Regional distributed energy storage optimization scheduling method

Abstract

The invention discloses a regional distributed energy storage optimization scheduling method, which comprises the following steps: S1, model construction: constructing a mixed integer nonlinear programming model containing a multi-objective optimization function according to physical characteristics, operation constraints and economic objectives of a regional power distribution network, a micro-grid, an energy storage system and renewable energy; s2, scene generation and processing: adopting a Latin hypercube sampling method to generate a multi-scene data set of renewable energy output, load demand and electricity price fluctuation, performing clustering analysis on scene data, and screening out representative typical scenes so as to reduce calculation complexity; and S3, carrying out optimization solution. In the application, multi-dimensional constraints such as energy storage life attenuation, network loss, carbon emission and the like are added in constraint conditions, the practicability of the model is improved, and the limitation of traditional single-target scheduling is broken through through multi-target optimization and multi-constraint coordination.

Description

Technical Field

[0001] The present invention relates to the technical field of plastic bottles, and in particular to a method for optimizing and scheduling regional distributed energy storage. Background Art

[0002] Due to natural constraints, distributed energy generation is highly random and uncertain. When a large number of distributed energy sources are connected to the power grid, it can significantly impact the grid's stability and economic operation. Therefore, in the process of optimizing the dispatch of distributed power sources, fully considering the uncertainty of distributed power sources and improving their absorption capacity have become one of the main topics of current distributed power source research.

[0003] Chinese patent publication number CN109494794B, publication date 20200503, discloses a regional distributed energy storage optimization scheduling method and device based on multi-source day-ahead prediction uncertainty. It fully considers the uncertainty of day-ahead prediction of distributed power sources and loads such as photovoltaic and wind power, and proposes a judgment rule for multi-source day-ahead prediction uncertainty; considering the impact of multi-source prediction uncertainty on the power quality of the distribution network, multi-source prediction uncertainty is used as a boundary condition of the regional distributed energy storage optimization scheduling method to ensure that the regional power quality changes are within a controllable range; based on the above-mentioned evaluation conclusions, a regional distributed energy storage economic scheduling method is proposed, and a penalty mechanism for multi-source prediction uncertainty is constructed to reduce the impact of prediction uncertainty on the economic operation of distributed energy storage, thereby optimizing the overall economic efficiency of the regional distributed energy storage system.

[0004] Existing regional distributed energy storage optimization scheduling methods such as the above are mostly optimized with a single objective, ignoring the comprehensive balance of system reliability, energy storage life and environmental benefits, resulting in the problem of "losing sight of one thing while focusing on another" in actual application of scheduling strategies. Therefore, it is urgent to propose corresponding regional distributed energy storage optimization scheduling methods to solve the above problems. Summary of the Invention

[0005] The purpose of the present invention is to propose a regional distributed energy storage optimization scheduling method in order to solve the above problems.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] The regional distributed energy storage optimization scheduling method includes the following steps:.

[0008] S1. Model construction:

[0009] Based on the physical characteristics, operational constraints, and economic objectives of the regional distribution network, microgrid, energy storage system, and renewable energy, a mixed integer nonlinear programming model with multi-objective optimization functions is constructed.

[0010] S2. Scene generation and processing:

[0011] Use the Latin hypercube sampling method to generate a multi-scenario dataset of renewable energy output, load demand, and electricity price fluctuations, perform cluster analysis on the scenario data, and screen out representative typical scenarios to reduce the computational complexity;

[0012] S3. Optimization and solution, including the following stages:

[0013] Introduce the particle swarm optimization algorithm, combined with the dynamic weight adjustment strategy, to solve the multi-objective optimization model;

[0014] Set the particle position as the energy storage charge and discharge power, the microgrid operation mode, and the power purchase and sale decision variables. The particle velocity is adjusted by the inertia weight and the constraint penalty term;

[0015] Embed the constraint conditions into the objective function through the penalty function method to ensure the feasibility of the solution;

[0016] Perform sensitivity analysis on the solution results to verify the robustness of the model to key parameters;

[0017] S4. Strategy update and verification, including the following stages:

[0018] Adjust the energy storage system operation strategy, the microgrid grid connection mode, and the regional power trading plan according to the optimization results;

[0019] Through the rolling horizon optimization method, update the optimization model in real time to cope with dynamic environmental changes;

[0020] Use the simulation tool to verify the stability and economy of the scheduling strategy under transient conditions.

[0021] Preferably, the multi-objective optimization function in the step S1 includes:

[0022] Economic objective:

[0023] min(C total = C pu + C st + C lo + C em ), where C pu is the power purchase cost, C st is the energy storage investment and operation and maintenance cost, C lo is the network loss cost; C em is the carbon emission cost;

[0024] Reliability objective:

[0025] Where R sy is the system power supply reliability, P su is the reagent power supply power in the i-th period, P lois the load demand in the i-th period;

[0026] Environmental goal:

[0027] Among them, E em is the total carbon emission, P gr,i is the main network power purchase, ∈ gr is the main network carbon emission factor, P dg,i is the distributed power generation, ∈ dg is the distributed power source carbon emission factor.

[0028] Preferably, the mixed-integer nonlinear programming model includes the following constraint conditions:

[0029] Power balance constraint: The total regional load demand is equal to the sum of regional power generation, energy storage charge and discharge, main network power purchase and sale, and renewable energy output;

[0030] Energy storage system constraint: including energy storage charge and discharge power limit, state of charge upper and lower limits, charge and discharge efficiency, and life decay model;

[0031] Renewable energy output uncertainty constraint: Based on the scenario generation method, quantify the output fluctuations of wind power, photovoltaic power and renewable energy;

[0032] Network constraint: including line power flow constraint, node voltage constraint and network loss limit;

[0033] Economic constraint: including power purchase price, power sale price, energy storage investment cost, operation and maintenance cost and carbon emission cost;

[0034] Multi-objective optimization goal: Minimize the total regional cost, maximize the system reliability and minimize the carbon emission as the optimization goals, and complete multi-objective coordination through the weighted coefficient method.

[0035] Preferably, the energy storage system constraint includes:

[0036] Charge and discharge power constraint:

[0037] P disc,min ≤P disc,i ≤P disc,max ,P c,min ≤P c,i ≤P c,max ;

[0038] State of charge constraint:

[0039] Among them, η c is the charging efficiency, η d is the discharging efficiency, Δt is the time step;

[0040] Life decay constraint:

[0041] Among them, L cycle is the allowable charge and discharge cycle times of the energy storage system over its entire life cycle.

[0042] Preferably, the step S2 further includes:

[0043] S21. Modeling the output of renewable energy. The output of wind power adopts the Weibull distribution, and the output of photovoltaic power adopts the normal distribution. Combining historical meteorological data to generate random scenarios;

[0044] S22. Modeling the load demand, adopting the elastic load model under time-of-use electricity prices;

[0045] S23. Adopting the real-time electricity price model and combining the predicted data of the day-ahead market to generate electricity price scenarios.

[0046] Preferably, the particle swarm optimization algorithm includes the following steps:

[0047] Initializing parameters: setting the particle swarm size N, the maximum number of iterations T max , the inertia weight ω, the acceleration coefficients c1 and c2;

[0048] Fitness function design: f(x i ) = ω1·C total +ω2·(1 - R sys )+ω3·E em , where ω1, ω2, and ω3 are weight coefficients, satisfying ω1 + ω2 + ω3 = 1;

[0049] Constraint handling strategy:

[0050] For particles that violate the constraints, adopt the dynamic penalty function method:

[0051] Among them, ρ is the penalty factor, and is the violation amount of the g k (x i )th constraint;

[0052] Convergence determination: When the change rate of the fitness function is less than the threshold ∈ in M consecutive iterations, terminate the algorithm.

[0053] Preferably, the rolling horizon optimization method includes the following steps:

[0054] Divide the optimization period into N sub-periods, and perform an optimization solution once within each sub-period;

[0055] At the end of each sub-period, update the prediction scenario of the next sub-period according to the real-time data and re-solve the optimization model;

[0056] Ensure the dynamic adaptability of the scheduling strategy through the rolling window mechanism.

[0057] Preferably, according to the second aspect of the present disclosure, there is provided a regional distributed energy storage optimization scheduling device, including:

[0058] A data acquisition module: used to collect power data, meteorological data, and market electricity price information in the region in real time;

[0059] A model construction module: used to construct a mixed integer nonlinear programming model including multi-objective optimization functions and constraint conditions;

[0060] A scenario generation module: used to generate a multi-scenario data set of renewable energy output, load demand, and electricity price fluctuations;

[0061] An optimization solution module: used to execute an improved mixed integer programming or particle swarm optimization algorithm to solve the multi-objective optimization model;

[0062] A strategy execution module: used to convert the optimization results into energy storage charge and discharge instructions, microgrid operation modes, and power trading plans, and send them to the execution devices through a communication interface.

[0063] Preferably, according to the third aspect of the present disclosure, there is provided an electronic device, including:

[0064] At least one processor;

[0065] And a memory communicatively connected to the processor; wherein, the memory stores instructions executable by the processor, and when the instructions are executed by the processor, the processor is enabled to execute the method according to any one of claims 1 to 7.

[0066] Preferably, according to the third aspect of the present disclosure, there is provided a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the method according to any one of claims 1 to 7.

[0067] In summary, due to the adoption of the above technical solutions, the beneficial effects of the present invention are:

[0068] 1. In this application, multi-dimensional constraints such as energy storage life attenuation, network loss, and carbon emissions are newly added to the constraint conditions to improve the practicability of the model. Through multi-objective optimization and multi-constraint coordination, the limitations of traditional single-objective scheduling are broken through.

[0069] 2. In this application, the optimization algorithm introduces dynamic weight adjustment and penalty function method to enhance the solution efficiency and robustness, and proposes an uncertainty modeling method based on scenario generation and clustering to improve the adaptability of the scheduling strategy. Brief Description of the Drawings

[0070] Figure 1 It shows a flowchart provided according to an embodiment of the present invention. Detailed implementation manners

[0071] 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 of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0072] Please refer to Figure 1 , the present invention provides a technical solution: a regional distributed energy storage optimization scheduling method, including the following steps:.

[0073] S1. Model construction:

[0074] According to the physical characteristics, operation constraints and economic objectives of the regional distribution network, microgrid, energy storage system and renewable energy, a mixed integer nonlinear programming model including a multi-objective optimization function is constructed;

[0075] S2. Scenario generation and processing:

[0076] The Latin hypercube sampling method is used to generate a multi-scenario dataset of renewable energy output, load demand and electricity price fluctuations, and the scenario data is clustered and analyzed to screen out representative typical scenarios to reduce the computational complexity;

[0077] S3. Optimization solution, including the following stages:

[0078] The particle swarm optimization algorithm is introduced, combined with the dynamic weight adjustment strategy, to solve the multi-objective optimization model;

[0079] The particle position is set as the energy storage charge and discharge power, the microgrid operation mode and the purchase and sale electricity decision variables, and the particle velocity is regulated by the inertia weight and the constraint penalty term;

[0080] The constraint conditions are embedded into the objective function through the penalty function method to ensure the feasibility of the solution;

[0081] The sensitivity analysis is performed on the solution results to verify the robustness of the model to the key parameters;

[0082] S4. Strategy update and verification, including the following stages:

[0083] Adjust the energy storage system operation strategy, the microgrid grid connection mode and the regional power trading plan according to the optimization results;

[0084] Through the rolling horizon optimization method, the optimization model is updated in real time to cope with the dynamic environment changes;

[0085] Use simulation tools to verify the stability and economy of the scheduling strategy under transient conditions;

[0086] The present invention simultaneously considers economy (minimizing the total cost), reliability (maximizing power supply reliability), and environmental friendliness (minimizing carbon emissions), and realizes multi-objective coordination through the weighted coefficient method or the Pareto frontier method. For example, by introducing a carbon emission cost coefficient, the environmental protection objective is quantified into a computable economic index, so that the scheduling strategy achieves a balance between economy and sustainability.

[0087] The multi-objective optimization function in step S1 includes:

[0088] Economic objective:

[0089] min(C total = C pu + C st + C lo + C em ), where C pu is the power purchase cost, C st is the energy storage investment and operation and maintenance cost, C lo is the network loss cost; C em is the carbon emission cost;

[0090] Reliability objective:

[0091] where R sy is the system power supply reliability, P su is the reagent power supply power in the i-th period, P lo is the load demand in the i-th period;

[0092] Environmental objective:

[0093] where E em is the total carbon emission, P gr,i is the main grid power purchase quantity, ∈ gr is the main grid carbon emission factor, P dg,i is the distributed power generation quantity, ∈ dg is the distributed power source carbon emission factor.

[0094] The mixed integer nonlinear programming model includes the following constraint conditions:

[0095] Power balance constraint: The total regional load demand is equal to the sum of the regional power generation, energy storage charge and discharge, main grid power purchase and sale, and renewable energy output;

[0096] Energy storage system constraint: including energy storage charge and discharge power limit, state of charge upper and lower limits, charge and discharge efficiency, and life decay model;

[0097] Renewable energy output uncertainty constraint: Based on the scenario generation method, quantify the output fluctuations of wind power generation, photovoltaic power generation and renewable energy;

[0098] Network constraint: including line power flow constraint, node voltage constraint and network loss limit;

[0099] Economic constraint: including the purchase electricity price, the selling electricity price, the energy storage investment cost, the operation and maintenance cost and the carbon emission cost;

[0100] Multi-objective optimization goal: Minimize the regional total cost, maximize the system reliability and minimize the carbon emission as the optimization goals, and complete multi-objective coordination through the weighted coefficient method;

[0101] In addition to the traditional power balance and network constraints, the present invention adds the following key constraints:

[0102] Energy storage life attenuation constraint: By restricting the charge and discharge cycle times, avoid the shortening of the battery life caused by frequent charge and discharge;

[0103] Renewable energy output uncertainty constraint: Combine probability distribution modeling (such as Weibull distribution, Copula function) to quantify the volatility of wind and light output, and enhance the robustness of the scheduling strategy;

[0104] Time-of-use electricity price and user response constraint: Introduce an elastic load model, consider the user's response to the time-of-use electricity price, and optimize the load curve to reduce the peak-valley difference.

[0105] The energy storage system constraints include:

[0106] Charge and discharge power constraint:

[0107] P disc,min ≤P disc,i ≤P disc,max ,P c,min ≤P c,i ≤P c,max ;

[0108] State of charge constraint:

[0109] Where η c is the charging efficiency, η d is the discharging efficiency, and Δt is the time step;

[0110] Life attenuation constraint:

[0111] Where L cycle is the allowable charge and discharge cycle times of the energy storage system over the entire life cycle.

[0112] Step S2 further includes:

[0113] S21. Modeling the output of renewable energy. The output of wind power follows the Weibull distribution, and the output of photovoltaic power follows the normal distribution. Combining historical meteorological data to generate random scenarios;

[0114] S22. Modeling the load demand, adopting the elastic load model under time-of-use electricity price;

[0115] S23. Adopting the real-time electricity price model and combining the predicted data of the day-ahead market to generate electricity price scenarios.

[0116] The particle swarm optimization algorithm includes the following steps:

[0117] Initializing parameters: setting the size N of the particle swarm and the maximum number of iterations T max , the inertia weight ω, and the acceleration coefficients c1 and c2;

[0118] Designing the fitness function: f(x i ) = ω1·C total + ω2·(1 - R sys ) + ω3·E em , where ω1, ω2, and ω3 are weight coefficients satisfying ω1 + ω2 + ω3 = 1;

[0119] Constraint handling strategy:

[0120] For particles violating the constraints, using the dynamic penalty function method:

[0121] where ρ is the penalty factor, and is the violation amount of the g k (x i )th constraint;

[0122] Convergence determination: When the change rate of the fitness function is less than the threshold ∈ in continuous M iterations, terminate the algorithm.

[0123] The rolling horizon optimization method includes the following steps:

[0124] Dividing the optimization period into N sub-periods and performing an optimization solution once within each sub-period;

[0125] At the end of each sub-period, update the predicted scenario of the next sub-period according to the real-time data and re-solve the optimization model;

[0126] Through the rolling window mechanism, ensure the dynamic adaptability of the scheduling strategy.

[0127] The regional distributed energy storage optimal scheduling device includes:

[0128] Data acquisition module: used to collect real-time power data, meteorological data, and market electricity price information within the region;

[0129] Model construction module: used to construct a mixed-integer non-linear programming model containing multi-objective optimization functions and constraint conditions;

[0130] Scenario generation module: used to generate a multi-scenario dataset of renewable energy output, load demand, and electricity price fluctuations;

[0131] Optimization and solution module: used to execute an improved mixed-integer programming or particle swarm optimization algorithm to solve the multi-objective optimization model;

[0132] Strategy execution module: used to convert the optimization results into energy storage charge and discharge instructions, microgrid operation modes, and power trading plans, and send them to the execution devices through the communication interface.

[0133] An electronic device, comprising:

[0134] At least one processor;

[0135] And a memory communicatively connected to the processor; wherein the memory stores instructions executable by the processor, and when the instructions are executed by the processor, the processor is enabled to execute the method according to any one of claims 1 to 7. A non-transitory computer-readable storage medium stores computer instructions for causing a computer to execute the method according to any one of claims 1 to 7.

[0136] The above description of the embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. Regional distributed energy storage optimization scheduling method, characterized in that It includes the following steps: S1. Model construction: According to the physical characteristics, operation constraints and economic objectives of the regional distribution network, microgrid, energy storage system and renewable energy, a mixed-integer nonlinear programming model containing a multi-objective optimization function is constructed; S2. Scenario generation and processing: The Latin hypercube sampling method is used to generate a multi-scenario dataset of renewable energy output, load demand and electricity price fluctuations. Cluster analysis is performed on the scenario data to screen out representative typical scenarios to reduce the computational complexity; S3. Optimization solution, including the following stages: Introduce the particle swarm optimization algorithm, combined with the dynamic weight adjustment strategy, to solve the multi-objective optimization model; Set the particle position as the energy storage charge and discharge power, the microgrid operation mode and the purchase and sale electricity decision variables. The particle velocity is adjusted by the inertia weight and the constraint penalty term; Embed the constraint conditions into the objective function through the penalty function method to ensure the feasibility of the solution; Perform sensitivity analysis on the solution results to verify the robustness of the model to key parameters; S4. Strategy update and verification, including the following stages: Adjust the energy storage system operation strategy, the microgrid grid connection mode and the regional power trading plan according to the optimization results; Through the rolling horizon optimization method, the optimization model is updated in real time to cope with dynamic environmental changes; Use simulation tools to verify the stability and economy of the scheduling strategy under transient conditions.

2. The regional distributed energy storage optimal scheduling method according to claim 1, wherein The multi-objective optimization function in step S1 includes: Economic objective: min(C total = C pu + C st + C lo + C em ), where C pu is the electricity purchase cost, C st is the energy storage investment and operation and maintenance cost, C lo is the network loss cost; C em is the carbon emission cost; Reliability objective: wherein it is R sy System power supply reliability, P su is the reagent power supply power in the i-th period, P lo is the load demand in the i-th period; Environmental objective: Among them, E em is the total carbon emission, P gr,i is the main online power purchase quantity, ∈ gr is the main grid carbon emission factor, P dg,i is the distributed power generation quantity, ∈ dg is the distributed power source carbon emission factor.

3. The regional distributed energy storage optimal scheduling method according to claim 1, wherein The mixed-integer nonlinear programming model contains the following constraint conditions: Power balance constraint: The total regional load demand is equal to the sum of the regional power generation, the energy storage charge and discharge amount, the main grid purchase and sale electricity amount and the renewable energy output; Energy storage system constraints: including energy storage charge and discharge power limits, state of charge upper and lower limits, charge and discharge efficiency and life attenuation model; Renewable energy output uncertainty constraint: Based on the scenario generation method, quantify the output fluctuations of wind power generation, photovoltaic power generation and renewable energy; Network constraints: including line power flow constraints, node voltage constraints and network loss limits; Economic constraints: including purchase electricity price, sale electricity price, energy storage investment cost, operation and maintenance cost and carbon emission cost; Multi-objective optimization goal: Minimize the regional total cost, maximize the system reliability and minimize the carbon emission as the optimization goals, and complete multi-objective coordination through the weighted coefficient method.

4. The regional distributed energy storage optimal scheduling method according to claim 3, wherein The energy storage system constraints include: Charge and discharge power constraint: P disc,min ≤P disc,i ≤P disc,max ,P c,min ≤P c,i ≤P c,max ; State of charge constraint: where η c is the charging efficiency, η d is the discharging efficiency, and Δt is the time step; Life attenuation constraint: Where L cycle is the number of charge-discharge cycles allowed during the entire life cycle of the energy storage system.

5. The regional distributed energy storage optimization scheduling method according to claim 1, wherein Step S2 further includes: S21. Renewable energy output modeling. The wind power output adopts the Weibull distribution, the photovoltaic output adopts the normal distribution, and random scenarios are generated in combination with historical meteorological data; S22. Load demand modeling, using the elastic load model under time-of-use electricity price; S23. Adopt the real-time electricity price model, and generate electricity price scenarios in combination with the day-ahead market forecast data.

6. The regional distributed energy storage optimization scheduling method according to claim 1, wherein The particle swarm optimization algorithm includes the following steps: Initialization parameters: Set the particle swarm size N and the maximum number of iterations T max , the inertia weight ω, and the acceleration coefficients c1 and c2; Fitness function design: f(x i ) = ω1·C total + ω2·(1 - R sys ) + ω3·E em , where ω1, ω2, and ω3 are weight coefficients, satisfying ω1 + ω2 + ω3 = 1; Constraint handling strategy: For particles that violate the constraints, adopt the dynamic penalty function method: where ρ is the penalty factor and is g k (x i ) the violation amount of the i-th constraint; Convergence determination: When the change rate of the fitness function is less than the threshold ∈ in M consecutive iterations, terminate the algorithm.

7. The regional distributed energy storage optimal scheduling method according to claim 1, characterized in that The rolling horizon optimization method includes the following steps: Divide the optimization period into N sub-periods, and perform an optimization solution once within each sub-period; At the end of each sub-period, update the prediction scenario for the next sub-period based on real-time data and re-solve the optimization model; Ensure the dynamic adaptability of the scheduling strategy through a rolling window mechanism.

8. Regional distributed energy storage optimization scheduling device, characterized in that It includes: Data acquisition module: used to collect power data, meteorological data and market electricity price information in the region in real time; Model construction module: used to construct a mixed-integer nonlinear programming model containing multi-objective optimization functions and constraint conditions; Scenario generation module: used to generate a multi-scenario data set of renewable energy output, load demand and electricity price fluctuations; Optimization solution module: used to execute an improved mixed-integer programming or particle swarm optimization algorithm to solve the multi-objective optimization model; Strategy execution module: used to convert the optimization results into energy storage charge and discharge instructions, microgrid operation modes and power trading plans, and send them to the execution devices through a communication interface.

9. An electronic device, characterized in that, It includes: At least one processor; And a memory communicatively connected to the processor; wherein, the memory stores instructions executable by the processor, and when the instructions are executed by the processor, the processor is enabled to execute the method according to any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium, characterized in that, The storage medium stores computer instructions for causing the computer to execute the method according to any one of claims 1 to 7.

Images