-valley difference F, and the optimization method is a single-objective sparrow search algorithm. An improved Lévy flight strategy is used to update the positions of the discoverers in the population Q2.

[0099] S4-4 determines whether the demand response conditions have ended. If the maximum number of iterations has been reached, obtain the day-ahead load curve under the optimal peak shaving and valley filling scheme, and calculate the demand-side response cost; otherwise, return to step S43 to continue running.

[0100] S4-5 Initializes and optimizes the scheduling population Q1. Based on the updated load curve and demand response cost, and using the power output of controllable DG and energy storage at different time periods as decision variables, the position of the demand response population Q1 is initialized within constraints using Tent chaotic mapping and a back-learning strategy. The fitness values ​​F1 and F2 of each individual are calculated according to the objective function.

[0101] S4-6 obtains non-dominated solutions based on individual dominance relationships and stores them in the external archive set arav; an improved Levy flight strategy is used to update the positions of discoverers in population Q1.

[0102] S4-7 External Archives Update. When the number of individuals in an archive exceeds the maximum size limit, it is reduced based on the crowding distance.

[0103] S4-8 determines whether the termination condition is met. If the condition is met, the optimal scheduling scheme is output; otherwise, proceed to step S4-6 to continue running.

[0104] It should be noted that the above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A multi-objective optimization scheduling method for an active distribution network including distributed energy storage, characterized in that, Includes the following steps: S1. With peak shaving and valley filling as the goal and time-of-use pricing as the means of regulation, a demand-side time-of-use electricity price response model is established based on the theory of demand price elasticity. S2. Using time-of-use pricing schemes, distributed generation and energy storage day-ahead planned output as decision variables, an active distribution network multi-objective optimization scheduling model considering demand response was established. Step S2 includes: Based on electricity price response, S2-1 performs proactive distribution network "source-load-storage" coordinated optimization scheduling. The scheduling objective is to minimize the overall operating cost of the distribution network, which includes the cost of purchasing electricity from the upper-level grid, generation costs, operation and maintenance costs of energy storage, and demand response costs. The objective function for the economic optimization scheduling of the distribution network is: ； In the formula, F1 is the economic objective function for optimal dispatching of the distribution network. For the scheduling period, take Hour; and for The time period refers to the electricity purchase price and volume of the upper-level power grid; DG represents distributed generator units. The number of distributed generator sets. and For the first Taiwan-controlled DG in the first The cost of effort and the amount of work done over a given time period; The number of energy storage devices, and For the first Taiwan energy storage device in the first Operating costs and active power output over a given time period; for Demand-side response costs during specific time periods; The voltage at node S2-2 is one of the important indicators for measuring system stability and power quality. The voltage at each node in the distribution network should be kept near its rated value and the fluctuation should be limited to a certain range. Therefore, the second objective of the optimization model is to minimize the voltage deviation, and the objective function is as follows: ； In the formula, Optimize the objective function for voltage deviation. This represents the number of system nodes. for Time Node Voltage amplitude at that point for The average voltage of the node during the scheduling period; S3. Based on the sparrow search algorithm, an improved multi-objective sparrow search algorithm is constructed from the perspectives of improving the initial population diversity and enhancing the global search capability of the population. S4. First, use the improved multi-objective sparrow search algorithm to optimize the demand response and obtain the optimal load curve. Then, solve the multi-objective optimization scheduling scheme of the distribution network under the load curve.

2. The multi-objective optimization scheduling method for an active distribution network with distributed energy storage according to claim 1, characterized in that, Step S1 specifically includes the following steps: The S1-1 theory of price elasticity of demand describes the behavioral patterns of users in demand response, specifically the ratio of the change in demand to the change in electricity price per unit time. The price elasticity coefficient represents the demand-side capacity to reduce load. The price elasticity coefficient is divided into self-elasticity coefficient and reciprocal elasticity coefficient, and its expression is: ； In the formula, the subscripts i and j are positive integers used to represent time periods, i = 1, 2, 3…24, j = 1, 2, 3…24. Let be the self-elasticity coefficient, if Electricity prices decrease during certain time periods, leading to increased electricity consumption by users. Let be the mutual elasticity coefficient, if Electricity prices rise during certain periods. During periods when user electricity consumption decreases, it is shifted to Time period Increased user electricity consumption during a certain period, i.e. , ; , , ; and for The original load and the load after response during the time period, ΔD i for The difference between the load after the time-period response and the original load; and for Initial electricity price and changed electricity price for the time period, Δp i for The difference between the electricity price after the time period change and the initial electricity price; and for Initial electricity price and changed electricity price for the time period, Δp j Let be the difference between the electricity price after the change in time period j and the initial electricity price; S1-2 uses an electricity price elasticity matrix to express the impact of electricity price changes on electricity consumption. The day is divided into 24 time periods, and the expression for the elasticity matrix is: ； S1-3 uses the uniform electricity price as a reference price and makes appropriate adjustments to the electricity price, reasonably increasing the price during peak electricity consumption periods and reasonably decreasing the price during off-peak periods. The expression for determining the electricity price for each period is as follows: ； In the formula, A 24-hour price matrix; The revised electricity price for peak load periods; The revised electricity price for the period of grid parity; The revised electricity price for off-peak hours; The initial uniform electricity price; , These are the electricity price adjustment parameters for peak and off-peak hours, respectively. Peak hour electricity price and... It is directly proportional to the electricity price during off-peak hours. Inversely proportional; S1-4, derived from the electricity consumption change matrix, provides the electricity demand for each time period after the price response. The user electricity consumption for each time period after the price response is as follows: ； S1-5 uses time-of-use pricing as the control method and minimizes the peak-valley load difference as the objective function to establish a price response model. The objective function is: ； In the formula, minF is the minimum load peak-to-valley difference. Let max(q) and min(q) be the time series of user electricity consumption within the load scheduling period, and max(q) and min(q) be the maximum and minimum load values ​​within the scheduling period. The requirements response constraints are as follows: (1) Electricity price upper and lower limits during peak and off-peak periods: ； In the formula, p max and p min These are the upper and lower limits of electricity prices, respectively. (2) Considering the interests of both the power supply side and the power consumption side, the peak-valley electricity price ratio satisfies the following constraint: ； In the formula, ξ is the peak-valley electricity price ratio, ξ=p f / p v , and The minimum and maximum peak-valley electricity price ratio; Meanwhile, based on user electricity consumption and electricity prices before and after demand response, the expression for demand-side response cost can be derived as follows: ； In the formula, C tou For demand-side response costs, for Original electricity price at any time The load demand at the original electricity price. After the demand-side response Electricity price at any time The electricity price is The load demand after the optimal demand-side response at that time.

3. The multi-objective optimization scheduling method for an active distribution network including distributed energy storage according to claim 1, characterized in that, Step S2 further includes: In solving S2-3 for distribution network optimization, equality constraints, i.e., power flow constraints at distribution network nodes, must be satisfied. Inequality constraints must also be satisfied, i.e., all state variables and control variables must remain within defined ranges. Specific constraints are as follows: System power balance constraints: ； In the formula, , and These represent the nodes respectively. Generator active power output, DG active power output, and load active power demand; , and They are at the nodes Generator reactive power output, DG reactive power output, load reactive power demand; For nodes voltage, For nodes The voltage; and Representing nodes respectively and Between conductance and susceptance; For nodes and The phase angle difference; Controllable DG output constraints: ， ； In the formula, , and The first A generator in Output during a given time period, minimum output, and maximum output; and The first Taiwan-controlled DG and Power during a given time period; and The first The upper and lower limits of the ramp rate of Taiwan DG; Node voltage constraints: ； In the formula, and They are nodes The upper and lower limits of voltage amplitude; Transmission line power constraints: ； In the formula, For the first Line transmission power, For the first Maximum allowable transmission power of the line; Energy storage operation constraints: ； In the formula, For the first Energy storage The charging and discharging power at any given moment; , The first The upper and lower limits of the energy storage charging and discharging power; For the first Energy storage State of charge at time t, , The first The upper and lower limits of the state of charge of each energy storage are set. Considering the impact of charging and discharging power on the lifespan of energy storage, the state of charge of energy storage at the beginning and end of a scheduling cycle is set to 0.

4. Time interval, for Time-of-use energy storage charging power, for Time-of-use energy storage discharge power, , These are the energy storage charging and discharging efficiencies, respectively. This represents the self-discharge current rate.

4. The multi-objective optimization scheduling method for an active distribution network with distributed energy storage according to claim 1, characterized in that, Step S3 specifically includes the following steps: Before the algorithm begins iteration, the initial sparrow population in S3-1 is generated using a Tent chaotic mapping and a reverse learning strategy. However, Tent chaotic iteration has unstable periodic points. To avoid this problem, a random variable is introduced for improvement. The improved expression is: ； In the formula, for Random numbers between The result after mapping, It takes values ​​between (0,1). The range is indicated in Random numbers between; Let be the number of elements in the chaotic sequence, if , , Let be the upper and lower bounds of the population search space, respectively. Mapping the generated chaotic sequence to the solution space, the expression for an individual sparrow is: ； In the formula, For the first Each sparrow, As the result of the Tent chaotic mapping, backward learning constructs a backward solution from the feasible solutions at the current position to increase the diversity of the population. It then selects the better solution from the current solution and the backward solution as the next generation of individuals. The formula for calculating the backward solution is: ； In the formula, The reverse solution for the individual sparrow. To take values ​​in The dynamic coefficients on the above; the specific steps for initializing the population using Tent chaotic mapping and reverse learning strategies are as follows: Step 1: Map the sequence of chaotic mappings to individual sparrows using the Tent chaotic mapping sequence and calculate the fitness of the population; Step 2: Calculate the inverse solution population of the population after the Tent chaotic mapping, and calculate the fitness of the population; Step 3: Merge the Tent chaotic map population and the reverse solution population, sort them according to their fitness values, and select the top... One sparrow individual was used as the initial population; In the sparrow search algorithm S3-2, the discoverer plays a guiding role in the population. If the discoverer gets stuck in a local optimum, it can easily cause the entire population to stagnate. Therefore, a Lévy flight perturbation strategy is introduced to update the discoverer's position. Based on the basic Lévy flight strategy, the step size factor is changed to a dynamic value that changes with the number of iterations. The expression for the step size factor is: ； In the formula, γ(i) is the first... Step size in the next iteration This represents the current iteration number. The maximum number of iterations, To adjust the parameters, exp(·) represents an exponential function with base e, and sinh(·) represents a hyperbolic sine function. The improved expression for updating the finder's position using the Levy flight strategy is: ； ； In the formula, This is the current optimal solution; ; and The distribution follows a normal random distribution; Let г(·) be the perturbation constant, and г(·) be the gamma function; Indicates the current iteration number; Represent a A vector of dimension 1 with all elements equal to 1; Indicates the first The first generation Only sparrows in the first The position in the middle, Indicates the first The first generation Only sparrows in the first The position in the dimension; It is obedience Normally distributed random numbers; and These are the warning value and the safety value, respectively. When there are no natural predators nearby, the discoverer can conduct a wider search; when When sparrows discover a predator, the entire population will adjust its search strategy and quickly move to a safe area. The joiner updates its position by following the discoverer, and the formula is: ； In the formula, For the first In the next iteration, the sparrow was at the... The worst position in the dimension; Indicates the first In the next iteration, the discoverer was at the... The optimal position in the dimension; Represent a A dimensional vector, where each element is randomly assigned the value 1 or -1. for The transpose of , and ; The number of sparrows in the population, when When this occurs, it indicates that less adaptable participants need to fly to other areas in search of food; when When this occurs, it indicates that the current participant will choose to forage near the discoverer's optimal location; When an early warning animal senses danger, it will exhibit anti-predation behavior, and its location will be updated as follows: ； In the formula, For the first The global optimal position at the next iteration; It is obedience The step size control parameters; ; This represents the current fitness value of the sparrow. This represents the current globally optimal fitness value. This represents the current worst-case fitness value globally. It is a constant close to 0 to avoid the denominator being 0; S3-3 uses an external archive outside the population to store and update the Pareto optimal solution set with the non-dominated solutions generated in each iteration of the algorithm. Since the external archive has limited capacity, and the population size in the external archive may increase during the update process, to maintain population diversity and prevent the population size from exceeding the limit, a method is used to remove similar individuals based on the crowding distance to maintain the balance of the Pareto solution set. The crowding distance is calculated using the following formula: ； In the formula, For the Pareto solution set, the first The crowded distance of a single sparrow The number of objective functions; and For the first Only the first two adjacent individuals of a sparrow The objective function value, , The first The maximum and minimum values ​​of each objective function.

5. The multi-objective optimization scheduling method for an active distribution network including distributed energy storage according to claim 1, characterized in that, Step S4 specifically involves: Demand response optimization based on time-of-use pricing is performed on the original load to obtain the optimized load and electricity price, and the load response cost is calculated. Then, an improved sparrow search algorithm is used to solve the multi-objective optimization scheduling model. The individuals in the sparrow population and the number of update iterations are set, and they are updated in each subsequent update until the iteration is completed, so as to obtain a non-dominated solution set that satisfies complex constraints.

Images