A resource planning method for source, grid, load and storage considering flexible supply and demand balance

Abstract

The present invention relates to the field of load-storage resource planning, and specifically discloses a source-grid-load-storage resource planning method that considers the balance of flexible supply and demand, including the following steps: obtaining various resource data for the planning year, constructing a resource flexibility balance model that balances the changes between flexible resources and controllable resources; analyzing the temperature characteristics of battery energy storage, and constructing a battery energy storage operation model with temperature control; constructing a two-layer planning model, establishing a flexible supply and demand balance index for the power system, and performing flexibility inspection according to the operation results output by the lower-layer optimization scheduling model, and then iteratively solving the two-layer planning model until the verification conditions are met to obtain the optimal planning solution. The technical solution of the present invention solves the problem of fully considering energy storage loss in energy storage planning that is difficult to solve in the prior art.

Description

Technical Field

[0001] This invention relates to the field of load-storage resource planning, and specifically discloses a source-grid-load-storage resource planning method that considers flexible supply and demand balance. Background Art

[0002] Currently, the proportion of renewable energy in my country's power system is increasing year by year, especially wind power and photovoltaic power generation, which are characterized by significant randomness and intermittency. To mitigate power fluctuations, the construction of large-scale energy storage stations is imperative. However, existing pumped storage power stations and compressed air energy storage are limited by geographical conditions, making large-scale construction difficult; although battery energy storage technology has made significant progress, long-term economical and efficient operation still faces many challenges.

[0003] Specifically, existing battery energy storage planning methods mainly establish deterministic or stochastic models for stochastic loads and stochastic renewable energy, aiming at optimizing energy storage station capacity configuration and charging / discharging strategies. However, they rarely consider the inherent losses of the energy storage station itself, such as charging / discharging losses and temperature rise losses. This will lead to discrepancies between the energy storage planning results and the actual operating performance.

[0004] On the other hand, existing energy storage loss assessment studies are mainly based on individual energy storage devices, rarely combined with large-scale system energy storage planning. This makes the optimization results unsuitable for directly guiding energy storage planning. Therefore, in the context of economical and efficient power system operation, how to fully consider energy storage loss issues in energy storage planning and achieve a match between theoretical calculations and actual operation is a current technical challenge. Summary of the Invention

[0005] In view of this, the present invention provides a source-grid-load-storage resource planning method that considers flexible supply and demand balance, which can solve the problem that existing technologies cannot fully consider energy storage losses when planning energy storage. The method provided by the present invention specifically includes the following steps:

[0006] S10. Obtain basic data: Obtain various resource data for the planning year, including load, wind power, and installed capacity. Based on historical data and actuarial theory, construct scenarios and calculate the probability of extreme scenarios occurring, and obtain and input typical daily scenarios for the planning year.

[0007] S20. Construct a resource flexibility balance model: Classify flexible resources into energy-type flexible resources, power-type flexible resources, and frequency-type flexible resources, and construct a resource flexibility balance model that balances the changes between flexible resources and controllable resources.

[0008] S30. Construct a battery energy storage operation model: Analyze the temperature characteristics of battery energy storage and construct a battery energy storage operation model with temperature control.

[0009] S40. Build a two-layer model: Based on various constraints, the resource flexibility balance model, and the battery energy storage operation model with temperature control, build a two-layer planning model, establish power system flexibility supply and demand balance indicators, and perform flexibility testing based on the operation results output by the lower-layer optimization scheduling model;

[0010] S50, solving the model: using a non-dominated sorting genetic algorithm to solve the bi-level programming model to obtain multiple optimal planning schemes that meet the flexibility requirements of the power system; wherein, in the process of solving the problem using the non-dominated sorting genetic algorithm, the offspring population obtained in each iteration is optimized using a loss evaluation method; wherein the step of optimizing using the loss evaluation method includes:

[0011] Constructing a loss model: A daily power generation-energy storage loss model is established using a linear regression model to calculate the throughput of battery energy storage based on daily power generation, power consumption, and energy storage.

[0012] Fitting the loss model: The daily power generation-energy storage loss model is fitted using historical annual data to obtain the daily power generation-energy storage loss model, and the annual power loss for the planning year is calculated.

[0013] Offspring population optimization: Select the offspring individuals with the lowest annual power loss in the planning year of the corresponding planning scheme as the optimized offspring population.

[0014] S60. Optimal Plan: From multiple preferred planning plans, select the planning plan with the minimum annual power loss for the planning year as the optimal planning plan.

[0015] The acquisition of various resource data for the planning year includes:

[0016] The acquired resource data includes the maximum load value for the planning year, annual electricity consumption, wind power installed capacity, single unit capacity of electrochemical energy storage, annual utilization hours, unit investment cost and unit operating cost of each resource, discount rate for the planning year, upper and lower limits of conventional unit output power, ramp rate of conventional units, charging and discharging power limits of electrochemical energy storage, call-up cost of day-ahead adjustable load, unit penalty cost of load shedding, coal consumption coefficient, start-up cost, and start-up / shutdown variables; the optimal planning scheme for the power system to meet flexibility requirements includes:

[0017] The planning scheme includes the capacity of conventional units (including thermal power units and hydropower units), the capacity of electrochemical energy storage units, the capacity of pumped storage units, the capacity of wind turbine units, and the interruptible load capacity; investment cost, operating cost and comprehensive cost;

[0018] The process of constructing scenarios and calculating the probability of extreme scenarios based on historical data and actuarial theory includes:

[0019] The extreme scenario is defined as the gradual withdrawal of thermal power from the market and its use as backup power.

[0020] Collect and analyze historical and current observation data;

[0021] Calculate the probability of extreme scenarios based on insurance actuarial theory.

[0022] The calculation of extreme scenario probabilities based on insurance actuarial theory includes:

[0023] Determine the number of thermal power companies and the number of companies that will be delisted within a certain time frame. These companies can be regarded as independent experiments, and delisting can be regarded as the success of the experiment;

[0024] Calculate the relative frequency. The relative frequency is the ratio of the number of successes to the number of trials, that is, the ratio of the number of delisted companies to the total number of thermal power companies. When the number of trials is large enough, the relative frequency will approach the probability of the event. It can be considered that the relative frequency of delisted companies is the probability of thermal power delisting.

[0025] By analyzing historical data and the current market environment, we estimate the expected revenue and expected expenditure of thermal power companies. If the expected revenue is much lower than the expected expenditure, the possibility of the thermal power company being delisted increases;

[0026] Combining the approximate value of relative frequency and the economic evaluation of thermal power enterprises, the probability of delisting is adjusted. Among them, the premium is determined based on the probability that the thermal power enterprise may delist, which can be expressed as E pg =E claims +E costs ;

[0027] Among them, E pg E represents expected premium income; claims E represents the expected compensation payment. costs This indicates the expected operating costs.

[0028] The flexibility resources are classified into the following categories:

[0029] Based on the physical characteristics of various types of flexibility resources, flexibility resources are classified into energy-type flexibility resources, power-type flexibility resources and frequency-type flexibility resources. Among them, energy-type flexibility resources refer to resources that can provide longer usage time on a single charge, power-type flexibility resources refer to resources that can release a large amount of power in a short period of time, and frequency-type flexibility resources have the ability to respond quickly in seconds.

[0030] The resource flexibility balance model, which balances the changes between flexible resources and controllable resources, includes:

[0031] Use flexible resources to adjust the changes in controllable resources, based on ΔP.new +ΔP l =ΔP g +ΔP ess +ΔP DR Construct a resource flexibility balance model that balances the changes between flexible resources and controllable resources;

[0032] Wherein, ΔP new Indicates the change in new energy, ΔP l ΔP represents the change in load. g ΔP represents the change in thermal power generation. ess ΔP represents the change in energy storage. DR This indicates the amount of change in demand response.

[0033] The construction of the battery energy storage operation model with temperature control includes:

[0034] The battery energy storage output power is the sum of the output power of all individual cells, which can be expressed as: P bess (t)=nP(t);

[0035] Among them, P bess P(t) represents the battery energy storage power during time period t; P(t) represents the power of a single battery cell during time period t.

[0036] Considering battery energy storage temperature control, the battery energy storage temperature can be expressed as:

[0037]

[0038] Among them, T bess (t+1),T bess (t) represents the battery storage temperature at time t+1 and time t, respectively; T0(t) represents the ambient temperature at time t; h represents the convective heat transfer coefficient; s represents the battery surface area; C represents the battery specific heat capacity; M represents the battery weight; q(t) represents the battery thermal power at time t; H tc (t), C tc (t) represent the heat power and cooling power generated by the temperature control system during time period t, respectively. bess (t) represents the thermal power of waste heat recovery, n represents the number of individual cells in the battery energy storage compartment, and Δt represents the duration of time period t.

[0039] The battery energy storage life loss cost can be expressed as:

[0040] Among them, Z bess For the cost of battery energy storage lifespan loss, c bess For battery energy storage price, SOC max SOC min These represent the upper and lower limits of the permissible SOC for battery energy storage charging and discharging, respectively, and N. bessFor battery energy storage in the SOC range [SOC max , SOC min The number of times the program can run within a given timeframe, where n is the number of scheduling periods.

[0041] Wherein, the construction of the two-level programming model includes:

[0042] Upper-level investment decision-making model and lower-level optimization scheduling model;

[0043] Among them, the upper-level investment decision model is constructed with the optimal comprehensive cost of the power system as the objective function, and the resource allocation plan is solved and used as the input of the lower-level model;

[0044] By introducing fuzzy opportunity parameters, and based on the resource allocation scheme obtained from the upper layer and the relaxed power balance constraints, constraints for the lower-level optimal scheduling model are established. The lower-level optimal scheduling model is constructed with the optimal operating cost of the power system as the objective function.

[0045] The construction of the lower-level optimal scheduling model with the objective function of optimizing power system operating cost includes:

[0046] The cost objective function Z is calculated based on the operation scheduling results of the typical day s. s Then, based on the number of days D in the planning year of a typical day. s For N s The summation of typical daily scenarios ultimately yields the annual operating cost C. b ,according to Construct the objective function that minimizes the operating cost of the system;

[0047] According to Z s =min(Z) l +Z g +Z ess +Z v +Z w +Z P +Z bess Establish an economic cost objective function;

[0048] Among them, Z l represents the user load cost, Z g Z represents the operating cost of a conventional generating unit. ess Z represents the operating cost of electrochemical energy storage. v Z represents the operating cost of pumped storage hydroelectric power. P Z represents the operating cost of wind turbine units. P Z represents the cost of power curtailment penalties. bess Represents the life loss cost of battery energy storage.

[0049] according to constructing the user load cost;

[0050] Where, N s ρ represents the number of scenes. s c represents the probability of scenario s occurring. dl p represents the actual amount of adjustable load that is called up the day before. dl,t,s c represents the current day's adjustable load call cost. il p represents the actual dispatch cost of adjustable load within the day. il,t,s c represents the actual amount of adjustable load that can be dispatched within the day. sl p represents the unit penalty cost for load shedding. sl,t,s This represents the actual load shedding amount at time t under scenario s.

[0051] according to The operating cost of the aforementioned conventional generating unit;

[0052] Where, N g f represents the planned number of conventional generating units. c Indicates the price of coal, a i b i 、c i P represents the coal consumption coefficient of the i-th unit. gi,t,s This represents the output power of a conventional unit i during time period t in scenario s.

[0053] according to The operating cost of constructing electrochemical energy storage;

[0054] Where, N ess c represents the planned number of electrochemical energy storage devices. essi p represents the operating cost of electrochemical energy storage equipment. essi,t,s Let represent the power generation of the i-th electrochemical energy storage power station at time t in scenario s. This represents the start-up and shutdown power of the i-th electrochemical energy storage power station at time t in scenario s, and is a 0-1 variable. This represents the depreciation cost of an electrochemical energy storage power station.

[0055] according to Construct the operating cost of pumped storage; among which, N v c represents the planned number of pumped storage units. vi p represents the operating cost of a pumped storage unit. vi,t,s Let represent the power generation of the i-th pumped storage unit at time t in scenario s. This represents the start-up and shutdown power consumption of the i-th pumped storage unit at time t in scenario s, and is a 0-1 variable. This represents the depreciation cost of pumped storage units.

[0056] according to Construction and operation costs of wind turbine units;

[0057] Where, N w c represents the planned number of wind power installations. wi p represents the operating cost of wind turbine units. wi,t,s Let represent the output value of the i-th wind turbine at time t in scenario s. U represents the startup cost of the i-th unit. wi,t,s u wi,t-1,s They represent the start and stop variables of the i-th wind turbine at time t and time t-1, respectively, and are 0-1 variables;

[0058] according to Constructing penalty costs for curtailment;

[0059] Among them, f w f pv f h These represent the penalty fees for abandoning wind, solar, and hydropower, respectively. They represent the power generation forecast values ​​of wind farm, photovoltaic power station and hydropower station in scenario s during period t, respectively. w,t,s 、P pv,t,s 、P h,t,s They represent the scheduled power generation of wind farms, photovoltaic power stations, and hydropower stations respectively, and T represents the total number of time periods.

[0060] The upper-level investment decision model is constructed with the optimal comprehensive cost of the power system as the objective function, including:

[0061] According to C = min(C b +C t Construct the objective function of the upper-level investment decision-making model;

[0062] Among them, C b C represents operating costs. t represents the annual investment cost;

[0063] according to Construction of annual investment costs;

[0064] Among them, C i X represents the annual investment cost of the resource to be developed. i Let λ represent a 0-1 variable, where i represents the index, N represents the total number of resources to be built, and λ represents the total number of resources to be built. i represents the annual investment cost of resources to be built, ω i Let n represent the planned capacity, n represent the useful life, and r represent the discount rate.

[0065] The establishment of power system flexibility supply and demand balance indicators includes:

[0066] Select g FP As a flexibility index, its calculation formula is:

[0067] Among them, g FP This indicates the probability of insufficient flexibility. These represent the probabilities of insufficient upward and downward flexibility, respectively. These represent periods of insufficient upward and downward flexibility, respectively, with T representing the number of uphill periods.

[0068] Compared with existing technologies, this invention first establishes a flexibility resource balance model, treating renewable energy sources such as wind power as power-type flexibility resources and loads as frequency-type flexibility resources, fully considering their random fluctuation characteristics. Based on this, a balance relationship between controllable resources and flexibility resources is constructed to achieve a dynamic balance of power system flexibility demands and enhance system immunity. Second, a temperature-constrained battery energy storage model is established. A generalized temperature constraint model is provided to address the differences in temperature requirements for different types of batteries. By controlling the battery health state factor, the degradation characteristics of the batteries are considered, enabling them to operate efficiently under economic and safe conditions. Third, an energy storage optimization configuration model under complex constraints is established. This model fully considers the constraints of random loads, random renewable energy sources, and the batteries themselves, providing an optimal energy storage configuration scheme under the goal of minimizing operating costs. This scheme guides energy storage planning to achieve coordination and mutual benefit with the main power grid. Finally, an energy storage power loss assessment model is established. Based on a linear regression model, this model considers historical load data, renewable energy data, and actual energy storage operation data to quantitatively assess the power loss level of different energy storage configuration schemes, providing a basis for selecting an economical energy storage operation scheme. In summary, the technical solution provided by this invention solves the problem that existing technologies struggle to fully consider energy storage losses during energy storage planning. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0070] Figure 1 This is a flowchart of the method provided by the present invention. Detailed Implementation

[0071] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0072] The method provided by this invention specifically includes the following steps:

[0073] S10. Obtain basic data: Obtain various resource data for the planning year, including load, wind power, and installed capacity. Based on historical data and actuarial theory, construct scenarios and calculate the probability of extreme scenarios occurring, and obtain and input typical daily scenarios for the planning year.

[0074] S20. Construct a resource flexibility balance model: Classify flexible resources into energy-type flexible resources, power-type flexible resources, and frequency-type flexible resources, and construct a resource flexibility balance model that balances the changes between flexible resources and controllable resources.

[0075] S30. Constructing a battery energy storage operation model: analyzing the temperature characteristics of the battery energy storage and constructing a battery energy storage operation model with temperature control;

[0076] S40. Construct a two-layer model: Based on various constraints, a resource flexibility balance model, and a battery energy storage operation model with temperature control, construct a two-layer planning model, establish a power system flexibility supply and demand balance index, and verify flexibility based on the operation results output by the lower-level optimization scheduling model.

[0077] S50. Solving the Model: The bi-level programming model is solved using a non-dominated sorting genetic algorithm to obtain multiple preferred planning schemes that meet the flexibility requirements of the power system. During the solution process using the non-dominated sorting genetic algorithm, the offspring population obtained in each iteration is optimized using a loss assessment method. The steps for optimization using the loss assessment method include:

[0078] Constructing a loss model: Using a linear regression model, a daily power generation-storage loss model is constructed to calculate the battery storage throughput based on daily power generation, power consumption, and storage capacity.

[0079] Fitting the loss model: The daily power generation-energy storage loss model is fitted using historical annual data to obtain the daily power generation-energy storage loss model, and the annual power loss for the planning year is calculated.

[0080] Offspring population optimization: Select the offspring individuals with the lowest annual power loss in the planning year of the corresponding planning scheme as the optimized offspring population.

[0081] S60. Optimal Plan: From multiple preferred planning plans, select the planning plan with the minimum annual power loss for the planning year as the optimal planning plan.

[0082] Specifically, the steps for solving the bilevel programming model are as follows: using the Non-dominated Sorting Genetic Algorithm (NSGA) to perform multi-objective optimization on the bilevel programming model.

[0083] 1. Individual coding

[0084] Real number coding is used, each variable is represented by a real number, and all variables are connected to form an individual. The length of the individual is the number of all variables.

[0085] 2. Population initialization

[0086] Multiple individuals are randomly generated as the initial population, and the population size is set to 100.

[0087] 3. Fitness evaluation

[0088] Each individual is decoded and brought into the bi-level programming model to calculate the objective function value. The objective function includes minimizing the power generation cost, minimizing the amount of wind power curtailment, and minimizing the load response deficiency.

[0089] 4. Non-dominated sorting

[0090] According to the objective function value of each individual, the population is non-dominated and stratified using the fast non-dominated sorting algorithm, and the first frontier is the best layer.

[0091] 5. Select

[0092] (1) The crowding distance method is used as the second-level selection criterion to retain individuals with high fitness and generate the basic offspring population;

[0093] (2) Construct loss model:

[0094] ① Collect historical data, including daily power generation P i , power consumption L i and battery energy storage C i , a total of n days of data.

[0095] ②Establish a linear regression model:

[0096] C i =aP i +bL i +c+∈

[0097] Among them, a, b, c are the parameters of the model, and ∈ is the random error.

[0098] ③ Estimate parameters using ordinary least squares:

[0099] Construct the objective function:

[0100]

[0101] Solve for the parameters a, b, and c to obtain the loss model.

[0102] (3) Fitting loss model:

[0103] ① Collect daily power generation, power consumption, and energy storage data from historical years.

[0104] ② Substitute the data into the loss model to calculate the daily energy storage.

[0105] ③ Calculate the difference between the daily calculated energy storage and the actual energy storage, and use it as the fitting error.

[0106] ④ Adjust the parameters in the loss model to minimize the fitting error and obtain the final daily power generation-storage power loss model.

[0107] (4) Selection of offspring populations:

[0108] ① For each offspring individual of the basic offspring population, calculate the daily power generation and power consumption according to its corresponding planning scheme.

[0109] ② Substitute the daily power generation and power consumption into the daily power generation-energy storage loss model to calculate the annual energy storage loss.

[0110] ③ Select the top x% of offspring individuals with the lowest annual power loss as the optimized offspring population. Here, x can be around 10%.

[0111] 6. Crossover and Mutation

[0112] A new population is generated by performing single-point crossover and mutation operations on the offspring population. The crossover probability is set to 0.8, and the mutation probability is set to 0.1.

[0113] 7. Termination Conditions

[0114] The termination conditions were set to 100 generations or 2 hours of computation time.

[0115] The goal of this algorithm is to eventually obtain a non-dominated frontier through iterative evolution of the population, making it as close as possible to the true Pareto frontier, and thus obtaining multiple optimal planning schemes. These schemes achieve different degrees of balance and trade-offs among various objectives, providing decision-makers with options for reference.

[0116] The acquisition of various resource data for the planning year includes:

[0117] The acquired resource data includes the maximum load value for the planning year, annual electricity consumption, wind power installed capacity, single unit capacity of electrochemical energy storage, annual utilization hours, unit investment cost and unit operating cost of each resource, discount rate for the planning year, upper and lower limits of conventional unit output power, ramp rate of conventional units, charging and discharging power limits of electrochemical energy storage, call-up cost of day-ahead adjustable load, unit penalty cost of load shedding, coal consumption coefficient, start-up cost, and start-up / shutdown variables; the optimal planning scheme for the power system to meet flexibility requirements includes:

[0118] The planning scheme includes the capacity of conventional units (including thermal power units and hydropower units), the capacity of electrochemical energy storage units, the capacity of pumped storage units, the capacity of wind turbine units, and the interruptible load capacity; investment cost, operating cost and comprehensive cost;

[0119] The process of constructing scenarios and calculating the probability of extreme scenarios based on historical data and actuarial theory includes:

[0120] The extreme scenario is defined as the gradual withdrawal of thermal power from the market and its use as backup power.

[0121] Collect and analyze historical and current observation data;

[0122] Calculate the probability of extreme scenarios based on insurance actuarial theory.

[0123] The calculation of extreme scenario probabilities based on insurance actuarial theory includes:

[0124] Determine the number of thermal power companies and the number of companies that will be delisted within a certain time frame. These companies can be regarded as independent experiments, and delisting can be regarded as the success of the experiment;

[0125] Calculate the relative frequency. The relative frequency is the ratio of the number of successes to the number of trials, that is, the ratio of the number of delisted companies to the total number of thermal power companies. When the number of trials is large enough, the relative frequency will approach the probability of the event. It can be considered that the relative frequency of delisted companies is the probability of thermal power delisting.

[0126] By analyzing historical data and the current market environment, we estimate the expected revenue and expected expenditure of thermal power companies. If the expected revenue is much lower than the expected expenditure, the possibility of the thermal power company being delisted increases;

[0127] Combining the approximate value of relative frequency and the economic evaluation of thermal power enterprises, the probability of delisting is adjusted. Among them, the premium is determined based on the probability that the thermal power enterprise may delist, which can be expressed as E pg =E claims +E costs ;

[0128] Among them, E pg E represents expected premium income; claims E represents the expected compensation payment. costs This indicates the expected operating costs.

[0129] The flexibility resources are classified into the following categories:

[0130] Based on the physical characteristics of various flexible resources, they are classified into energy-type flexible resources, power-type flexible resources, and frequency-type flexible resources. Energy-type flexible resources refer to those that can provide longer usage time on a single charge, power-type flexible resources refer to those that can release a large amount of power in a short time, and frequency-type flexible resources have a rapid response capability on the order of seconds.

[0131] The resource flexibility balance model, which balances the changes between flexible resources and controllable resources, includes:

[0132] Use flexible resources to adjust the change of controllable resources according to ΔP new +ΔP l =ΔP g +ΔP ess +ΔP DR Construct a resource flexibility balance model that balances the changes between flexible resources and controllable resources;

[0133] Wherein, ΔP new Indicates the change in new energy, ΔP l ΔP represents the change in load. g ΔP represents the change in thermal power generation. ess ΔP represents the change in energy storage. DR This indicates the amount of change in demand response.

[0134] The construction of the battery energy storage operation model with temperature control includes:

[0135] The battery energy storage output power is the sum of the output power of all individual cells, which can be expressed as: P bess (t)=nP(t);

[0136] Among them, P bess P(t) represents the battery energy storage power during time period t; P(t) represents the power of a single battery cell during time period t.

[0137] Considering battery energy storage temperature control, the battery energy storage temperature can be expressed as:

[0138]

[0139] Among them, T bess (t+1),T bess (t) represents the battery storage temperature at time t+1 and time t, respectively; T0(t) represents the ambient temperature at time t; h represents the convective heat transfer coefficient; s represents the battery surface area; C represents the battery specific heat capacity; M represents the battery weight; q(t) represents the battery thermal power at time t; H tc (t), C tc (t) represent the heat power and cooling power generated by the temperature control system during time period t, respectively. bess(t) represents the thermal power of waste heat recovery, n represents the number of individual cells in the battery energy storage compartment, and Δt represents the duration of time period t.

[0140] The battery energy storage life loss cost can be expressed as:

[0141] Among them, Z bess For the cost of battery energy storage lifespan loss, c bess For battery energy storage price, SOC max SOC min These represent the upper and lower limits of the permissible SOC for battery energy storage charging and discharging, respectively, and N. bess For battery energy storage in the SOC range [SOC max , SOC min The number of times the program can run within a given timeframe, where n is the number of scheduling periods.

[0142] Wherein, the construction of the two-level programming model includes:

[0143] Upper-level investment decision-making model and lower-level optimization scheduling model;

[0144] Among them, the upper-level investment decision model is constructed with the optimal comprehensive cost of the power system as the objective function, and the resource allocation plan is solved and used as the input of the lower-level model;

[0145] By introducing fuzzy opportunity parameters, and based on the resource allocation scheme obtained from the upper layer and the relaxed power balance constraints, constraints for the lower-level optimal scheduling model are established. The lower-level optimal scheduling model is constructed with the optimal operating cost of the power system as the objective function.

[0146] The construction of the lower-level optimal scheduling model with the objective function of optimizing power system operating cost includes:

[0147] according to Construct power balance constraints;

[0148] Among them, P gi P represents the power of a conventional generating unit. w P represents wind power output. ess P represents the electrochemical energy storage power. v This represents the power output of the pumped storage unit, ΔP. w P represents the power of wind curtailment. L Indicates the annual maximum load, γ L This represents the capacity reserve factor.

[0149] according to Construct power balance constraints;

[0150] Among them, T gi T represents the number of hours of thermal power plant utilization. w Indicates the number of wind power utilization hours, Tess T represents the number of hours of electrochemical energy storage utilization. v ΔT represents the number of hours of pumped storage utilization. w Indicates the number of hours of abandoned electricity utilization, γ E represents the reserve rate, E year Indicates the total power demand.

[0151] Since power balance and energy balance calculations are based on the economics of traditional power sources, and only consider the planning balance constraints of these two methods, they lack calculations for the economics of flexible resources. Therefore, further calculations are needed based on... Constructing flexibility supply and demand balance constraints;

[0152] Where, N g Indicates the number of conventional units after planning, N ess Represents the number of electrochemical energy storage devices after planning, N v N represents the planned number of pumped storage units. w This indicates the number of wind power installations after the plan is completed. This indicates the maximum upward adjustment rate of a conventional unit. Indicates the maximum upward adjustment rate of the electrochemical energy storage device, Indicates the maximum upward regulation rate of the pumped storage unit, This indicates the maximum upward adjustment rate of the wind turbine. This indicates the maximum upward adjustment rate of load-related resources. This represents the maximum rate of change of the system's net upward load. This indicates the maximum downward adjustment rate for conventional generating units. Indicates the maximum downward adjustment rate of the electrochemical energy storage device, This indicates the maximum downward adjustment rate of the pumped storage unit. This indicates the maximum downward adjustment rate of the wind turbine. This indicates the maximum downward adjustment rate for load-related resources. P represents the maximum rate of change of the system's net downward load. gi P represents the capacity of a conventional generating unit. essi Represents the capacity of electrochemical energy storage, P wi P represents the capacity of the wind turbine. DR This represents the capacity of load-type resources, (L) t -P Rt ) max Indicates the annual maximum net load.

[0153] The cost objective function Z is calculated based on the operation scheduling results of a typical day s. s Then, based on the number of days D in the planning year of a typical day. s , for N s The sum of typical daily scenarios is the final annual operating cost Cb ,according to Construct the objective function that minimizes the operating cost of the system;

[0154] According to Z s =min(Z) l +Z g +Z ess +Z v +Z w +Z P +Z bess Establish an economic cost objective function;

[0155] Among them, Z l represents the user load cost, Z g Z represents the operating cost of a conventional generating unit. ess Z represents the operating cost of electrochemical energy storage. v Z represents the operating cost of pumped storage hydroelectric power. P Z represents the operating cost of wind turbine units. P Z represents the cost of power curtailment penalties. bess Represents the life loss cost of battery energy storage.

[0156] according to constructing the user load cost;

[0157] Where, N s ρ represents the number of scenes. s c represents the probability of scenario s occurring. dl p represents the actual amount of adjustable load that is called up the day before. dl,t,s c represents the current day's adjustable load call cost. il p represents the actual dispatch cost of adjustable load within the day. il,t,s c represents the actual amount of adjustable load that can be dispatched within the day. sl p represents the unit penalty cost for load shedding. sl,t,s This represents the actual load shedding amount at time t under scenario s.

[0158] according to The operating cost of the aforementioned conventional generating unit;

[0159] Where, N g f represents the planned number of conventional generating units. c Indicates the price of coal, a i b i 、c i P represents the coal consumption coefficient of the i-th unit. gi,t,s This represents the output power of a conventional unit i during time period t in scenario s.

[0160] according to The operating cost of constructing electrochemical energy storage;

[0161] Where, N ess c represents the planned number of electrochemical energy storage devices. essi p represents the operating cost of electrochemical energy storage equipment. essi,t,s represents the power generation of the i-th electrochemical energy storage power station at time t under scenario s, This represents the start-up and shutdown power of the i-th electrochemical energy storage power station at time t in scenario s, and is a 0-1 variable. This represents the depreciation cost of an electrochemical energy storage power station.

[0162] according to Calculate the operating costs of pumped storage hydroelectric power.

[0163] Where, N v c represents the planned number of pumped storage units. vi p represents the operating cost of a pumped storage unit. vi,t,s represents the power generation of the i-th pumped storage unit at time t under scenario s, It represents the start-stop power of the i-th pumped storage unit at time t under scenario s, which is a 0-1 variable. This represents the depreciation cost of pumped storage units.

[0164] according to The operating cost of wind turbine units is constructed; among which, N w c represents the planned number of wind power installations. wi p represents the operating cost of wind turbine units. wi,t,s Let represent the output value of the i-th wind turbine at time t in scenario s. U represents the startup cost of the i-th unit. wi,t,s u wi,t-1,s They represent the start and stop variables of the i-th wind turbine at time t and time t-1, respectively, and are 0-1 variables;

[0165] according to Constructing penalty costs for curtailment;

[0166] Among them, f w f pv f h These represent the penalty fees for abandoning wind, solar, and hydropower, respectively. They represent the power generation forecast values ​​of wind farm, photovoltaic power station and hydropower station in scenario s during period t, respectively. w,t,s 、P pv,t,s 、P h,t,s They represent the scheduled power generation of wind farms, photovoltaic power stations, and hydropower stations respectively, and T represents the total number of time periods.

[0167] The upper-level investment decision model is constructed with the optimal comprehensive cost of the power system as the objective function, including:

[0168] according to Establish power balance constraints; among which, This represents the predicted user load.

[0169] Since the power balance of the system contains uncertainties, the impact of uncertainties should be considered when scheduling the day-ahead dispatch of the power system. Therefore, wind power fuzzy parameters and load fuzzy parameters are introduced to relax the above power balance into a power constraint balance under a certain confidence level α, so that the probability of this balance constraint being true is not less than α. In this way, a set of uncertainties is constructed: Among them, P r {·} indicates the credibility of the event;

[0170] Based on the need to ensure high security in system operation, the confidence level cannot be set too low. When , the power balance constraint is clear and equivalent to the triangular fuzzy parameter class:

[0171]

[0172] Among them, P wt1 、P wt2 、P lt3 、P lt4 They are represented as membership parameters of wind power and load respectively, P gi,t,s represents the output of the i-th conventional unit at time t under scenario s, P essi,t,s represents the output value of the i-th energy storage device at time t under scenario s, P vi,t,s represents the output value of the i-th pumped storage unit at time t under scenario s;

[0173] according to Establish output constraints for conventional units;

[0174] in, This represents the lower limit of the output of the i-th conventional unit. This represents the upper limit of the output of the i-th conventional unit.

[0175] According to ΔD g ≤P gi,t+1,s -P gi,t,s ≤ΔU g Construct ramp rate constraints;

[0176] Where, ΔD g Indicates the upward climbing rate of unit i, ΔU g This represents the rate at which unit i climbs downhill.

[0177] according to Construct operating constraints for wind turbine units;

[0178] Among them, Pwi,t This represents the grid-connected power of the i-th wind farm at time t. Let represent the predicted output of the i-th wind turbine at time t in scenario s.

[0179] according to and Constructing operational constraints for electrochemical energy storage;

[0180] in, These represent the charging and discharging power limits of electrochemical energy storage, respectively. This represents the output of electrochemical energy storage i at time t in scenario s. Respectively represent the lower and upper limits of the electrochemical energy storage capacity, This represents the energy storage capacity of electrochemical energy storage i at time t under scenario s.

[0181] according to Establish operational constraints for pumped storage units;

[0182] in, These represent the charging and discharging power limits of the pumped storage unit i and V, respectively. vi,t,s represents the water storage capacity of pumped storage unit i at time t under scenario s, r represents the lower and upper limits of the water storage capacity of pumped storage units. vi This indicates the ramp rate limit for pumped storage unit i;

[0183] According to -P ij,max ≤B ij (θ i,t -θ j,t )≤P ij,max Establish line transmission capacity constraints;

[0184] Among them, B ij θ is the admittance between nodes i and j; i,t and θ j,t The voltage phase angles at nodes i and j are respectively; P ij,max is the maximum transmission capacity of the line between nodes i and j.

[0185] according to Construct load-side resource constraints;

[0186] in, and They represent the minimum and maximum call amounts of the adjustable load on the day before, and These represent the minimum and maximum call amounts for the adjustable load within a day, respectively.

[0187] According to C = min(C b +C tConstruct the objective function of the upper-level investment decision-making model;

[0188] Among them, C b C represents operating costs. t represents the annual investment cost;

[0189] according to Build annual investment cost;

[0190] Among them, C i represents the annual investment cost of resources to be built, X i Let λ represent a 0-1 variable, where i represents the index, N represents the total number of resources to be built, and λ represents the total number of resources to be built. i ω represents the annual investment cost of the resource to be developed. i Let n represent the planned capacity, n represent the useful life, and r represent the discount rate.

[0191] The establishment of power system flexibility supply and demand balance indicators includes:

[0192] Select g FP is the flexibility index, and its calculation formula is

[0193] Among them, g FP This indicates the probability of insufficient flexibility. These represent the probabilities of insufficient upward and downward flexibility, respectively. These represent periods of insufficient upward and downward flexibility, respectively, with T representing the number of uphill periods.

[0194] The technical principle of the present invention is:

[0195] First, this invention takes the perspective of power system flexibility balance, rationally allocates various controllable and uncontrollable resources, and places energy storage in the core position of resource balance in order to meet the balance requirements under random fluctuations.

[0196] Then, as the main energy storage device, the temperature and health status of the battery significantly affect losses. This invention establishes a temperature-constrained model while determining the energy storage scale and strategy, thereby achieving economical and safe operation of energy storage.

[0197] Then, using a linear regression model and based on historical measured data, the power loss level of different energy storage planning schemes is evaluated to achieve a match between theoretical calculations and actual loss values.

[0198] Finally, the energy storage configuration schemes were evaluated based on the power loss index, and the scheme with the lowest loss was selected to obtain the final optimal energy storage planning scheme.

[0199] In this way, the present invention fully considers the actual possible losses of energy storage stations when planning energy storage, greatly improving the reliability of decision-making results and achieving a high degree of matching between theory and practice.

Claims

1. A source-grid-load-storage resource planning method considering flexibility supply-demand balance, characterized in that: The steps include: S10. Obtain basic data: Obtain various resource data for the planning year, including load, wind power, and installed capacity. Based on historical data and insurance actuarial theory, construct scenarios and calculate the probability of occurrence of extreme scenarios, and obtain and input typical daily scenarios for the planning year; S20. Constructing a resource flexibility balance model: classifying flexibility resources into energy-type flexibility resources, power-type flexibility resources, and frequency-type flexibility resources, and constructing a resource flexibility balance model that balances the variation between flexibility resources and controllable resources; S30, constructing a battery energy storage operation model: analyzing the temperature characteristics of the battery energy storage, and constructing a battery energy storage operation model with temperature control; S40, construct a two-layer model: Based on various constraints, resource flexibility balance model and battery energy storage operation model with temperature control, construct a two-layer planning model, establish the power system flexibility supply and demand balance index, and perform flexibility inspection according to the operation results output by the lower-layer optimization scheduling model; S50, solving the model: using a non-dominated sorting genetic algorithm to solve the bi-level programming model to obtain multiple optimal planning schemes that meet the flexibility requirements of the power system; wherein, in the process of solving the problem using the non-dominated sorting genetic algorithm, the offspring population obtained in each iteration is optimized by loss evaluation; wherein the step of optimizing by loss evaluation includes: Constructing a loss model: Using a linear regression model, a daily power generation-energy storage loss model is constructed to calculate the throughput of battery energy storage based on daily power generation, power consumption, and energy storage. Fitting the loss model: fitting the daily power generation-energy storage loss model using historical data to obtain the daily power generation-energy storage loss model, and calculating the annual power loss in the planning year; Optimization of offspring population: selecting multiple offspring individuals with the lowest annual electricity loss in the planning year of the planning scheme corresponding to the offspring population as the optimized offspring population; S60, selecting a preferred plan: selecting a plan with the smallest annual electricity loss in the planning year from among multiple preferred planning plans as the optimal planning plan; The classification of flexibility resources includes: According to the physical characteristics of various types of flexibility resources, flexibility resources are classified into energy-type flexibility resources, power-type flexibility resources and frequency-type flexibility resources. Among them, energy-type flexibility resources refer to resources that can provide longer usage time on a single charge, power-type flexibility resources refer to resources that can release a large amount of power in a short period of time, and frequency-type flexibility resources have the ability to respond quickly within seconds. The construction of the two-layer planning model includes: an upper-layer investment decision model and a lower-layer optimization scheduling model; Among them, the upper-level investment decision model is constructed with the optimal comprehensive cost of the power system as the objective function, and the resource allocation plan is solved and used as the input of the lower-level model; Fuzzy opportunity parameters are introduced to establish the constraints of the lower-level optimization dispatch model based on the resource allocation scheme obtained from the upper layer and the relaxed power balance constraints. The lower-level optimization dispatch model is constructed with the optimal operation cost of the power system as the objective function. The construction of a resource flexibility balance model that balances the variation between flexible resources and controllable resources includes: Use flexible resources to adjust the change of controllable resources according to ΔP new +ΔP l =ΔP g +ΔP ess +ΔP DR Construct a resource flexibility balance model that balances the changes between flexible resources and controllable resources; Among them, ΔP new Represents the change in new energy, ΔP l Indicates the change in load, ΔP g Indicates the change in thermal power, ΔP ess Indicates the change in stored energy, ΔP DR represents the amount of change in demand response; Wherein, the construction of a battery energy storage operation model with temperature control includes: The battery energy storage output power is the sum of the output power of all single cells, expressed as: P bess (t) = nP(t); Among them, P bess (t) is the battery energy storage power during period t; P(t) is the single battery power during period t; Considering the battery energy storage temperature control, the battery energy storage temperature is expressed as: Among them, T bess (t+1), T bess (t) are the battery energy storage temperatures in periods t+1 and t, respectively; T0(t) is the ambient temperature in period t; h is the convection heat transfer coefficient; s is the battery surface area; C is the battery specific heat capacity; M is the battery weight; q(t) is the battery thermal power in period t; H tc (t), C tc (t) are the heating power and cooling power generated by the temperature control system during period t, H bess (t) is the thermal power of waste heat recovery, n is the number of single cells in the battery energy storage bin, and Δt represents the duration of the t period; The battery energy storage life loss cost is expressed as: Among them, Z bess is the battery energy storage life loss cost, c bess is the battery energy storage price, SOC max , SOC min are the upper and lower limits of SOC allowed for battery energy storage charging and discharging, N bess For battery energy storage in SOC range [SOC min , SOC max ] can be run within the specified time. is the number of scheduling periods; The construction of the lower-level optimization dispatching model with the optimal operation cost of the power system as the objective function includes: The cost objective function Z is calculated based on the operation scheduling results of the typical day s. s , and then according to the number of days in the planning year D s , for N s The sum of typical daily scenarios is finally calculated to get the annual operating cost C b ,according to Construct the objective function of minimizing system operation cost; According to Z s =min(Z l +Z g +Z ess +Z v +Z w +Z P +Z bess ) Establish economic cost objective function; Among them, Z l represents the user load cost, Z g represents the operating cost of conventional units, Z ess represents the operating cost of electrochemical energy storage, Z v represents the operating cost of pumped storage, Z w represents the operating cost of wind turbines, Z P represents the penalty cost of power curtailment, Z bess Represents the lifetime loss cost of battery energy storage; according to constructing the user load cost; Among them, N s represents the number of scenes, ρ s represents the probability of scene s occurring, c dl represents the actual cost of calling the adjustable load on the day before, p dl,t,s represents the actual call amount of the adjustable load on the day before, c il represents the actual cost of calling the adjustable load within the day, p il,t,s represents the actual call amount of the adjustable load within the day, c sl represents the unit penalty cost of load shedding, p sl,t,s represents the actual load shedding amount at time t under scenario s; according to Constructing the operating cost of the conventional unit; Among them, N g represents the number of conventional units after planning, f c represents the coal price, a i 、b i 、c i represents the coal consumption coefficient of the i-th conventional unit, P gi,t,s represents the output power of conventional unit i at time t under scenario s; according to Constructing electrochemical energy storage operating costs; Among them, N ess represents the number of electrochemical energy storage power stations after planning, c essi represents the operating cost of the electrochemical energy storage power station, p essi,t,s represents the power generation of the i-th electrochemical energy storage power station at time t under scenario s, It represents the start-stop power of the i-th electrochemical energy storage power station at time t under scenario s, which is a 0-1 variable. represents the depreciation cost of the electrochemical energy storage power station; according to constructing pumped storage operating costs; Among them, N v represents the number of pumped storage units after planning, c vi represents the operating cost of the pumped storage unit, p vi,t,s represents the power generation of the ith pumped storage unit at time t under scenario s, It represents the start-stop power of the i-th pumped storage unit at time t under scenario s, which is a 0-1 variable. represents the depreciation cost of the pumped storage unit; according to Construct wind turbine operating costs; Among them, N w represents the number of wind turbines after planning, c wi represents the operating cost of wind turbines, p wi,t,s represents the output value of the i-th wind turbine at time t in scenario s, represents the startup cost of the i-th wind turbine, u wi,t,s 、u wi,t-1,s They represent the start and stop variables of the i-th wind turbine at time t and time t-1, respectively, and are 0-1 variables; according to Constructing the penalty cost for curtailment of electricity; Where T represents the total number of time periods, f w 、f pv 、f h They represent the penalty costs for abandoning wind, solar and hydro respectively. They represent the predicted power generation values ​​of wind power station, photovoltaic power station and hydropower station in scenario s during period t, respectively. w,t,s , P pv,t,s , P h,t,s They represent the dispatching planned power generation of wind power station, photovoltaic power station and hydropower station respectively; The upper-level investment decision model is constructed with the optimal comprehensive cost of the power system as the objective function, including: According to C = min (C b +C t ) Construct the objective function of the upper investment decision model; Among them, C b represents the operating cost, C t represents the annual investment cost; according to Build annual investment cost; Among them, C i represents the annual investment cost of resources to be built, X i represents a 0-1 variable, i represents the sequence number, N represents the total number of resources to be built, and λ i represents the annual investment cost per unit capacity of the resource to be built, ω i represents the planned capacity, ξ represents the service life, and r represents the discount rate; The establishment of power system flexibility supply and demand balance indicators includes: Select g FP is the flexibility index, and its calculation formula is Among them, g FP represents the probability of insufficient flexibility, They represent the probability of insufficient upward and downward flexibility, T l up , T l down They represent the periods of insufficient upward and downward flexibility respectively, and T' represents the number of climbing periods.

2. The source-grid-load-storage resource planning method considering the flexibility supply-demand balance according to claim 1 is characterized in that: The various resource data obtained for the planning year include: The various resource data obtained include the maximum load value of the planning year, annual electricity consumption, wind power installed capacity, electrochemical energy storage unit capacity, annual utilization hours, and unit investment cost and unit operating cost of each resource, the discount rate of the planning year, the upper and lower limits of the output power of conventional units, the ramp rate of conventional units, the charging and discharging power limit of electrochemical energy storage, the call cost of adjustable load on the day before, the unit penalty cost of load shedding, coal consumption coefficient, startup cost, and start-stop variables; the optimal planning scheme for the power system to meet the flexibility requirements is obtained, including: The planning scheme includes the capacity of conventional units, the capacity of electrochemical energy storage, the capacity of pumped storage units, the capacity of wind turbine units, and the interruptible load capacity; investment cost, operating cost and comprehensive cost, wherein the conventional units include thermal power units and hydropower units; The construction of scenarios based on historical data and insurance actuarial theory and calculation of the probability of occurrence of extreme scenarios include: The extreme scenario is defined as the gradual withdrawal of thermal power from the market and its use as backup power; Collect and analyze historical data and current observation data; Calculate the probability of extreme scenarios based on insurance actuarial theory.

3. The source-grid-load-storage resource planning method considering flexibility supply-demand balance according to claim 2 is characterized in that: The calculation of extreme scenario probabilities based on insurance actuarial theory includes: Determine the number of thermal power companies and the number of companies that will be delisted within a certain period of time. These companies are regarded as independent experiments, and delisting is regarded as the success of the experiment; Calculate the relative frequency. The relative frequency is the ratio of the number of successes to the number of trials, that is, the ratio of the number of delisted companies to the total number of thermal power companies. When the number of trials is large enough, the relative frequency will approach the probability of the event. It is believed that the relative frequency of delisted companies is the probability of thermal power delisting. By analyzing historical data and the current market environment, the expected income and expected expenditure of thermal power companies are estimated. If the expected income is much lower than the expected expenditure, the possibility of thermal power companies delisting will increase; The probability of delisting is adjusted by combining the approximate value of relative frequency and the economic evaluation of thermal power enterprises. The premium is determined based on the probability that the thermal power enterprises may delist, which is expressed as AND pg =And claims +E costs ; Among them, E pg represents the expected premium income; E claims represents the expected compensation expenditure, E costs Represents the expected operating cost.

Images