Power system scheduling method and device based on Pareto optimization, terminal equipment and storage medium

Abstract

The invention discloses an electric power system dispatching method and device based on Pareto optimization, terminal equipment and a storage medium, and belongs to the field of electric power system dispatching.The method comprises the steps that unit power data, unit cost data, node power data and branch power data of a to-be-dispatched system are obtained; constructing a power system scheduling model and corresponding security constraints according to the unit power data, the unit cost data, the node power data and the branch power data; solving the power system dispatching model to obtain a Pareto solution set according to the unit power data, the unit cost data, the node power data and the branch power data under the security constraint by taking the minimum sum of the unit operation cost, the carbon emission and the valley-peak difference as a target; calculating the relative proximity of each Pareto solution according to the Pareto solution set; and scheduling the system to be scheduled by taking the Pareto solution corresponding to the minimum relative proximity as a scheduling strategy parameter. By implementing the method, the multi-target scheduling requirement of the power system is met.

Description

Technical Field

[0001] The present invention relates to the technical field of power system scheduling, and in particular, to a power system scheduling method, device, terminal device and storage medium based on Pareto optimization. Background Art

[0002] In the current global energy structure, the power industry, as a major field of fossil energy consumption, generates a huge amount of carbon dioxide emissions during its operation, which has a great impact on the natural environment. Therefore, the traditional single-objective economic scheduling of power systems should be transformed into multi-objective scheduling. For the multi-objective scheduling problem of power systems, the current single-objective transformation methods cannot evaluate the relationship of mutual conflict and restriction between each objective, and it is difficult to find an effective balance point among multiple conflicting objectives. This leads to difficulty in meeting the requirements of multi-objective scheduling of power systems in practical applications, thus restricting the overall performance of power systems. Summary of the Invention

[0003] Embodiments of the present invention provide a power system scheduling method, device, terminal device and storage medium based on Pareto optimization, which can effectively solve the problem that it is difficult to find an effective balance point among multiple conflicting objectives in the prior art and it is difficult to meet the requirements of multi-objective scheduling of power systems.

[0004] An embodiment of the present invention provides a power system scheduling method based on Pareto optimization, including:

[0005] Obtaining the unit power data, unit cost data, node power data and branch power data of the system to be scheduled;

[0006] Constructing a power system scheduling model and the corresponding security constraints of the power system scheduling model according to the unit power data, the unit cost data, the node power data and the branch power data;

[0007] Taking the sum of the unit operation cost, carbon emissions and valley-peak difference as the objective, and solving the power system scheduling model according to the unit power data, the unit cost data, the node power data and the branch power data under the security constraints to obtain the Pareto solution set of the power system scheduling model;

[0008] Calculating the relative proximity of each Pareto solution in the Pareto solution set according to the Pareto solution set;

[0009] Taking the Pareto solution corresponding to the minimum relative proximity as the scheduling strategy parameter of the system to be scheduled;

[0010] Scheduling the system to be scheduled according to the scheduling strategy parameter.

[0011] Furthermore, the security constraints include: generator output constraints, generator ramp constraints, system reserve demand constraints, new energy output constraints, branch transmission constraints, and system power balance constraints;

[0012] The generator output constraints are:

[0013]

[0014] where i is the generator number of the system to be scheduled; p i,t is the output of generator i at time t; is the downward reserve reserved by generator i to cope with power fluctuations at time t; is the upward reserve reserved by generator i to cope with power fluctuations at time t; P i is the lower bound of the output of generator i; is the upper bound of the output of generator i;

[0015] The generator ramp constraints are:

[0016]

[0017] where p i，t+1 is the output of generator i at time t + 1; is the upward reserve reserved by generator i to cope with power fluctuations at time t + 1; is the downward ramp limit of generator i at time t; is the downward reserve reserved by generator i to cope with power fluctuations at time t + 1; is the upward ramp limit of generator i at time t; is the upper bound of the downward ramp of generator i; is the upper bound of the downward ramp of generator i;

[0018] The system reserve demand constraints are:

[0019]

[0020] where, is the minimum demand that the total reserve provided by the generators should reach; is the maximum demand that the total reserve provided by the generators should reach;

[0021] The new energy output constraints are:

[0022]

[0023] where w j,t is the actual output of new energy j at time t; W j,t is the available output of new energy j at time t;

[0024] The branch transmission constraint is as follows:

[0025]

[0026] where P l is the lower bound of the transmission power of branch l; is the upper bound of the transmission power of branch l; g l,i is the influence coefficient of the power of unit i on the transmission power of branch l; g l,j is the influence coefficient of the power of new energy j on the transmission power of branch l; g l,n is the influence coefficient of the injection power of node n on the transmission power of branch l; D n,t is the load of node n;

[0027] The system power balance constraint is as follows:

[0028]

[0029] Furthermore, with the goal of minimizing the sum of the unit operation cost, carbon emissions, and valley-peak difference, under the above safety constraints, according to the unit power data, the unit cost data, the node power data, and the branch power data, the power system scheduling model is solved to obtain the Pareto solution set of the power system scheduling model, including:

[0030] Respectively, with the goals of minimizing the unit operation cost, minimizing carbon emissions, and minimizing the valley-peak difference, according to the unit power data, the unit cost data, the node power data, and the branch power data, under the above safety constraints, the power system scheduling model is solved to obtain the first objective solution, the second objective solution, and the third objective solution;

[0031] According to the first objective solution, the second objective solution, and the third objective solution, normalization processing is respectively performed to obtain a number of normalized objective vectors; and the normalized objective vectors are used as the Pareto front;

[0032] According to the Pareto front and the preset combination coefficients, the reference point set of the solution of the power system scheduling model is determined;

[0033] According to the reference point set, the first objective solution, the second objective solution, and the third objective solution, parallel axes are generated;

[0034] According to the reference point set and the parallel axes, the solution is performed to make the Pareto solutions corresponding to the parallel axes within the Pareto front boundary fall on the parallel axes, and the Pareto solutions corresponding to the parallel axes outside the Pareto front boundary fall on the boundary, to obtain the initial Pareto solution;

[0035] Project the initial Pareto solution onto the reference plane formed by the set of reference points along the direction parallel to the axis, and perform screening processing to obtain the Pareto solution set of the power system scheduling model.

[0036] Further, according to the Pareto front and the preset combination coefficient, determine the set of reference points of the solution of the power system scheduling model, including:

[0037] Calculate the internal reference point of the solution of the power system scheduling model according to the Pareto front and the preset combination coefficient;

[0038] Calculate the difference vector according to the Pareto front and the preset number of segments;

[0039] Generate the external reference point of the solution of the power system scheduling model according to the internal reference point and the difference vector;

[0040] Merge according to the internal reference point and the external reference point to obtain the set of reference points of the solution of the power system scheduling model.

[0041] Further, project the initial Pareto solution onto the reference plane formed by the set of reference points along the direction parallel to the axis, and perform screening processing to obtain the Pareto solution set of the power system scheduling model, including:

[0042] Generate several reference planes according to the set of reference points;

[0043] Project the initial Pareto solution onto the reference plane to generate several projection points;

[0044] Calculate the projection distance between every two projection points according to the projection points;

[0045] Calculate the reference point neighborhood according to the set of reference points, the difference vector and the preset projection point distance threshold;

[0046] In the reference point neighborhood, retain the initial Pareto solution corresponding to the projection point closest to the reference point to obtain the target Pareto solution;

[0047] Delete the target Pareto solutions corresponding to the projection distances less than the preset projection point distance threshold to obtain the Pareto solution set of the power system scheduling model.

[0048] Further, according to the Pareto solution set, calculate the relative closeness of each Pareto solution in the Pareto solution set, including:

[0049] Calculate the normalization value of each Pareto solution according to the Pareto solution set to obtain the target normalization value;

[0050] Construct an evaluation matrix based on the target normalization value to obtain the target evaluation matrix;

[0051] Take the minimum value of the target normalization value as the positive ideal point and the maximum value of the target normalization value as the negative ideal point;

[0052] Calculate the first weight according to the target evaluation matrix and the number of solutions corresponding to the Pareto solution set;

[0053] Calculate the relative closeness according to the positive ideal point, the negative ideal point, the first weight and the preset weight.

[0054] Further, the scheduling strategy parameters include: unit output, downward reserve reserved by the unit, upward reserve reserved by the unit, and new energy output;

[0055] Schedule the system to be scheduled according to the scheduling strategy parameters, including:

[0056] Adjust the unit output allocation of the system to be scheduled according to the unit output;

[0057] Determine the downward reserve and upward reserve of the units in the system to be scheduled according to the downward reserve reserved by the unit and the upward reserve reserved by the unit;

[0058] Adjust the new energy output allocation of the system to be scheduled according to the new energy output.

[0059] As an improvement of the above solution, another embodiment of the present invention correspondingly provides a power system scheduling device based on Pareto optimization, including:

[0060] A system data acquisition module, configured to acquire unit power data, unit cost data, node power data, and branch power data of the system to be scheduled;

[0061] A model and constraint construction module, configured to construct a power system scheduling model and the corresponding security constraints of the power system scheduling model according to the unit power data, the unit cost data, the node power data, and the branch power data;

[0062] A scheduling model solving module, configured to take the sum of the minimum unit operating cost, carbon emissions, and valley-peak difference as the goal, and solve the power system scheduling model according to the unit power data, the unit cost data, the node power data, and the branch power data under the security constraints to obtain the Pareto solution set of the power system scheduling model;

[0063] A relative closeness calculation module, configured to calculate the relative closeness of each Pareto solution in the Pareto solution set according to the Pareto solution set;

[0064] A scheduling policy parameter determination module, configured to use the Pareto solution corresponding to the minimum relative proximity as the scheduling policy parameter of the system to be scheduled;

[0065] A system scheduling module, configured to schedule the system to be scheduled according to the scheduling policy parameter.

[0066] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, the method for scheduling a power system based on Pareto optimization as described in the above embodiment is implemented.

[0067] Another embodiment of the present invention provides a computer-readable storage medium. The computer-readable storage medium includes a stored computer program. When the computer program runs, the device where the computer-readable storage medium is located is controlled to execute the method for scheduling a power system based on Pareto optimization as described in the above embodiment.

[0068] By implementing the present invention, at least the following beneficial effects are achieved:

[0069] The present invention provides a power system scheduling method, device, terminal device and storage medium based on Pareto optimization. The method can obtain the unit power data, unit cost data, node power data and branch power data of the system to be scheduled; construct a power system scheduling model and the corresponding security constraints of the power system scheduling model according to the unit power data, the unit cost data, the node power data and the branch power data; with the goal of minimizing the sum of unit operating costs, carbon emissions and valley-peak differences, under the security constraints, solve the power system scheduling model according to the unit power data, the unit cost data, the node power data and the branch power data to obtain the Pareto solution set of the power system scheduling model; calculate the relative proximity of each Pareto solution in the Pareto solution set according to the Pareto solution set; use the Pareto solution corresponding to the minimum relative proximity as the scheduling strategy parameter of the system to be scheduled; and schedule the system to be scheduled according to the scheduling strategy parameter. In power system scheduling, three conflicting objectives of cost, carbon emissions and valley-peak differences are optimized simultaneously. By using the unit power data, unit cost data, node power data and branch power data, the power system scheduling model is solved under security constraints to achieve the economic, low-carbon and safe operation of the power system with the goal of minimizing the sum of cost, carbon emissions and valley-peak differences. The Pareto solution corresponding to the minimum relative proximity is used as the scheduling strategy parameter of the system to be scheduled, covering the Pareto solutions among cost, carbon emissions and valley-peak differences. The minimum relative proximity indicates that each objective reaches the optimal value under the mutual constraints of cost, carbon emissions and valley-peak difference objectives, so as to find an effective balance point among multiple objectives, meet the multi-objective scheduling requirements of the power system, and improve the overall performance of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 is a schematic flowchart of a power system scheduling method based on Pareto optimization provided by an embodiment of the present invention;

[0071] Figure 2 is a schematic diagram of the equal division adjacent point method provided by an embodiment of the present invention;

[0072] Figure 3 is a flowchart of the equal division adjacent point method provided by an embodiment of the present invention;

[0073] Figure 4 is another schematic flowchart of the solution process provided by an embodiment of the present invention;

[0074] Figure 5 is a schematic structural diagram of a power system scheduling device based on Pareto optimization provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0075] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0076] See Figure 1 , to solve the problem that it is difficult to find an effective balance among multiple conflicting goals in the prior art and it is difficult to meet the multi-objective scheduling requirements of the power system, a flowchart of a power system scheduling method based on Pareto optimization is provided in an embodiment of the present invention, including:

[0077] S1. Obtain the unit power data, unit cost data, node power data, and branch power data of the system to be scheduled;

[0078] Specifically, the unit power data includes: the lower bound of unit output, the upper bound of unit output, the upward ramp limit of the unit, the downward ramp limit of the unit, the minimum demand for the total unit reserve, the maximum demand for the total unit reserve, and the available output of new energy; the unit cost data includes: the unit power generation cost, the unit reserve cost, the new energy abandonment penalty cost, and carbon emissions; the node power data includes: the influence coefficient of the node on the branch transmission power and the node load; the branch power data includes: the lower bound of the branch transmission power, the upper bound of the branch transmission power, and the influence coefficient of the unit on the branch transmission power.

[0079] S2. Construct a power system scheduling model and the corresponding security constraints of the power system scheduling model according to the unit power data, the unit cost data, the node power data, and the branch power data;

[0080] Specifically, the security constraints include: unit output constraints, unit ramp constraints, system reserve demand constraints, new energy output constraints, branch transmission constraints, and system power balance constraints;

[0081] The unit output constraint is:

[0082]

[0083] where i is the unit serial number of the system to be scheduled; p i,t is the output of unit i at time t; is the downward reserve reserved by unit i at time t to cope with power fluctuations; is the upward reserve reserved by unit i at time t to cope with power fluctuations; P i is the lower bound of the output of unit i; is the upper bound of the output of unit i;

[0084] The ramp constraint of the unit is as follows:

[0085]

[0086] where p i，t+1 is the output of unit i in the (t + 1)th period; is the upward reserve reserved by unit i in the (t + 1)th period to cope with power fluctuations; is the downward ramp limit of unit i in the tth period; is the downward reserve reserved by unit i in the (t + 1)th period to cope with power fluctuations; is the upward ramp limit of unit i in the tth period; is the upper bound of the downward ramp of unit i; is the upper bound of the downward ramp of unit i;

[0087] The system reserve demand constraint is as follows:

[0088]

[0089] where is the minimum demand that the total reserve provided by the units should reach; is the maximum demand that the total reserve provided by the units should reach;

[0090] The new energy output constraint is as follows:

[0091]

[0092] where w j,t is the actual output of new energy j in the tth period; W j,t is the available output of new energy j in the tth period;

[0093] The branch transmission constraint is as follows:

[0094]

[0095] where P l is the lower bound of the transmission power of branch l; is the upper bound of the transmission power of branch l; g l,i is the influence coefficient of the power of unit i on the transmission power of branch l; g l,j is the influence coefficient of the power of new energy j on the transmission power of branch l; g l,n is the influence coefficient of the injection power of node n on the transmission power of branch l; D n,tis the load of node n; in a system with a fixed topology, the influence coefficient of the injection power of node n on the transmission power of branch l is fixed, that is, for every unit of power injected by node n, the transmission power of branch l increases by the same amount.

[0096] The system power balance constraint is as follows:

[0097]

[0098] Specifically, the generator output constraint limits the upper and lower limits of the output of each generator; the generator ramping constraint means that the reserve provided by the generator cannot exceed the ramping limit; the system reserve constraint means that the sum of the reserves provided by all generators should reach the minimum demand and the maximum demand The new energy output constraint means that the actual output w of the new energy j,t cannot exceed the available output W j,t ; the branch transmission constraint means the restrictions on branch transmission by generators and nodes; the system power balance constraint means that the power generation capacity must be able to meet the load demand, and the power generation power must be equal to the load demand power.

[0099] S3. With the goal of minimizing the sum of the unit operating cost, carbon emissions, and valley - peak difference, under the above - mentioned security constraints, solve the power system scheduling model according to the unit power data, the unit cost data, the node power data, and the branch power data, and obtain the Pareto solution set of the power system scheduling model;

[0100] Specifically, with the goal of minimizing the sum of the unit operating cost, carbon emissions, and valley - peak difference, under the above - mentioned security constraints, solve the power system scheduling model according to the unit power data, the unit cost data, the node power data, and the branch power data, and obtain the Pareto solution set of the power system scheduling model, including:

[0101] Solve the power system scheduling model with the goals of minimizing the unit operating cost, minimizing carbon emissions, and minimizing the valley - peak difference respectively, according to the unit power data, the unit cost data, the node power data, and the branch power data, under the above - mentioned security constraints, to obtain the first - objective solution, the second - objective solution, and the third - objective solution;

[0102] Perform normalization processing on the first - objective solution, the second - objective solution, and the third - objective solution respectively to obtain a number of normalized objective vectors; and use the normalized objective vectors as the Pareto front;

[0103] Determine the reference point set of the solutions of the power system scheduling model according to the Pareto front and the preset combination coefficients;

[0104] Generate parallel axes based on the reference point set, the first objective solution, the second objective solution, and the third objective solution;

[0105] Solve according to the reference point set and the parallel axes, so that the Pareto solutions corresponding to the parallel axes within the Pareto front boundary fall on the parallel axes, and the Pareto solutions corresponding to the parallel axes outside the Pareto front boundary fall on the boundary, to obtain the initial Pareto solution;

[0106] Project the initial Pareto solution onto the reference plane formed by the reference point set along the direction of the parallel axes, and perform screening processing to obtain the Pareto solution set of the power system scheduling model.

[0107] Preferably, the unit operating cost f1: is the unit power generation cost, a i , b i , c i are the power generation cost coefficients of unit i respectively, is the reserve cost, and are the downward reserve cost coefficient and upward reserve cost coefficient of unit i respectively, ∑ t ∑ j ρ j (W j，t -w j，t ) is the penalty cost for new energy abandonment, ρ j is the penalty coefficient of new energy j. Carbon emissions f2: f2 = ∑ t ∑ i e i p i,t , e i is the carbon emissions generated by unit i per unit output. Valley - peak difference f3: f3 = ∑ t (ΔP t ) 2 , ΔP t = ∑ i p i,t - ∑ i p i,t-1 , ΔP t is the change in the total output of the unit in adjacent time periods, that is, the total output of unit i in time period t ∑ i p i，t and the total output of unit i in time period t - 1 ∑ i p i，t-1 The multi - objective optimization based on the Pareto theory is defined as follows:

[0108] min F = {f1, f2,..., f m ,..., f M}

[0109] such that \(G(x)\leq0\)

[0110] where \(F\) is the objective set, i.e., with the objective of minimizing the sum of unit operating costs, carbon emissions, and valley - peak differences; \(f\) m is a sub - objective, and there can be multiple sub - objectives. The sub - objectives in this embodiment are unit operating costs, carbon emissions, and valley - peak differences; \(M\) is the number of objectives, which is 3 in this embodiment; \(x\) is the decision variable vector, i.e., the scheduling strategy parameters of the system to be scheduled, including the output \(p\) of each unit i,t , the reserved downward reserve and upward reserve as well as the output \(w\) of each new energy source j,t ; \(G\) is the constraint condition, i.e., the security constraint of the power system scheduling model in this embodiment. There is a dominance relationship among the solutions in Pareto optimization:

[0111]

[0112] \(i,j\) represent the \(i\) - th and \(j\) - th solutions; \(m,n\) represent the \(m\) - th and \(n\) - th objective function values; \(f\) i,m represents the \(i\) - th solution of the \(m\) - th objective function value; \(f\) j,m represents the \(j\) - th solution of the \(m\) - th objective function value; \(f\) i,n represents the \(i\) - th solution of the \(n\) - th objective function value; \(f\) j,n represents the \(j\) - th solution of the \(n\) - th objective function value. If the above conditions are met, then the \(i\) - th solution dominates the \(j\) - th solution. The solution that is not dominated by other solutions is called the Pareto solution (or non - dominated solution). All solutions within the feasible region form the Pareto solution set, and the corresponding objective vectors form the Pareto front. Since the Pareto front is continuously distributed, it is impossible to obtain all Pareto solutions. In this case, the Pareto solution set is used to approximately represent all Pareto solutions. To ensure the approximation degree, the Pareto front corresponding to the Pareto solution set should not only be non - dominated, but also be evenly and widely distributed. For a multi - objective problem, it is impossible to find a solution that minimizes all objective values (the smaller the value, the better), so there are many solutions. Solution \(i\) dominates solution \(j\), which means that all objective values of solution \(i\) are less than or equal to those of solution \(j\), and there is at least one objective value less than that of solution \(j\), that is, it represents that the result of solution \(i\) is better than that of solution \(j\). The solution is the value of the decision variable \(x\), and the Pareto solution set is actually evaluated according to the objective function values corresponding to the solutions, obtaining a set without superiority or inferiority.

[0113] In a preferred embodiment of the present invention, an equal - division adjacent - point method is proposed to solve the Pareto solution set of the power system scheduling model. First, with each objective minimized, the power system scheduling model is solved to obtain an initial solution, that is, each objective is optimized separately to obtain M extreme solutions, namely the first - objective solution, the second - objective solution, and the third - objective solution corresponding to each objective; then, normalization processing is performed on the first - objective solution, the second - objective solution, and the third - objective solution respectively to obtain a number of normalized objective vectors And the normalized objective vectors are used as the Pareto front; according to the Pareto front and a preset combination coefficient, a reference - point set of the solutions of the power system scheduling model is determined, is the i - th reference point, c i,m is the preset combination coefficient, Ω R is the reference - point set. The reference points are sequentially taken from the set [-S E ,…,0,1 / D,…,1 + S E / D] (where D is the preset number of segments, and S E is the expansion factor) to ensure that the reference points are evenly distributed and cover the entire Pareto front. Then, according to the reference - point set, the first - objective solution, the second - objective solution, and the third - objective solution, parallel axes are generated, and the direction vector e of the parallel axes is also calculated from these endpoints: Solve according to the reference - point set and the parallel axes, so that the Pareto solutions corresponding to the parallel axes within the Pareto - front boundary fall on the parallel axes, and the Pareto solutions corresponding to the parallel axes outside the Pareto - front boundary fall on the boundary, to obtain the initial Pareto solutions.

[0114] Two sequential optimization problems are used to obtain the initial Pareto solution corresponding to each reference point:

[0115]

[0116] In the formula, is the optimal value obtained by optimizing maxλ i , ; r i,m represents the objective - normalization value of objective m of the i - th Pareto solution; R i represents the i - th Pareto solution. The above two sequential optimization problems ensure that the Pareto solutions corresponding to the parallel axes within the Pareto - front boundary exactly fall on the axes, while the Pareto solutions corresponding to the parallel axes outside the Pareto - front boundary will fall on the boundary. Then, the initial Pareto solutions are projected onto the reference plane formed by the reference - point set along the direction of the parallel axes, and screening processing is performed to obtain the Pareto solution set of the power system scheduling model. The reference points, Pareto solutions, parallel axes, and endpoints of the Pareto front are as Figure 2As shown, the Pareto solution set obtained by integrating the operating cost, carbon emissions, and valley-peak difference target of the combined unit As Figure 3 Shown in the flowchart, by solving the reference point set to map the Pareto solution set, and then using sequential optimization to solve the Pareto solution corresponding to each reference point and eliminating the redundant Pareto solutions, the Pareto solution set of the power system scheduling model is obtained. By optimizing the operating cost and carbon emissions of the unit, it helps to reduce the overall cost and environmental impact of the power system. At the same time, by considering the minimum valley-peak difference, it helps to balance the load of the power system and improve the stability and efficiency of the power system.

[0117] Specifically, according to the Pareto front and the preset combination coefficient, determine the reference point set of the solution of the power system scheduling model, including:

[0118] According to the Pareto front and the preset combination coefficient, calculate the internal reference point of the solution of the power system scheduling model;

[0119] According to the Pareto front and the preset number of segments, calculate the difference vector;

[0120] According to the internal reference point and the difference vector, generate the external reference point of the solution of the power system scheduling model;

[0121] Merge according to the internal reference point and the external reference point to obtain the reference point set of the solution of the power system scheduling model.

[0122] In a preferred embodiment of the present invention, the equal division of adjacent points method constructs a maximum possible reference point set composed of internal and external reference points. The internal reference point is obtained by the formula And needs to meet the following additional conditions: Is the normalized value of the i-th internal reference point in the m-th target dimension; Is the normalized value of the i-th internal reference point in the n-th target dimension; e m Is the m-th target reference value; e n Is the n-th target reference value.

[0123] According to the Pareto front and the preset number of segments, calculate the difference vector. Taking the three-objective optimization problem as an example, select two shortest difference vectors d1 and d2 from Then the generation of the external reference point is:

[0124] Is the first external reference point, Is the second external reference point, It is an internal reference point. Based on the internal reference point and the external reference point, they are merged, and the overlapping points are removed to obtain the set of reference points for the solution of the power system scheduling model.

[0125] Specifically, project the initial Pareto solution onto the reference plane formed by the set of reference points along the parallel axis direction, and perform a screening process to obtain the Pareto solution set of the power system scheduling model, including:

[0126] Generate a number of reference planes according to the set of reference points;

[0127] Project the initial Pareto solution onto the reference plane to generate a number of projection points;

[0128] Calculate the projection distance between every two projection points according to the projection points;

[0129] Calculate the reference point neighborhood according to the set of reference points, the difference vector, and the preset projection point distance threshold;

[0130] In the reference point neighborhood, retain the initial Pareto solution corresponding to the projection point closest to the reference point to obtain the target Pareto solution;

[0131] Delete the target Pareto solutions corresponding to the projection distances less than the preset projection point distance threshold to obtain the Pareto solution set of the power system scheduling model.

[0132] In a preferred embodiment of the present invention, first, project the boundary Pareto solution (initial Pareto solution) onto the reference plane along the direction of the parallel axis to generate a number of projection points. The projection point is represented as According to the projection point and the reference point distribution, delete the redundant boundary Pareto solutions. To ensure the uniformity of the Pareto solution set, only one Pareto solution is retained near each reference point , that is, the Pareto solution corresponding to the closest projection point in its neighborhood is retained to obtain the target Pareto solution. The reference point neighborhood of each reference point is defined as:

[0133]

[0134] |c1| ≤ 0.5, |c2| ≤ 0.5;

[0135] Secondly, the boundary Pareto solutions that are relatively close to each other are deleted to further ensure uniformity. If the projection distance between two projection points that are close to each other is less than the preset projection point distance threshold, then one of the two target Pareto solutions associated with these two projection points will be deleted, and the Pareto solution set of the power system scheduling model is obtained. The preset projection point distance threshold is set as: |c1| < 0.25, |c2| < 0.25.

[0136] S4. Calculate the relative closeness of each Pareto solution in the Pareto solution set according to the Pareto solution set;

[0137] Preferably, calculating the relative closeness of each Pareto solution in the Pareto solution set according to the Pareto solution set includes:

[0138] Calculate the normalization value of each Pareto solution according to the Pareto solution set to obtain the target normalization value;

[0139] Construct an evaluation matrix according to the target normalization value to obtain the target evaluation matrix;

[0140] Take the minimum value of the target normalization value as the positive ideal point and the maximum value of the target normalization value as the negative ideal point;

[0141] Calculate the first weight according to the target evaluation matrix and the number of solutions corresponding to the Pareto solution set;

[0142] Calculate the relative closeness according to the positive ideal point, the negative ideal point, the first weight and the preset weight.

[0143] In a preferred embodiment of the present invention, calculate the normalization value of each Pareto solution according to the Pareto solution set to obtain the target normalization value, and then construct an evaluation matrix according to the target normalization value to obtain the target evaluation matrix. The element r in the evaluation matrix i,m represents the target normalization value of the target m of the i-th Pareto solution. Take the minimum value of the target normalization value as the positive ideal point and the maximum value of the target normalization value as the negative ideal point. The positive and negative ideal points are respectively and Calculate the first weight according to the target evaluation matrix and the number of solutions corresponding to the Pareto solution set. The first weight represents the objective weight, and the objective weight is determined by the entropy weight σ m The entropy weight represents the amount of information provided by each target, and its calculation formula is as follows:

[0144] N PF is the number of solutions corresponding to the Pareto solution set; ρ i，mis the normalized value of the i-th Pareto solution on the m-th index, obtained by normalizing the original normalized value r i，m through normalization. The preset weight is the subjective weight and can be set customarily. Then, the total comprehensive weight is composed of the first weight and the preset weight. Finally, calculations are performed according to the positive ideal point, the negative ideal point, the first weight, and the preset weight to obtain the relative closeness. The relative closeness of the i-th Pareto solution can be calculated as follows:

[0145]

[0146] In the formula, the smaller T i represents a better solution. Then, the Pareto solution corresponding to the minimum T i is the optimal compromise solution, that is, the scheduling strategy parameter of the system to be scheduled.

[0147] S5. Use the Pareto solution corresponding to the minimum relative closeness as the scheduling strategy parameter of the system to be scheduled;

[0148] In a preferred embodiment of the present invention, the evenly divided neighboring point method is used to obtain the Pareto solution set, providing diversified scheduling decision options for dispatchers. In practical applications, an optimal compromise solution should be selected from the Pareto solution set as the final scheduling decision. The entropy weight ideal point method is used to select the optimal compromise solution, and the Pareto solution corresponding to the minimum relative closeness is used as the scheduling strategy parameter of the system to be scheduled. As Figure 4 shown, by establishing a multi-objective low-carbon scheduling model (power system scheduling model) of the power system, then solving the Pareto solution set of the model, and selecting the best compromise solution from the Pareto solution set according to the entropy weight ideal point method as the scheduling strategy parameter of the system to be scheduled, the proposed evenly divided neighboring point method can solve a uniformly distributed and extensive Pareto solution set, thereby providing more diversified decision options for decision-makers. The proposed entropy weight ideal point method fully considers the actual scheduling requirements and explores the information contained in the distribution of the Pareto solution set, thereby realizing a more scientific compromise solution selection.

[0149] S6. Schedule the system to be scheduled according to the scheduling strategy parameter.

[0150] Specifically, the scheduling strategy parameter includes: unit output, downward reserve reserved by the unit, upward reserve reserved by the unit, and new energy output;

[0151] Scheduling the system to be scheduled according to the scheduling strategy parameter includes:

[0152] Adjusting the unit output distribution of the system to be scheduled according to the unit output;

[0153] Determine the downward reserve and upward reserve of the units in the system to be dispatched according to the downward reserve and upward reserve reserved for the units.

[0154] Adjust the new energy output distribution of the system to be dispatched according to the new energy output.

[0155] By implementing this embodiment, obtain the unit power data, unit cost data, node power data, and branch power data of the system to be dispatched; according to the unit power data, the unit cost data, the node power data, and the branch power data, construct a power system dispatch model and the corresponding security constraints of the power system dispatch model; with the goal of minimizing the sum of unit operating costs, carbon emissions, and valley-peak differences, under the security constraints, solve the power system dispatch model according to the unit power data, the unit cost data, the node power data, and the branch power data to obtain the Pareto solution set of the power system dispatch model; according to the Pareto solution set, calculate the relative proximity of each Pareto solution in the Pareto solution set; take the Pareto solution corresponding to the minimum relative proximity as the dispatch strategy parameter of the system to be dispatched; dispatch the system to be dispatched according to the dispatch strategy parameter. In the power system dispatch, simultaneously optimize three conflicting objectives of cost, carbon emissions, and valley-peak differences. Through the unit power data, unit cost data, node power data, and branch power data, solve the power system dispatch model under the security constraints to achieve the economic, low-carbon, and safe operation of the power system with the goal of minimizing the sum of cost, carbon emissions, and valley-peak differences, find an effective balance point among multiple objectives, that is, the Pareto solution corresponding to the minimum relative proximity, meet the multi-objective dispatch requirements of the power system, and improve the overall performance of the power system.

[0156] See Figure 5 , which is a schematic structural diagram of a power system dispatch device based on Pareto optimization provided by an embodiment of the present invention, includes:

[0157] A system data acquisition module, configured to acquire the unit power data, unit cost data, node power data, and branch power data of the system to be dispatched;

[0158] A model and constraint construction module, configured to construct a power system dispatch model and the corresponding security constraints of the power system dispatch model according to the unit power data, the unit cost data, the node power data, and the branch power data;

[0159] A scheduling model solving module, which is used to take the minimum of the sum of the unit operating cost, carbon emissions, and valley-peak difference as the objective, and under the said security constraints, solve the power system scheduling model according to the unit power data, the unit cost data, the node power data, and the branch power data, so as to obtain the Pareto solution set of the power system scheduling model;

[0160] A relative proximity calculation module, which is used to calculate the relative proximity of each Pareto solution in the Pareto solution set according to the Pareto solution set;

[0161] A scheduling strategy parameter determination module, which is used to take the Pareto solution corresponding to the minimum relative proximity as the scheduling strategy parameter of the system to be scheduled;

[0162] A system scheduling module, which is used to schedule the system to be scheduled according to the scheduling strategy parameter.

[0163] The present invention provides a power system scheduling device based on Pareto optimization. According to the system data acquisition module, the unit power data, the unit cost data, the node power data, and the branch power data of the system to be scheduled are acquired; in the model and constraint construction module, according to the unit power data, the unit cost data, the node power data, and the branch power data, a power system scheduling model and the security constraints corresponding to the power system scheduling model are constructed; through the scheduling model solving module, taking the minimum of the sum of the unit operating cost, carbon emissions, and valley-peak difference as the objective, under the said security constraints, solve the power system scheduling model according to the unit power data, the unit cost data, the node power data, and the branch power data, so as to obtain the Pareto solution set of the power system scheduling model; then in the relative proximity calculation module, calculate the relative proximity of each Pareto solution in the Pareto solution set according to the Pareto solution set; in the scheduling strategy parameter determination module, take the Pareto solution corresponding to the minimum relative proximity as the scheduling strategy parameter of the system to be scheduled; finally in the system scheduling module, schedule the system to be scheduled according to the scheduling strategy parameter. When scheduling the power system, three conflicting objectives of cost, carbon emissions, and valley-peak difference are optimized simultaneously. By using the unit power data, the unit cost data, the node power data, and the branch power data, under the security constraints, the power system scheduling model is solved, so as to achieve the economic, low-carbon, and secure operation of the power system with the objective of minimizing the sum of cost, carbon emissions, and valley-peak difference, find an effective balance point among multiple objectives, that is, the Pareto solution corresponding to the minimum relative proximity, meet the multi-objective scheduling requirements of the power system, and improve the overall performance of the power system.

[0164] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the attached drawings of the device embodiments provided by the present invention, the connection relationships between the modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art can understand and implement it without creative efforts.

[0165] Those skilled in the art can clearly understand that for the sake of convenience and brevity, the specific working process of the device described above can refer to the corresponding process in the foregoing method embodiment, and will not be elaborated herein.

[0166] Another embodiment of the present invention also provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a power system scheduling method based on Pareto optimization as described in the above embodiment. The terminal device can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The terminal device may include, but is not limited to, a processor and a memory.

[0167] The processor may be a central processing unit (CPU), or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The processor is the control center of the terminal device, and connects various parts of the entire terminal device through various interfaces and lines.

[0168] The memory can be used to store the computer program. By running or executing the computer program stored in the memory and invoking the data stored in the memory, the processor realizes various functions of the terminal device. The memory mainly includes a program storage area and a data storage area. Among them, the program storage area can store the operating system, application programs required for at least one function, etc.; the data storage area can store the data created according to the use of the mobile phone, etc. In addition, the memory can include high-speed random access memory, and can also include non-volatile memory, such as hard disks, memory, plug-in hard disks, smart media cards (SMC), secure digital (SD) cards, flash cards, at least one magnetic disk storage device, flash device or other volatile solid-state storage devices.

[0169] Another embodiment of the present invention provides a computer-readable storage medium. The computer-readable storage medium includes a stored computer program. When the computer program runs, it controls the device where the computer-readable storage medium is located to execute a power system scheduling method based on Pareto optimization described in the above embodiment.

[0170] The storage medium is a computer-readable storage medium, and the computer program is stored in the computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of the above method embodiments. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0171] The above is the preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements are also regarded as the protection scope of the present invention.

Claims

1. A power system scheduling method based on Pareto optimization, characterized in that, Including: Obtaining the unit power data, unit cost data, node power data, and branch power data of the system to be scheduled; Constructing a power system scheduling model and the corresponding security constraints of the power system scheduling model according to the unit power data, the unit cost data, the node power data, and the branch power data; Taking the minimum sum of the unit operating cost, carbon emissions, and valley-peak difference as the objective, under the security constraints, solving the power system scheduling model according to the unit power data, the unit cost data, the node power data, and the branch power data to obtain the Pareto solution set of the power system scheduling model; Calculating the relative proximity of each Pareto solution in the Pareto solution set according to the Pareto solution set; Taking the Pareto solution corresponding to the minimum relative proximity as the scheduling strategy parameter of the system to be scheduled; Scheduling the system to be scheduled according to the scheduling strategy parameter.

2. The power system scheduling method based on Pareto optimization according to claim 1, wherein The security constraints include: unit output constraint, unit ramp constraint, system reserve demand constraint, new energy output constraint, branch transmission constraint, and system power balance constraint; The unit output constraint is: where i is the unit number of the system to be scheduled; p i,t is the output of unit i at time t; is the downward reserve reserved by unit i to cope with power fluctuations at time t; is the upward reserve reserved by unit i to cope with power fluctuations at time t; P i is the lower bound of the output of unit i; is the upper bound of the output of unit i; The unit ramp constraint is: where p i，t+1 is the output of unit i in the (t + 1)th period; is the upward reserve reserved by unit i for power fluctuation in the (t + 1)th period; is the downward ramp rate limit of unit i in the tth period; is the downward reserve reserved by unit i for power fluctuation in the (t + 1)th period; is the upward ramp rate limit of unit i in the tth period; is the upper bound of the downward ramp rate of unit i; is the upper bound of the downward ramp rate of unit i; The system reserve demand constraint is: Among them, the minimum demand that the total standby provided for the unit should reach; the maximum demand that the total standby provided for the unit should reach; The new energy output constraint is: where, w j,t is the actual output of new energy j at time period t; W j,t is the available output of new energy j at time period t; The branch transmission constraint is: Among them, P l is the lower bound of the transmission power of branch l; is the upper bound of the transmission power of branch l; g l,i is the influence coefficient of the power of unit i on the transmission power of branch l; g l,j is the influence coefficient of the power of new energy j on the transmission power of branch l; g l,n is the influence coefficient of the injection power of node n on the transmission power of branch l; D n,t is the load of node n; The system power balance constraint is:

3. The method for scheduling a power system based on Pareto optimization according to claim 1, characterized in that Taking the minimum sum of the unit operating cost, carbon emissions, and valley-peak difference as the objective, under the security constraints, solving the power system scheduling model according to the unit power data, the unit cost data, the node power data, and the branch power data to obtain the Pareto solution set of the power system scheduling model, including: Taking the minimum unit operating cost, minimum carbon emissions, and minimum valley-peak difference as the objectives respectively, and solving the power system scheduling model according to the unit power data, the unit cost data, the node power data, and the branch power data under the security constraints to obtain the first objective solution, the second objective solution, and the third objective solution; Performing normalization processing on the first objective solution, the second objective solution, and the third objective solution respectively to obtain a number of normalized objective vectors; and taking the normalized objective vectors as the Pareto front; Determining a reference point set of the solutions of the power system scheduling model according to the Pareto front and a preset combination coefficient; Generating parallel axes according to the reference point set, the first objective solution, the second objective solution, and the third objective solution; Solving according to the reference point set and the parallel axes, so that the Pareto solutions corresponding to the parallel axes within the Pareto front boundary fall on the parallel axes, and the Pareto solutions corresponding to the parallel axes outside the Pareto front boundary fall on the boundary, to obtain the initial Pareto solutions; Projecting the initial Pareto solutions onto the reference plane formed by the reference point set along the direction of the parallel axes, and performing screening processing to obtain the Pareto solution set of the power system scheduling model.

4. The method for scheduling a power system based on Pareto optimization according to claim 3, characterized in that, Determining a reference point set of the solutions of the power system scheduling model according to the Pareto front and a preset combination coefficient, including: Calculate an internal reference point of the solution of the power system scheduling model according to the Pareto front and a preset combination coefficient; Calculate a difference vector according to the Pareto front and a preset number of segments; Generate an external reference point of the solution of the power system scheduling model according to the internal reference point and the difference vector; Merge according to the internal reference point and the external reference point to obtain a reference point set of the solution of the power system scheduling model.

5. The power system scheduling method based on Pareto optimization according to claim 4, characterized in that Project the initial Pareto solution onto a reference plane formed by the reference point set along the parallel axis direction and perform a screening process to obtain the Pareto solution set of the power system scheduling model, including: Generate a number of reference planes according to the reference point set; Project the initial Pareto solution onto the reference plane to generate a number of projection points; Calculate the projection distance between every two projection points according to the projection points; Calculate a reference point neighborhood according to the reference point set, the difference vector, and a preset projection point distance threshold; Retain the initial Pareto solution corresponding to the projection point closest to the reference point within the reference point neighborhood to obtain a target Pareto solution; Delete the target Pareto solutions corresponding to projection distances less than the preset projection point distance threshold to obtain the Pareto solution set of the power system scheduling model.

6. The power system scheduling method based on Pareto optimization according to claim 1, characterized in that, Calculate the relative closeness of each Pareto solution in the Pareto solution set according to the Pareto solution set, including: Calculate the normalization value of each Pareto solution according to the Pareto solution set to obtain a target normalization value; Construct an evaluation matrix according to the target normalization value to obtain a target evaluation matrix; Take the minimum value of the target normalization value as the positive ideal point and the maximum value of the target normalization value as the negative ideal point; Calculate a first weight according to the target evaluation matrix and the number of solutions corresponding to the Pareto solution set; Calculate the relative closeness according to the positive ideal point, the negative ideal point, the first weight, and a preset weight.

7. A power system scheduling method based on Pareto optimization according to claim 1, characterized in that The scheduling strategy parameters include: unit output, unit reserved downward reserve, unit reserved upward reserve, and new energy output; Schedule the system to be scheduled according to the scheduling strategy parameters, including: Adjust the unit output allocation of the system to be scheduled according to the unit output; Determine the unit downward reserve and unit upward reserve of the system to be scheduled according to the unit reserved downward reserve and the unit reserved upward reserve; Adjust the new energy output allocation of the system to be scheduled according to the new energy output.

8. A power system scheduling device based on Pareto optimization, characterized in that, Including: A system data acquisition module for acquiring unit power data, unit cost data, node power data, and branch power data of the system to be scheduled; A model and constraint construction module for constructing a power system scheduling model and the safety constraints corresponding to the power system scheduling model according to the unit power data, the unit cost data, the node power data, and the branch power data; A dispatching model solving module, which is used to take the minimum of the sum of the unit operation cost, carbon emissions and valley-peak difference as the objective, and under the said security constraints, solve the power system dispatching model according to the unit power data, the unit cost data, the node power data and the branch power data, so as to obtain the Pareto solution set of the power system dispatching model; A relative proximity calculation module, which is used to calculate the relative proximity of each Pareto solution in the Pareto solution set according to the Pareto solution set; A dispatching strategy parameter determination module, which is used to take the Pareto solution corresponding to the minimum relative proximity as the dispatching strategy parameter of the system to be dispatched; A system dispatching module, which is used to dispatch the system to be dispatched according to the dispatching strategy parameter; 9. A terminal device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements a power system dispatching method based on Pareto optimization as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein when the computer program runs, it controls the device where the computer-readable storage medium is located to execute a power system dispatching method based on Pareto optimization as described in any one of claims 1 to 7.

Images