Hybrid energy storage capacity configuration and operation joint decision-making method for multiple electricity markets

Abstract

The invention discloses a hybrid energy storage capacity configuration and operation joint decision-making method for multiple power markets, which is used for realizing collaborative optimization of a hybrid energy storage system in planning and operation aspects. According to the method, a double-layer stochastic programming model is constructed, wherein an upper-layer programming model takes annual net income maximization of the hybrid energy storage system as a target, and the rated power and capacity configuration of an energy type battery and a power type battery are optimized; and the lower-layer operation model optimizes a day-ahead market declaration strategy and a real-time response power strategy of the hybrid energy storage system under a day-ahead-real-time two-stage random scene set by taking daily operation expected net income maximization as a target. And the upper-layer planning model and the lower-layer operation model carry out iterative interaction through capacity parameters and operation feasibility information, so that combined optimization of capacity configuration and a multi-market operation strategy is realized. The technical problem that in the prior art, it is difficult for a hybrid energy storage system to achieve planning operation collaborative optimization and reasonable distribution of internal power under multiple uncertainties is solved.

Description

Technical Field

[0001] This invention relates to the field of energy storage technology, and in particular to a joint decision-making method for hybrid energy storage capacity configuration and operation for multiple electricity markets. Background Technology

[0002] With the transformation of the global energy structure, the penetration rate of new energy sources, represented by wind and solar power, in the power system is increasing. However, the intermittency and volatility of new energy output pose significant challenges to the safe and stable operation of the power grid. Energy storage systems, especially electrochemical energy storage, have become a key technology for mitigating new energy fluctuations and improving the flexibility and stability of the power grid due to their advantages such as fast response and flexible configuration. Treating energy storage systems as independent market players, participating in the electricity market and ancillary service markets such as frequency regulation, is an important way to realize their commercial value and promote their large-scale application.

[0003] To address diverse energy and power demands, hybrid energy storage systems (HESS) have emerged, comprising energy storage devices such as batteries and power storage devices such as supercapacitors. The core idea is to leverage the complementary advantages of different energy storage components in their technical characteristics to achieve overall performance improvement, with coordinated power distribution being a key technical challenge.

[0004] In existing technologies, hybrid energy storage power allocation largely relies on fixed filtering strategies, distributing power demand signals to supercapacitors and batteries based on frequency characteristics. While this can protect batteries to some extent, it lacks dynamic response to battery SOC and market prices, potentially leading to overuse of batteries even under adverse conditions, accelerating performance degradation, and reducing system economics. Meanwhile, energy storage capacity planning and operation strategies are generally disconnected: the planning phase relies on simplified models to estimate benefits, often neglecting life-cycle costs and multi-market coupling, resulting in capacity configurations deviating from optimal levels; the operation phase is constrained by predetermined capacity, making it difficult to maximize economic benefits throughout the entire lifecycle.

[0005] Furthermore, the uncertainties in electricity market prices and renewable energy output are key factors influencing HESS decision-making. Existing technologies often struggle to simultaneously account for uncertainties at two different time scales: daily macroeconomic operational patterns and daily real-time random deviations, resulting in insufficient robustness and accuracy of the constructed models. Summary of the Invention

[0006] The purpose of this invention is to provide a joint decision-making method for hybrid energy storage capacity configuration and operation for multiple electricity markets, which solves the technical problem that hybrid energy storage systems in the prior art are unable to achieve coordinated optimization of planning and operation and reasonable allocation of internal power under multiple uncertainties.

[0007] To achieve the above objectives, the technical solution provided by this invention is as follows: a joint decision-making method for hybrid energy storage capacity configuration and operation oriented towards multiple electricity markets. This method constructs a two-level stochastic programming model that combines planning and operation. Considering the uncertainty of new energy output, electricity market price fluctuations, and real-time frequency regulation requirements, it achieves coordinated optimization of the capacity configuration and operation strategies of energy-type and power-type batteries in the hybrid energy storage system. The specific implementation steps are as follows:

[0008] S1: Acquire historical data on renewable energy output, market prices, and real-time frequency regulation signals; use cluster analysis to generate typical daily scenarios that characterize interannual operating characteristics, serving as a description of the uncertainty in the day-ahead phase; and construct real-time scenarios that characterize intraday renewable energy forecasting deviations and real-time frequency regulation demand under each typical daily scenario, thus constructing a two-stage random scenario of day-ahead and real-time.

[0009] S2: Taking the maximization of the annualized net income of the hybrid energy storage system as the upper-level optimization objective, and combining the day-ahead and real-time two-stage stochastic scenarios, the upper-level planning model is used to determine the capacity configuration of energy-type batteries and power-type batteries.

[0010] S3: Taking the maximization of the daily net operating income of the hybrid energy storage system as the lower-level optimization objective, based on the capacity configuration and the day-ahead-real-time two-stage random scenario, a lower-level operating model for the hybrid energy storage system to participate in the day-ahead-real-time market is constructed, and the corresponding constraints are determined.

[0011] S4: Combine the upper-level planning model and the lower-level operation model to form a two-level stochastic programming model that integrates planning and operation. By iteratively solving the two-level stochastic programming model, the optimal rated power and capacity configuration scheme of the hybrid energy storage system, as well as the corresponding day-ahead market reporting strategy and real-time power response strategy, are obtained.

[0012] Furthermore, in step S1, constructing the day-to-day and real-time two-stage stochastic scenario includes:

[0013] Acquire historical data on renewable energy output, market prices, and real-time frequency regulation signals, where the frequency regulation signals are real-time power adjustment instructions issued by the power system dispatching agency or the power market platform;

[0014] Based on the historical data of new energy output and market prices, cluster analysis is used to extract several typical daily scenarios representing different seasons and typical load characteristics from the annual operation data, which are used to describe the prediction uncertainty on an annual scale.

[0015] Within each typical daily scenario, multiple real-time scenarios are constructed based on real-time frequency modulation signals and new energy output prediction deviations to simulate potential power deviations and frequency modulation demands during daily operation, thereby forming a two-stage stochastic scenario that can simultaneously characterize both day-ahead prediction uncertainties and real-time operational uncertainties.

[0016] Furthermore, in step S2, the upper-level planning model aims to maximize the annualized net income of the hybrid energy storage system, where the annualized net income is the difference between the annualized profit of the hybrid energy storage system and the annualized investment cost and annualized operation and maintenance cost of the hybrid energy storage system.

[0017] The annualized profit of the hybrid energy storage system is calculated based on the operation results of the lower-level operation model under the day-ahead and real-time two-stage random scenarios, and the rated power and capacity configuration scheme of the power type battery and the energy type battery are determined accordingly.

[0018] Furthermore, the objective function of the upper-level planning model is:

[0019] ;

[0020] In the formula, To optimize the goals of the upper level, The annualized profit of hybrid energy storage is calculated by the underlying operation model. For the investment cost of hybrid energy storage, For hybrid energy storage operation and maintenance costs;

[0021] Hybrid energy storage investment costs The expression is:

[0022] ;

[0023] Hybrid energy storage operation and maintenance costs The expression is:

[0024] ;

[0025] In the formula, and These are the rated capacity and rated power of the energy-type battery, respectively. and These are the rated capacity and rated power of the power-type battery, respectively. , These refer to the investment cost per unit capacity and the investment cost per unit power of energy-type batteries, respectively. , These are the investment costs per unit capacity and per unit power of power-type batteries, respectively. and These are the unit capacity operation and maintenance costs and unit power operation and maintenance costs of energy-type batteries, respectively. , These are the unit capacity maintenance cost and unit power maintenance cost of power-type batteries, respectively. For the service life of the equipment, is the discount rate.

[0026] Furthermore, the constraints of the upper-level planning model include boundary constraints on the power and capacity configuration of hybrid energy storage, investment budget constraints on hybrid energy storage, energy-power ratio constraints, and feasibility constraints that satisfy the lower-level operation model.

[0027] The boundary constraints for the configuration of hybrid energy storage power and capacity mean that the rated power and rated capacity of both energy-type batteries and power-type batteries must be limited within the technically permissible range, including the minimum and maximum configurable values, in order to ensure the physical feasibility of the hybrid energy storage system.

[0028] The hybrid energy storage investment budget constraint means that the sum of the capacity investment cost and power investment cost of energy type battery and power type battery shall not exceed the preset investment budget, so as to ensure that the planned scheme is economically feasible.

[0029] The energy-power ratio constraint refers to the relationship between the capacity and power of energy-type batteries and power-type batteries, which must meet their technical characteristics. That is, energy-type batteries should have a high energy capacity-power ratio, while power-type batteries should have a high power density, so as to ensure that energy-type batteries and power-type batteries can undertake the corresponding energy support and rapid response tasks in the subsequent operation phase.

[0030] The aforementioned feasibility constraints for satisfying the lower-level operating model refer to the fact that the rated capacity and rated power configuration of the energy-type battery and the power-type battery given in the upper-level planning model must fall within the capacity-power feasible region that ensures that the lower-level operating model has a feasible solution in both the day-to-day and real-time stochastic scenarios. If a candidate configuration results in no feasible solution for the lower-level operating model, it will be excluded from the upper-level planning model.

[0031] Furthermore, the boundary constraints for the hybrid energy storage power and capacity configuration are expressed as follows:

[0032] ;

[0033] ;

[0034] In the formula, , These are the upper limits of the battery's power and capacity, respectively. , These are the upper limits for the power and capacity of a supercapacitor, respectively.

[0035] The investment budget constraint for hybrid energy storage is expressed as follows:

[0036] ;

[0037] In the formula, This represents the investment cost function for hybrid energy storage. This is the upper limit of the investment budget;

[0038] The energy-power ratio constraint is expressed as:

[0039] ;

[0040] ;

[0041] In the formula, The minimum allowable energy-power ratio for a battery. The maximum energy-power ratio allowed by the battery. This represents the minimum allowable energy-to-power ratio for supercapacitors. This represents the maximum allowable energy-to-power ratio for a supercapacitor.

[0042] The feasibility constraints that satisfy the lower-level operating model are expressed as follows:

[0043] ;

[0044] In the formula, To ensure that the lower-level operating model has a feasible capacity-power set under the given day-to-real-time two-stage stochastic scenario, a capacity-power set is required.

[0045] Furthermore, in step S3, the lower-level operation model aims to maximize the daily net operating revenue of the hybrid energy storage system. This daily net operating revenue includes day-ahead electricity trading revenue, real-time electricity deviation costs, frequency regulation market revenue, and related operating costs, and is calculated based on a two-stage stochastic scenario (day-ahead and real-time), wherein:

[0046] The objective function of the lower-level operating model is:

[0047] ;

[0048] In the formula, For the optimization goal of the lower level, This indicates that the expected value of a typical day d is taken according to probability. This means taking the expected value of the real-time scene ω according to probability under a typical day d. For typical day sets, This is a collection of real-time scenarios under typical daytime conditions. For the electricity market revenue of a typical day, For typical daily FM market revenue, For typical d, the volatility-smoothing return, For the electricity market revenue under a typical daily real-time scenario ω, For the frequency modulation market revenue in a typical daily real-time scenario ω, For the volatility mitigation return under a typical daily real-time scenario ω, The loss cost of a hybrid energy storage system under a typical daily real-time scenario ω;

[0049] Typical daily electricity market revenue The expression is:

[0050] ;

[0051] In the formula, The day-ahead electricity market price for a typical day d period t. , These represent the discharge power and charging power of an energy-type battery participating in the electricity market during a typical daily period t (d). The day-ahead scheduling cycle has a value of 24. The time step is the day-ahead time step, with a value of 1 hour.

[0052] Typical daily FM market revenue The expression is:

[0053] ;

[0054] In the formula, , These are the unit price of frequency regulation capacity and the unit price of frequency regulation mileage for a typical day d-time period t. This refers to the daytime frequency regulation capacity declared by the hybrid energy storage system during a typical daytime period t. The average FM mileage is obtained from historical FM signal data;

[0055] Typical daily volatility reduction return The expression is:

[0056] ;

[0057] In the formula, , These represent the discharge power and charging power of an energy-type battery during a typical daily period t, with the fluctuations smoothed out. , These represent the discharge power and charging power of a power-type battery during a typical daily period t, with the fluctuations smoothed out. The unit power price for leasing a hybrid energy storage system during a typical day's d-period;

[0058] Electricity market revenue in a typical daily real-time scenario ω The expression is:

[0059] ;

[0060] In the formula, For a typical daily real-time scenario ω period Real-time electricity market price This is the penalty coefficient for deviations in the electricity market. , These represent the typical daily real-time scenarios (ω-time period) of energy-type batteries in the electric energy market. The declared positive and negative imbalance amounts of charging and discharging power. The real-time scheduling period is set to 96. This is the real-time time step, with a value of 15 minutes.

[0061] Frequency modulation market revenue in typical daily real-time scenarios. The expression is:

[0062] ;

[0063] In the formula, , These represent typical daily real-time scenarios with time intervals ω. The unit price of frequency modulation capacity and the unit price of frequency modulation mileage, This is the frequency modulation market deviation penalty coefficient. , These represent typical daily real-time scenarios (ω-hours) for hybrid energy storage systems in the frequency regulation market. The declared frequency regulation capacity positive and negative imbalance;

[0064] Fluctuation mitigation returns in a typical daily real-time scenario ω The expression is:

[0065] ;

[0066] In the formula, For a typical daily real-time scenario ω period The unit power price for leasing hybrid energy storage systems. To mitigate fluctuation deviation penalty coefficient, , The hybrid energy storage system smooths out fluctuations during the ω period of a typical daily real-time scenario. Positive and negative imbalance in charging and discharging power.

[0067] Loss cost of hybrid energy storage system in typical daily real-time scenario ω The expression is:

[0068] ;

[0069] ;

[0070] ;

[0071] ;

[0072] ;

[0073] In the formula, This addresses the cost of energy-type batteries used to mitigate fluctuations and participate in the electricity market. This is to reduce the cost of power batteries used to mitigate fluctuations and participate in the electricity market. The cost of energy-type batteries participating in the frequency regulation market. The loss cost of power batteries participating in the frequency modulation market; For energy-type batteries during time periods The cost per unit of energy loss varies with the depth of discharge. For power batteries during time periods Average cost per unit of energy lost , The respective costs per unit frequency regulation mileage for energy-type batteries and power-type batteries are: For a typical daily real-time scenario ω period Frequency modulation mileage coefficient, , These represent energy-type batteries in a typical daily real-time scenario during the ω period. Charging and discharging power participating in the electricity market , These represent energy-type batteries in a typical daily real-time scenario during the ω period. Smooth out fluctuations in charging and discharging power. , These represent the power-type batteries during typical daily real-time scenarios at the ω-hour interval. Charging and discharging power participating in the electricity market , These represent the power-type batteries in a typical daily real-time scenario during the ω period. Smooth out fluctuations in charging and discharging power. For energy-type batteries during typical daily d periods The frequency modulation capacity recently declared For power batteries during typical daily d periods The frequency modulation capacity recently declared.

[0074] Furthermore, the constraints of the lower-level operation model include day-ahead stage constraints and real-time stage constraints. The day-ahead stage constraints include new energy power fluctuation smoothing constraints, frequency regulation capacity application constraints, hybrid energy storage power constraints, and hybrid energy storage energy constraints.

[0075] The aforementioned new energy power fluctuation smoothing constraint is used to couple the energy storage smoothing power with the new energy output in order to limit the variation of grid-connected power in adjacent time periods.

[0076] The frequency regulation capacity declaration constraint is used to determine the capacity that the hybrid energy storage system can declare to the frequency regulation market during the day-ahead phase, and requires that the declared capacity match the rated charge and discharge power capability of the hybrid energy storage system.

[0077] The hybrid energy storage power constraint is used to limit the charging and discharging power of the energy type battery and the power type battery at any time to not exceed the rated power, and to prevent charging and discharging from occurring simultaneously through state mutual exclusion conditions.

[0078] The hybrid energy storage energy constraint is used to apply upper and lower limits of the state of charge to the energy type battery and the power type battery respectively, and to ensure the energy continuity during the time-series operation through the energy balance equation, so as to avoid the energy exceeding the limit.

[0079] Furthermore, the real-time stage constraints are structurally consistent with the day-ahead stage constraints but are distinguished in terms of scenario and time scale. In addition, they also include power deviation constraints and internal power coordination strategy constraints of hybrid energy storage.

[0080] The power deviation constraint is used to quantify the deviation of real-time operation from the day-ahead plan, including power deviation, frequency regulation capacity deviation and power smoothing deviation, and to reduce the impact of real-time deviation on plan execution through penalty terms.

[0081] The internal power coordination strategy constraint of the hybrid energy storage is used to achieve complementary responses between the energy-type battery and the power-type battery. It includes: constructing a dynamic power-bearing weight coefficient based on the state-of-charge deviation of the energy-type battery, and adjusting the proportion of real-time frequency regulation power of the energy-type battery during the real-time operation phase accordingly; controlling the power-type battery to bear the remaining real-time frequency regulation power not borne by the energy-type battery through the power balance relationship, thereby maintaining the state of charge and operational feasibility of the energy-type battery while ensuring the real-time response capability of the hybrid energy storage system.

[0082] Furthermore, in step S4, the two-level stochastic programming model is solved iteratively, including the following solution process:

[0083] The upper-level planning model solves for the rated power and capacity configuration of the hybrid energy storage system, using the capacity configuration as the optimization variable. The operational feasibility information and operational benefit information returned by the lower-level operation model are transformed into cutting planes and added to the upper-level planning model. The capacity configuration scheme is gradually corrected by continuously updating the cutting planes.

[0084] Given the capacity configuration output by the upper-level planning model, the lower-level operation model is solved. The typical daily scenario and its included real-time scenario are divided into multiple sub-problems that can be solved in parallel. The solution results of each sub-problem are used to generate a cutting plane that characterizes the daily reporting revenue and real-time response revenue, and then returned to the main problem of the lower-level operation model.

[0085] The upper-level planning model and the lower-level operation model interact iteratively. The upper-level planning model updates the capacity configuration based on the feedback from the lower-level operation model, and the lower-level operation model re-solves the operation strategy based on the updated capacity configuration.

[0086] Repeat the above solution process until the objective functions of the upper-level planning model and the lower-level operation model converge or the addition of a cutting plane no longer changes the solution results. This will give you the capacity configuration scheme with the optimal life cycle benefit of the hybrid energy storage system and its corresponding operation strategy.

[0087] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0088] 1. Achieving coordinated optimization of hybrid energy storage planning and operation: This invention constructs a two-layer stochastic programming model that combines an upper-layer planning model and a lower-layer operation model. It optimizes the capacity configuration of energy-type batteries and power-type batteries with the operation strategies of multiple electricity markets in a unified manner, ensuring that the capacity configuration scheme has a feasible operating solution in typical daily-real-time scenarios and achieves optimal economic benefits throughout the entire life cycle.

[0089] 2. To achieve coordinated operation decisions across multiple electricity markets, comprehensively consider the benefits of hybrid energy storage systems in mitigating new energy fluctuations, the electricity market, and the frequency regulation market. Establish power decomposition and capacity coupling constraints, coordinate the output allocation of energy-type batteries and power-type batteries in different markets, and enable the benefits of multiple markets to achieve dynamic equilibrium and overall optimization in multiple scenarios.

[0090] 3. Refined modeling of hybrid energy storage system lifetime loss models: Differentiated lifetime loss models are constructed for two typical operating conditions: low-frequency deep charge / discharge and high-frequency shallow charge / discharge. For low-frequency deep charge / discharge, the traditional rainflow counting method is nonlinear and difficult to embed into the MILP optimization model; therefore, a rainflow-like counting method based on state memory is constructed. For high-frequency shallow charge / discharge, a linearized loss model based on frequency regulation mileage is constructed.

[0091] 4. Enhance the robustness of multi-market decision-making: This invention quantifies the power deviation, frequency regulation capacity deviation and power stabilization deviation through power deviation constraints, and sets deviation penalties in the operating targets, so that day-ahead decisions have stronger executability and robustness in the face of multiple uncertainties, thereby improving the operational reliability of hybrid energy storage systems in complex market environments. Attached Figure Description

[0092] Figure 1 This is a flowchart of the method of the present invention. Detailed Implementation

[0093] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0094] This embodiment discloses a joint decision-making method for hybrid energy storage capacity configuration and operation in a multi-electricity market environment. It constructs a two-level stochastic programming model combining planning and operation, simultaneously considering the uncertainty of new energy output, electricity market price fluctuations, and real-time frequency regulation requirements. This achieves coordinated optimization of the capacity configuration and operation strategies of energy-type and power-type batteries in the hybrid energy storage system. Figure 1 As shown, the specific implementation steps are as follows:

[0095] S1: Acquire historical data on renewable energy output, market prices, and real-time frequency regulation signals; use cluster analysis to generate typical daily scenarios that characterize interannual operating characteristics, serving as a description of the uncertainty in the day-ahead phase; and construct real-time scenarios that characterize intraday renewable energy forecasting deviations and real-time frequency regulation demand under each typical daily scenario, thus constructing a two-stage random scenario of day-ahead and real-time.

[0096] Among them, historical power output data for renewable energy refers to historical power data of renewable energy units such as wind farms and photovoltaic power plants, used to characterize the volatility of renewable energy; market price data includes day-ahead electricity market prices and real-time market prices, used to reflect market economic signals at different time scales; real-time frequency regulation signal data is used to characterize the frequency regulation demand of the regional power grid. Typical daily scenarios refer to selecting representative daily operating modes from a large number of historical characteristic days, used to reduce the computational scale while retaining operating characteristics; real-time scenarios refer to random samples that characterize the intraday renewable energy forecast deviation and frequency regulation demand, used to reflect the uncertainty of real-time operation.

[0097] S2: Taking the maximization of the annualized net income of the hybrid energy storage system as the upper-level optimization objective, and combining the day-ahead and real-time two-stage stochastic scenarios, the upper-level planning model is used to determine the capacity configuration of energy-type batteries and power-type batteries.

[0098] This embodiment takes maximizing the annualized net income of the hybrid energy storage system as the upper-level planning objective. By optimizing the rated power and capacity of energy-type batteries and power-type batteries, the economic optimization of energy storage investment planning is achieved.

[0099] Among them, annualized net income refers to the net benefit obtained by the hybrid energy storage system within a year after deducting the corresponding annualized investment cost and operation and maintenance cost, which is used to characterize the long-term economics of the energy storage configuration scheme; annualized profit is calculated by the lower-level operation model under a typical daily-real-time two-stage stochastic scenario, including the income obtained by the energy storage system from participating in the electric energy market, frequency regulation market and smoothing the fluctuations of new energy sources; energy-type batteries are used to provide energy support and have a large energy capacity; power-type batteries are used to provide fast power response and have a high power density.

[0100] In this embodiment, the objective function of the upper-level planning model is:

[0101] ;

[0102] In the formula, To optimize the goals of the upper level, The annualized profit of hybrid energy storage is calculated by the underlying operation model. For the investment cost of hybrid energy storage, For hybrid energy storage operation and maintenance costs;

[0103] Hybrid energy storage investment costs The expression is:

[0104] ;

[0105] Hybrid energy storage operation and maintenance costs The expression is:

[0106] ;

[0107] In the formula, and These are the rated capacity and rated power of the energy-type battery, respectively. and These are the rated capacity and rated power of the power-type battery, respectively. , These refer to the investment cost per unit capacity and the investment cost per unit power of energy-type batteries, respectively. , These are the investment costs per unit capacity and per unit power of power-type batteries, respectively. and These are the unit capacity operation and maintenance costs and unit power operation and maintenance costs of energy-type batteries, respectively. , These are the unit capacity maintenance cost and unit power maintenance cost of power-type batteries, respectively. For the service life of the equipment, is the discount rate.

[0108] The constraints of the upper-level planning model include boundary constraints on hybrid energy storage power and capacity configuration, hybrid energy storage investment budget constraints, energy-power ratio constraints, and feasibility constraints that satisfy the lower-level operation model; these are used to ensure that the capacity configuration scheme obtained by the hybrid energy storage system in the planning stage is feasible and economically reasonable in the operation stage.

[0109] The power and capacity configuration boundary constraints of hybrid energy storage refer to the requirement that the rated power and rated capacity of both energy-type batteries and power-type batteries must be limited within the technically permissible range, including the minimum and maximum configurable values, in order to ensure the physical feasibility of the hybrid energy storage system.

[0110] The investment budget constraint for hybrid energy storage means that the sum of the capacity investment cost and power investment cost of energy type batteries and power type batteries must not exceed the preset investment budget, so as to ensure that the planned scheme is economically feasible.

[0111] The energy-power ratio constraint refers to the relationship between the capacity and power of energy-type batteries and power-type batteries. For example, energy-type batteries should have a high energy capacity-power ratio, while power-type batteries should have a high power density, so as to ensure that these two types of energy storage units can undertake the corresponding energy support and rapid response tasks in the subsequent operation phase.

[0112] Satisfying the feasibility constraints of the lower-level operating model means that the capacity and rated power configuration of the energy-type and power-type batteries given by the upper-level planning model must fall within the capacity-power feasible region that guarantees a feasible operating solution for the lower-level operating model in both day-ahead and real-time stochastic scenarios. If a candidate configuration causes the lower-level operating model to fail to meet the power balance, energy constraint, or frequency regulation requirements in any scenario, then that configuration is deemed infeasible and eliminated in the upper-level planning model.

[0113] The boundary constraints for hybrid energy storage power and capacity configuration are expressed as follows:

[0114] ;

[0115] ;

[0116] in, , These are the upper limits of the battery's power and capacity, respectively. , These represent the upper limits of power and capacity for supercapacitors, respectively.

[0117] The investment budget constraint for hybrid energy storage is expressed as follows:

[0118] ;

[0119] in, This represents the investment cost function for hybrid energy storage. This is the upper limit of the investment budget.

[0120] The energy-power ratio constraint is expressed as:

[0121] ;

[0122] ;

[0123] in, The minimum allowable energy-power ratio for a battery. The maximum energy-power ratio allowed by the battery. This represents the minimum allowable energy-to-power ratio for supercapacitors. This represents the maximum allowable energy-to-power ratio for a supercapacitor.

[0124] The feasibility constraints for satisfying the lower-level operational model are expressed as follows:

[0125] ;

[0126] in, To ensure that the underlying operational problems have feasible solutions under the given typical daily and real-time scenario sets, a capacity-power set is required.

[0127] S3: Taking the maximization of the daily net operating income of the hybrid energy storage system as the lower-level optimization objective, based on the capacity configuration and the day-ahead-real-time two-stage random scenario, a lower-level operating model for the hybrid energy storage system to participate in the day-ahead-real-time market is constructed, and the corresponding constraints are determined.

[0128] In this embodiment, the lower-level operating model aims to maximize daily net operating profit, and its objective function is:

[0129] ;

[0130] In the formula, For the optimization goal of the lower level, This indicates that the expected value of a typical day d is taken according to probability. This means taking the expected value of the real-time scene ω according to probability under a typical day d. For typical day sets, This is a collection of real-time scenarios under typical daytime conditions. For the electricity market revenue of a typical day, For typical daily FM market revenue, For typical d, the volatility-smoothing return, For the electricity market revenue under a typical daily real-time scenario ω, For the frequency modulation market revenue in a typical daily real-time scenario ω, For the volatility mitigation return under a typical daily real-time scenario ω, The loss cost of a hybrid energy storage system under a typical daily real-time scenario ω;

[0131] Among them, the electricity market revenue of a typical day The expression is:

[0132] ;

[0133] In the formula, The day-ahead electricity market price for a typical day d period t. , These represent the discharge power and charging power of an energy-type battery participating in the electricity market during a typical daily period t (d). The day-ahead scheduling cycle has a value of 24. The time step is the day-ahead time step, with a value of 1 hour.

[0134] Typical daily FM market revenue The expression is:

[0135] ;

[0136] In the formula, , These are the unit price of frequency regulation capacity and the unit price of frequency regulation mileage for a typical day d-time period t. This refers to the daytime frequency regulation capacity declared by the hybrid energy storage system during a typical daytime period t. The average FM mileage is obtained from historical FM signal data;

[0137] Typical daily volatility reduction return The expression is:

[0138] ;

[0139] In the formula, , These represent the discharge power and charging power of an energy-type battery during a typical daily period t, with the fluctuations smoothed out. , These represent the discharge power and charging power of a power-type battery during a typical daily period t, with the fluctuations smoothed out. The unit power price for leasing a hybrid energy storage system during a typical day's d-period;

[0140] Electricity market revenue in a typical daily real-time scenario ω The expression is:

[0141] ;

[0142] In the formula, For a typical daily real-time scenario ω period Real-time electricity market price This is the penalty coefficient for deviations in the electricity market. , These represent the typical daily real-time scenarios (ω-time period) of energy-type batteries in the electric energy market. The declared positive and negative imbalance amounts of charging and discharging power. The real-time scheduling period is set to 96. This is the real-time time step, with a value of 15 minutes.

[0143] Frequency modulation market revenue in typical daily real-time scenarios. The expression is:

[0144] ;

[0145] In the formula, , These represent typical daily real-time scenarios with time intervals ω. The unit price of frequency modulation capacity and the unit price of frequency modulation mileage, This is the frequency modulation market deviation penalty coefficient. , These represent typical daily real-time scenarios (ω-hours) for hybrid energy storage systems in the frequency regulation market. The declared frequency regulation capacity positive and negative imbalance;

[0146] Fluctuation mitigation returns in a typical daily real-time scenario ω The expression is:

[0147] ;

[0148] In the formula, For a typical daily real-time scenario ω period The unit power price for leasing hybrid energy storage systems. To mitigate fluctuation deviation penalty coefficient, , The hybrid energy storage system smooths out fluctuations during the ω period of a typical daily real-time scenario. Positive and negative imbalance in charging and discharging power.

[0149] Loss cost of hybrid energy storage system in typical daily real-time scenario ω The expression is:

[0150] ;

[0151] ;

[0152] ;

[0153] ;

[0154] ;

[0155] In the formula, This addresses the cost of energy-type batteries used to mitigate fluctuations and participate in the electricity market. This is to reduce the cost of power batteries used to mitigate fluctuations and participate in the electricity market. The cost of energy-type batteries participating in the frequency regulation market. The loss cost of power batteries participating in the frequency modulation market; For energy-type batteries during time periods The cost per unit of energy loss varies with the depth of discharge. For power batteries during time periods Average cost per unit of energy lost , The respective costs per unit frequency regulation mileage for energy-type batteries and power-type batteries are: For a typical daily real-time scenario ω period Frequency modulation mileage coefficient, , These represent energy-type batteries in a typical daily real-time scenario during the ω period. Charging and discharging power participating in the electricity market , These represent energy-type batteries in a typical daily real-time scenario during the ω period. Smooth out fluctuations in charging and discharging power. , These represent the power-type batteries in a typical daily real-time scenario during the ω period. Charging and discharging power participating in the electricity market , These represent the power-type batteries in a typical daily real-time scenario during the ω period. Smooth out fluctuations in charging and discharging power. For energy-type batteries during typical daily d periods The frequency modulation capacity recently declared For power batteries during typical daily d periods The frequency modulation capacity recently declared.

[0156] Furthermore, this paper discusses in detail the various loss costs of hybrid energy storage: Energy storage systems operate significantly differently in the electricity market and the frequency regulation market, leading to inconsistent aging mechanisms and lifetime loss rates. The electricity market primarily utilizes low-frequency, deep charge-deep discharge energy regulation, while the frequency regulation market exhibits high-frequency, shallow charge-shallow discharge power regulation characteristics. These two approaches have significantly different impacts on the lifetime of energy storage systems. Traditional lifetime models based on constant depth of discharge (DOD) cycle testing are insufficient to accurately characterize the aging process dominated by high-frequency micro-cycles under frequency regulation conditions. Therefore, this paper constructs differentiated lifetime loss models for two typical operating conditions: low-frequency deep charge-deep discharge and high-frequency shallow charge-shallow discharge.

[0157] During arbitrage in the energy market, energy storage systems primarily exhibit low-frequency, deep-discharge behavior. Battery aging under these conditions is mainly driven by volume expansion and mechanical fatigue of electrode materials, with the degree of loss exhibiting a highly nonlinear relationship with the depth of discharge. Traditional rainflow counting methods typically rely on post-processing of full-time-domain waveforms, making it impossible to directly embed mixed-integer linear programming (MILP) models for real-time solutions. To accurately quantify this type of loss within the MILP framework, a state-memory-based rainflow counting method is constructed.

[0158] To accurately identify the start and end times of a charge / discharge half-cycle within a discrete time step, continuous state variables are introduced. Characterizes the charging and discharging trend of the battery. Unlike instantaneous state variables, It possesses a memory effect, meaning that when the energy storage system is idle or in a zero-power state, it maintains the trend state from the previous moment, thus preventing the charging (discharging) process before and after the zero-interaction state from being split into two charging (discharging) processes. Its state transition constraint is defined as:

[0159] ;

[0160] In the formula: For memory variables, , They are respectively The charging and discharging status bits at any given time, as described above, ensure that when the system is inactive ( When forced .

[0161] Based on this, define the semi-loop transition flag. If and only if the charging and discharging trends reverse ( When a step occurs This marks the end of the previous half-cycle and the beginning of the current half-cycle. To accurately capture the switching point between charge and discharge half-cycles, a linearized auxiliary constraint is introduced to describe its absolute value logic:

[0162] ;

[0163] ;

[0164] ;

[0165] ;

[0166] The first two constraints constitute The lower bound constraint. When and At different times, forced This allows for the identification of turning points. The latter two constraints constitute... The upper bound constraint. When Limit during (continuous discharge or idle) ;when Limitation during (continuous charging or idle) These two constraints ensure that, when the state has not reversed, the forced... This prevents incorrect loop counting.

[0167] To calculate the effective depth of discharge within the current half-cycle, define the variable... This represents the cumulative energy throughput since the last state transition. If no transition occurred ( If a turning point occurs, the accumulated amount from the previous moment is inherited and the current increment is added; if a turning point occurs ( If the historical cumulative amount is reset, only the current increment will be retained.

[0168] ;

[0169] In the formula, and They are respectively and The cumulative energy throughput of energy stored at all times; , They are respectively The charging and discharging power of time-limited energy storage; , These are the charging and discharging efficiencies of energy storage, respectively. This refers to the time interval.

[0170] To eliminate the above nonlinear product terms Introducing auxiliary variables This represents the historical cumulative amount that needs to be truncated at the turning point, i.e. At this point, the cumulative energy update formula transforms into a linear form:

[0171] ;

[0172] In the formula, This is an auxiliary variable for cumulative battery power.

[0173] A set of linear inequality constraints is constructed using the Big M method to accurately describe... and Logical relationship:

[0174] ;

[0175] In the formula, It is a sufficiently large positive number.

[0176] Calculate the current real-time depth of discharge of the energy storage :

[0177] ;

[0178] In the formula, For energy storage Depth of discharge at any given time; This refers to the rated energy storage capacity. Because the cycle life of a battery has a non-linear relationship with its depth of discharge, different depths of discharge need to be considered. The half-cycle is converted to the equivalent number of cycles at 100% depth of discharge. The conversion factor is defined as follows:

[0179] ;

[0180] In the formula, for The equivalent number of cycles at 100% depth of discharge at a given time; Depth of discharge The corresponding theoretical maximum number of loops, The cycle life is represented by 100% depth of discharge. A piecewise linearization strategy is used to handle this nonlinear mapping relationship.

[0181] To calculate the total loss throughout the day, it is necessary to calculate the loss at each time point. The resulting increase in losses. Due to This represents the total accumulated loss up to the present half-cycle, therefore at time... The incremental loss should be the current value minus the value at the previous moment. However, when the charging / discharging state changes ( When this occurs, it signifies the start of a new cycle, and the historical loss from the previous moment should not be subtracted (i.e., truncated). Therefore, the formula for calculating the loss increment is:

[0182] ;

[0183] In the formula, For energy storage The incremental lifespan loss at any given moment.

[0184] The above formula contains a nonlinear product term. Introducing auxiliary variables Instead of that, i.e. The following linear constraint set is constructed using the Big M method to implement logical judgments:

[0185] ;

[0186] In the formula, It serves as an auxiliary variable for the increment of lifespan loss.

[0187] Ultimately, the energy storage system is constantly The equivalent cyclic loss increment is expressed in linear form:

[0188] ;

[0189] When participating in the high-frequency component response of primary and secondary frequency regulation (AGC), the energy storage system is in a state of high-frequency, small-amplitude power fluctuations. Under such conditions, the battery undergoes a massive amount of micro-cycles. The traditional rainflow counting method based on complete deep charge and discharge is not only computationally burdensome but also difficult to accurately assess the cumulative impact of shallow charge and discharge on battery life. Considering computational efficiency and model accuracy, a linearized life loss model for frequency regulation mileage is established.

[0190] Frequency regulation mileage is a key indicator for measuring the cumulative workload of an energy storage system in response to high-frequency regulation signals. It is defined as the frequency regulation mileage over a certain period of time, which is the time integral of the change in system output power. With its declared frequency modulation capacity There exists a relatively stable linear relationship:

[0191] ;

[0192] In the formula, For energy storage at any time Frequency regulation mileage; for energy storage at any time Frequency modulation capacity; This refers to the unit frequency regulation mileage within this time period, and its physical meaning is the unit capacity within the time period. The theoretical response mileage accumulated within the system.

[0193] To convert FM mileage into quantifiable lifetime loss, it is first converted into the equivalent number of complete cycles. .

[0194] ;

[0195] In the formula: the coefficient 2 in the denominator indicates that a complete cycle includes two processes: charging and discharging; The average cycle depth for frequency modulation services.

[0196] Considering that the average depth of charge and discharge under frequency modulation conditions is much less than 100%, a micro-circulation weighting coefficient is introduced. The degree of damage is corrected. This coefficient reflects the proportion of unit damage of microcirculation relative to deep circulation under the same energy throughput, and its value varies for different energy storage systems. At this point, the equivalent lifetime loss cost generated by high-frequency microcirculation can be expressed as:

[0197] ;

[0198] In the formula: The equivalent lifetime loss cost caused by high-frequency microcirculation; This represents the theoretical cycle life at this average depth. This refers to the investment cost of energy storage.

[0199] Due to specific power grid and battery technology routes, , All parameters are constants, and all constant terms can be combined into a comprehensive unit frequency modulation loss coefficient:

[0200] ;

[0201] In the formula, Cost per unit frequency modulation mileage loss.

[0202] The constraints of the lower-level operation model include day-ahead stage constraints and real-time stage constraints. The day-ahead stage constraints include new energy power fluctuation smoothing constraints, frequency regulation capacity application constraints, hybrid energy storage power constraints, and hybrid energy storage energy constraints.

[0203] The aforementioned new energy power fluctuation smoothing constraint is used to couple the energy storage smoothing power with the new energy output in order to limit the variation of grid-connected power in adjacent time periods.

[0204] The frequency regulation capacity declaration constraint is used to determine the capacity that the hybrid energy storage system can declare to the frequency regulation market during the day-ahead phase, and requires that the declared capacity match the rated charge and discharge power capability of the hybrid energy storage system.

[0205] The hybrid energy storage power constraint is used to limit the charging and discharging power of the energy type battery and the power type battery at any time to not exceed the rated power, and to prevent charging and discharging from occurring simultaneously through state mutual exclusion conditions.

[0206] The hybrid energy storage energy constraint is used to apply upper and lower limits of the state of charge to the energy type battery and the power type battery respectively, and to ensure the energy continuity during the time-series operation through the energy balance equation, so as to avoid the energy exceeding the limit.

[0207] Recent constraints on smoothing out fluctuations in new energy sources:

[0208] ;

[0209] ;

[0210] ;

[0211] in, , These represent the output of the new energy generating units after smoothing out fluctuations during typical day d, time period t, and time period t-1, respectively. To smooth out fluctuations in the power output of new energy units during the typical daily d-period t-period, To mitigate power fluctuations in hybrid energy storage during a typical daily time period t, For maximum volatility, Rated capacity for new energy generating units.

[0212] The current power constraint for hybrid energy storage is:

[0213] ;

[0214] ;

[0215] ;

[0216] ;

[0217] ;

[0218] ;

[0219] ;

[0220] ;

[0221] in, , These represent the discharge and charging power of an energy-type battery during a typical daily period t (d). , These represent the discharge states of the battery and the power battery during a typical daily period t (1 is taken when discharging and 0 is taken when not discharging). , These represent the charging states of energy-type and power-type batteries during a typical day d-hour period t (1 is used when charging and 0 is used when not charging).

[0222] The current restrictions on frequency regulation capacity applications are as follows:

[0223] ;

[0224] ;

[0225] ;

[0226] ;

[0227] ;

[0228] The current energy constraint for hybrid energy storage is:

[0229] ;

[0230] ;

[0231] ;

[0232] ;

[0233] ;

[0234] ;

[0235] in, This represents the cumulative value of the frequency-modulated signal during a typical daily time period t. , These represent the energy of an energy-type battery during typical day d at time t and time t+1, respectively. , These represent the energy of a power battery during typical day d at time t and time t+1, respectively. , These represent the energy changes generated by energy-type batteries and power-type batteries participating in frequency modulation during a typical daily time period t (d). , These are the charging and discharging efficiency factors of energy-type batteries, respectively. , These are the charging and discharging efficiency factors for power batteries, respectively. , These are the minimum and maximum available SOC of an energy-type battery, respectively. , These are the minimum and maximum available SOC of a power battery, respectively.

[0236] The real-time stage constraints are structurally consistent with the day-ahead stage constraints but are distinguished in terms of scenario and time scale. In addition, they also include power deviation constraints and internal power coordination strategy constraints of hybrid energy storage.

[0237] The power deviation constraint is used to quantify the deviation of real-time operation from the day-ahead plan, including power deviation, frequency regulation capacity deviation and power smoothing deviation, and to reduce the impact of real-time deviation on plan execution through penalty terms.

[0238] The internal power coordination strategy constraint of the hybrid energy storage is used to achieve complementary responses between the energy-type battery and the power-type battery. It includes: constructing a dynamic power-bearing weight coefficient based on the state-of-charge deviation of the energy-type battery, and adjusting the proportion of real-time frequency regulation power of the energy-type battery during the real-time operation phase accordingly; controlling the power-type battery to bear the remaining real-time frequency regulation power not borne by the energy-type battery through the power balance relationship, thereby maintaining the state of charge and operational feasibility of the energy-type battery while ensuring the real-time response capability of the hybrid energy storage system.

[0239] The power deviation constraint is as follows:

[0240] ;

[0241] ;

[0242] ;

[0243] ;

[0244] ;

[0245] ;

[0246] in, , These represent energy-type batteries in a typical daily real-time scenario during the ω period. Discharge and charging power participating in the electricity market For hybrid energy storage in a typical daily real-time scenario during the ω period The declared frequency modulation capacity, For hybrid energy storage in a typical daily real-time scenario during the ω period Power to smooth out fluctuations.

[0247] The constraints of the internal power coordination strategy for energy storage are:

[0248] ;

[0249] ;

[0250] ;

[0251] In the formula, Indicates will Cut off to the interval [0,1]. For energy-type batteries in typical daily real-time scenarios during the ω period SOC deviation coefficient, The ideal SOC reference value for energy storage is taken as 0.5. The maximum allowable deviation is set to 0.4. As an exponential factor, it controls the steepness of the weight curve. For energy-type batteries in typical daily real-time scenarios during the ω period The actual frequency modulation power undertaken, For power batteries in typical daily real-time scenarios during the ω period The actual frequency modulation power undertaken.

[0252] S4: Combine the upper-level planning model and the lower-level operation model to form a two-level stochastic programming model that integrates planning and operation. By iteratively solving the two-level stochastic programming model, the optimal rated power and capacity configuration scheme of the hybrid energy storage system, as well as the corresponding day-ahead market reporting strategy and real-time power response strategy, are obtained.

[0253] This embodiment employs an iterative solution that alternates between upper-level planning and lower-level operation to achieve coordinated optimization of capacity planning and operation strategies.

[0254] Specifically, firstly, the upper-level planning model is solved. This model is responsible for determining the rated power and capacity configuration of the hybrid energy storage system. During the initial iterations or subsequent updates, the upper-level planning model receives cut-plane information from the lower-level operating model. These cut-planes, in the form of linear constraints, approximately express the feasible operating region and expected operating benefits of the lower-level operating model under different capacity configurations. Under the premise of satisfying these cut-plane constraints, the upper-level planning model, with the objective of maximizing net lifecycle benefits, solves for the current rated power and capacity configuration and passes it as fixed parameters to the lower level.

[0255] Next, the lower-level operational model is solved. After receiving the capacity configuration scheme from the upper layer, the lower-level operational model optimizes its strategies for the day-ahead and real-time market environments. To improve efficiency, this embodiment divides the multiple typical day-ahead and real-time scenarios included in the lower-level operational model into multiple independent sub-problems that can be solved in parallel. Each sub-problem independently calculates the day-ahead reporting strategy and real-time power response strategy for a specific scenario under the current capacity configuration, and calculates the corresponding operational benefits. Then, cut planes are generated and fed back. Based on the solution results of each independent sub-problem, cut planes are constructed that characterize the gradient of operational feasibility and benefit changes under the current capacity configuration. These cut planes are returned to the upper-level planning model to guide it in adjusting the capacity configuration towards higher operational benefits and better feasibility in the next iteration.

[0256] Finally, convergence is determined. The system repeats the above iterative process. As the number of iterations increases, the cutting plane constraints in the upper-level planning model are continuously improved, and the capacity configuration scheme gradually approaches the optimal solution. When the preset convergence condition is met, the iteration stops, and the solution obtained at this time is the capacity configuration scheme with the optimal life cycle benefit of the hybrid energy storage system and its corresponding operation strategy.

[0257] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A joint decision-making method for hybrid energy storage capacity configuration and operation in multiple electricity markets, characterized in that, A two-level stochastic programming model combining planning and operation is constructed to achieve coordinated optimization of capacity configuration and operation strategies for energy-type and power-type batteries in a hybrid energy storage system, while simultaneously considering the uncertainty of renewable energy output, fluctuations in electricity market prices, and real-time frequency regulation requirements. The specific implementation steps are as follows: S1: Acquire historical data on renewable energy output, market prices, and real-time frequency regulation signals; use cluster analysis to generate typical daily scenarios that characterize interannual operating characteristics, serving as a description of the uncertainty in the day-ahead phase; and construct real-time scenarios that characterize intraday renewable energy forecasting deviations and real-time frequency regulation demand under each typical daily scenario, thus constructing a two-stage random scenario of day-ahead and real-time. S2: Taking the maximization of the annualized net income of the hybrid energy storage system as the upper-level optimization objective, and combining the day-ahead and real-time two-stage stochastic scenarios, the upper-level planning model is used to determine the capacity configuration of energy-type batteries and power-type batteries. S3: Taking the maximization of the daily net operating income of the hybrid energy storage system as the lower-level optimization objective, based on the capacity configuration and the day-ahead-real-time two-stage random scenario, a lower-level operating model for the hybrid energy storage system to participate in the day-ahead-real-time market is constructed, and the corresponding constraints are determined. S4: Combine the upper-level planning model and the lower-level operation model to form a two-level stochastic programming model that integrates planning and operation. By iteratively solving the two-level stochastic programming model, the optimal rated power and capacity configuration scheme of the hybrid energy storage system, as well as the corresponding day-ahead market reporting strategy and real-time power response strategy, are obtained.

2. The joint decision-making method for hybrid energy storage capacity configuration and operation for multiple electricity markets as described in claim 1, characterized in that, In step S1, constructing the day-to-day and real-time two-stage stochastic scenario includes: Acquire historical data on renewable energy output, market prices, and real-time frequency regulation signals, where the frequency regulation signals are real-time power adjustment instructions issued by the power system dispatching agency or the power market platform; Based on the historical data of new energy output and market prices, cluster analysis is used to extract several typical daily scenarios representing different seasons and typical load characteristics from the annual operation data, which are used to describe the prediction uncertainty on an annual scale. Within each typical daily scenario, multiple real-time scenarios are constructed based on real-time frequency modulation signals and new energy output prediction deviations to simulate potential power deviations and frequency modulation demands during daily operation, thereby forming a two-stage stochastic scenario that can simultaneously characterize both day-ahead prediction uncertainties and real-time operational uncertainties.

3. The joint decision-making method for hybrid energy storage capacity configuration and operation for multiple electricity markets as described in claim 2, characterized in that, In step S2, the upper-level planning model aims to maximize the annualized net income of the hybrid energy storage system. The annualized net income is the difference between the annualized profit of the hybrid energy storage system and the annualized investment cost and annualized operation and maintenance cost of the hybrid energy storage system. The annualized profit of the hybrid energy storage system is calculated based on the operation results of the lower-level operation model under the day-ahead and real-time two-stage random scenarios, and the rated power and capacity configuration scheme of the power type battery and the energy type battery are determined accordingly.

4. The joint decision-making method for hybrid energy storage capacity configuration and operation for multiple electricity markets as described in claim 3, characterized in that, The objective function of the upper-level planning model is: ； In the formula, To optimize the goals of the upper level, The annualized profit of hybrid energy storage is calculated by the underlying operation model. For the investment cost of hybrid energy storage, For hybrid energy storage operation and maintenance costs; Hybrid energy storage investment costs The expression is: ； Hybrid energy storage operation and maintenance costs The expression is: ； In the formula, and These are the rated capacity and rated power of the energy-type battery, respectively. and These are the rated capacity and rated power of the power-type battery, respectively. , These refer to the investment cost per unit capacity and the investment cost per unit power of energy-type batteries, respectively. , These are the investment costs per unit capacity and per unit power of power-type batteries, respectively. and These are the unit capacity operation and maintenance costs and unit power operation and maintenance costs of energy-type batteries, respectively. , These are the unit capacity maintenance cost and unit power maintenance cost of power-type batteries, respectively. For the service life of the equipment, is the discount rate.

5. The joint decision-making method for hybrid energy storage capacity configuration and operation for multiple electricity markets as described in claim 4, characterized in that, The constraints of the upper-level planning model include boundary constraints on the power and capacity configuration of hybrid energy storage, investment budget constraints on hybrid energy storage, energy-power ratio constraints, and feasibility constraints that satisfy the lower-level operation model. The boundary constraints for the configuration of hybrid energy storage power and capacity mean that the rated power and rated capacity of both energy-type batteries and power-type batteries must be limited within the technically permissible range, including the minimum and maximum configurable values, in order to ensure the physical feasibility of the hybrid energy storage system. The hybrid energy storage investment budget constraint means that the sum of the capacity investment cost and power investment cost of energy type battery and power type battery shall not exceed the preset investment budget, so as to ensure that the planned scheme is economically feasible. The energy-power ratio constraint refers to the relationship between the capacity and power of energy-type batteries and power-type batteries, which must meet their technical characteristics. That is, energy-type batteries should have a high energy capacity-power ratio, while power-type batteries should have a high power density, so as to ensure that energy-type batteries and power-type batteries can undertake the corresponding energy support and rapid response tasks in the subsequent operation phase. The aforementioned feasibility constraints for satisfying the lower-level operating model refer to the fact that the rated capacity and rated power configuration of the energy-type battery and the power-type battery given in the upper-level planning model must fall within the capacity-power feasible region that ensures that the lower-level operating model has a feasible solution in both the day-to-day and real-time stochastic scenarios. If a candidate configuration results in no feasible solution for the lower-level operating model, it will be excluded from the upper-level planning model.

6. The joint decision-making method for hybrid energy storage capacity configuration and operation for multiple electricity markets according to claim 5, characterized in that, The boundary constraints for the hybrid energy storage power and capacity configuration are expressed as follows: ； ； In the formula, , These are the upper limits of the battery's power and capacity, respectively. , These are the upper limits for the power and capacity of a supercapacitor, respectively. The investment budget constraint for hybrid energy storage is expressed as follows: ； In the formula, This represents the investment cost function for hybrid energy storage. This is the upper limit of the investment budget; The energy-power ratio constraint is expressed as: ； ； In the formula, The minimum allowable energy-power ratio for a battery. The maximum energy-power ratio allowed by the battery. This represents the minimum allowable energy-to-power ratio for supercapacitors. This represents the maximum allowable energy-to-power ratio for a supercapacitor. The feasibility constraints that satisfy the lower-level operating model are expressed as follows: ； In the formula, To ensure that the lower-level operating model has a feasible capacity-power set under the given day-to-real-time two-stage stochastic scenario, a capacity-power set is required.

7. The joint decision-making method for hybrid energy storage capacity configuration and operation for multiple electricity markets as described in claim 6, characterized in that, In step S3, the lower-level operation model aims to maximize the daily net operating revenue of the hybrid energy storage system. This daily net operating revenue includes day-ahead electricity trading revenue, real-time electricity deviation costs, frequency regulation market revenue, and related operating costs. It is calculated based on a two-stage stochastic scenario (day-ahead and real-time), wherein: The objective function of the lower-level operating model is: ； In the formula, For the optimization goal of the lower level, This indicates that the expected value of a typical day d is taken according to probability. This means taking the expected value of the real-time scene ω according to probability under a typical day d. For typical day sets, This is a collection of real-time scenarios under typical daytime conditions. For the electricity market revenue of a typical day, For typical daily FM market revenue, For typical d, the volatility-smoothing return, For the electricity market revenue under a typical daily real-time scenario ω, For the frequency modulation market revenue in a typical daily real-time scenario ω, For the volatility mitigation return under a typical daily real-time scenario ω, The loss cost of a hybrid energy storage system under a typical daily real-time scenario ω; Typical daily electricity market revenue The expression is: ； In the formula, The day-ahead electricity market price for a typical day d period t. , These represent the discharge power and charging power of an energy-type battery participating in the electricity market during a typical daily period t (d). The current day's scheduling cycle, The time step is the day before yesterday; Typical daily FM market revenue The expression is: ； In the formula, , These are the unit price of frequency regulation capacity and the unit price of frequency regulation mileage for a typical day d-time period t. This refers to the daytime frequency regulation capacity declared by the hybrid energy storage system during a typical daytime period t. The average FM mileage is obtained from historical FM signal data; Typical daily volatility reduction return The expression is: ； In the formula, , These represent the discharge power and charging power of an energy-type battery during a typical daily period t, with the fluctuations smoothed out. , These represent the discharge power and charging power of a power-type battery during a typical daily period t, with the fluctuations smoothed out. The unit power price for leasing a hybrid energy storage system during a typical day's d-period; Electricity market revenue in a typical daily real-time scenario ω The expression is: ； In the formula, For a typical daily real-time scenario ω period Real-time electricity market price This is the penalty coefficient for deviations in the electricity market. , These represent the typical daily real-time scenarios (ω-time period) of energy-type batteries in the electric energy market. The declared positive and negative imbalance amounts of charging and discharging power. For real-time scheduling cycle, For real-time time steps; Frequency modulation market revenue in typical daily real-time scenarios. The expression is: ； In the formula, , These represent typical daily real-time scenarios with time intervals ω. The unit price of frequency regulation capacity and the unit price of frequency regulation mileage for hybrid energy storage systems. This is the frequency modulation market deviation penalty coefficient. , These represent typical daily real-time scenarios (ω-hours) for hybrid energy storage systems in the frequency regulation market. The declared frequency regulation capacity positive and negative imbalance; Fluctuation mitigation returns in a typical daily real-time scenario ω The expression is: ； In the formula, For a typical daily real-time scenario ω period The unit power price for rental To mitigate fluctuation deviation penalty coefficient, , The hybrid energy storage system smooths out fluctuations during the ω period of a typical daily real-time scenario. Positive and negative imbalance of charging and discharging power; Loss cost of hybrid energy storage system in typical daily real-time scenario ω The expression is: ； ； ； ； ； In the formula, This addresses the cost of energy-type batteries used to mitigate fluctuations and participate in the electricity market. This is to reduce the cost of power batteries used to mitigate fluctuations and participate in the electricity market. The cost of energy-type batteries participating in the frequency regulation market. The loss cost of power batteries participating in the frequency modulation market; For energy-type batteries during time periods The cost per unit of energy loss varies with the depth of discharge. For power batteries during time periods Average cost per unit of energy lost , The respective costs per unit frequency regulation mileage for energy-type batteries and power-type batteries are: For a typical daily real-time scenario ω period Frequency modulation mileage coefficient, , These represent energy-type batteries in a typical daily real-time scenario during the ω period. Charging and discharging power participating in the electricity market , These represent energy-type batteries in a typical daily real-time scenario during the ω period. Smooth out fluctuations in charging and discharging power. , These represent the power-type batteries in a typical daily real-time scenario during the ω period. Charging and discharging power participating in the electricity market , These represent the power-type batteries during typical daily real-time scenarios at the ω-hour interval. Smooth out fluctuations in charging and discharging power. For energy-type batteries during typical daily d periods The frequency modulation capacity recently declared For power batteries during typical daily d periods The frequency modulation capacity recently declared.

8. The joint decision-making method for hybrid energy storage capacity configuration and operation for multiple electricity markets according to claim 7, characterized in that, The constraints of the lower-level operation model include day-ahead stage constraints and real-time stage constraints. The day-ahead stage constraints include new energy power fluctuation smoothing constraints, frequency regulation capacity application constraints, hybrid energy storage power constraints, and hybrid energy storage energy constraints. The aforementioned new energy power fluctuation smoothing constraint is used to couple the energy storage smoothing power with the new energy output in order to limit the variation of grid-connected power in adjacent time periods. The frequency regulation capacity declaration constraint is used to determine the capacity that the hybrid energy storage system can declare to the frequency regulation market during the day-ahead phase, and requires that the declared capacity match the rated charge and discharge power capability of the hybrid energy storage system. The hybrid energy storage power constraint is used to limit the charging and discharging power of the energy type battery and the power type battery at any time to not exceed the rated power, and to prevent charging and discharging from occurring simultaneously through state mutual exclusion conditions. The hybrid energy storage energy constraint is used to apply upper and lower limits of the state of charge to the energy type battery and the power type battery respectively, and to ensure the energy continuity during the time-series operation through the energy balance equation, so as to avoid the energy exceeding the limit.

9. The joint decision-making method for hybrid energy storage capacity configuration and operation for multiple electricity markets as described in claim 8, characterized in that, The real-time stage constraints are structurally consistent with the day-ahead stage constraints but are distinguished in terms of scenario and time scale. In addition, they also include power deviation constraints and internal power coordination strategy constraints of hybrid energy storage. The power deviation constraint is used to quantify the deviation of real-time operation from the day-ahead plan, including power deviation, frequency regulation capacity deviation and power smoothing deviation, and to reduce the impact of real-time deviation on plan execution through penalty terms. The internal power coordination strategy constraint of the hybrid energy storage is used to achieve complementary responses between the energy-type battery and the power-type battery. It includes: constructing a dynamic power-bearing weight coefficient based on the state-of-charge deviation of the energy-type battery, and adjusting the proportion of real-time frequency regulation power of the energy-type battery during the real-time operation phase accordingly; controlling the power-type battery to bear the remaining real-time frequency regulation power not borne by the energy-type battery through the power balance relationship, thereby maintaining the state of charge and operational feasibility of the energy-type battery while ensuring the real-time response capability of the hybrid energy storage system.

10. The joint decision-making method for hybrid energy storage capacity configuration and operation for multiple electricity markets according to claim 9, characterized in that, In step S4, the two-level stochastic programming model is solved iteratively, including the following solution process: The upper-level planning model solves for the rated power and capacity configuration of the hybrid energy storage system, using the capacity configuration as the optimization variable. The operational feasibility information and operational benefit information returned by the lower-level operation model are transformed into cutting planes and added to the upper-level planning model. The capacity configuration scheme is gradually corrected by continuously updating the cutting planes. Given the capacity configuration output by the upper-level planning model, the lower-level operation model is solved. The typical daily scenario and its included real-time scenario are divided into multiple sub-problems that can be solved in parallel. The solution results of each sub-problem are used to generate a cutting plane that characterizes the daily reporting revenue and real-time response revenue, and then returned to the main problem of the lower-level operation model. The upper-level planning model and the lower-level operation model interact iteratively. The upper-level planning model updates the capacity configuration based on the feedback from the lower-level operation model, and the lower-level operation model re-solves the operation strategy based on the updated capacity configuration. Repeat the above solution process until the objective functions of the upper-level planning model and the lower-level operation model converge or the addition of a cutting plane no longer changes the solution results. This will give you the capacity configuration scheme with the optimal life cycle benefit of the hybrid energy storage system and its corresponding operation strategy.

Images