Collaborative peak-shaving methods, devices, equipment, and media based on heterogeneous energy storage systems

Abstract

This application relates to the field of power system dispatching technology, and particularly to a method, device, equipment, and medium for coordinated peak shaving based on heterogeneous energy storage systems. The method includes: obtaining the feasible region constraints of the heterogeneous energy storage system; constructing corresponding objective functions with total operating cost, total carbon emissions, and power system frequency stability indicators as targets; performing multi-objective optimization on decision variables based on the objective functions and feasible region constraints to generate a Pareto front solution set; determining the optimal dispatch scheme based on the value of each solution in the Pareto front solution set on each objective function; and analytically obtaining the output commands of the flywheel energy storage system, battery energy storage system, and molten salt thermal energy storage system at the target time based on the optimal dispatch scheme, so as to control the coordinated peak shaving of the heterogeneous energy storage system according to the output commands. This solves the problem in related technologies where a single type of energy storage cannot achieve full coverage of net load fluctuations across multiple time scales, leading to difficulties in simultaneously addressing high-frequency disturbances and long-term peak shaving.

Description

Technical Field

[0001] This application relates to the field of power system dispatching technology, and in particular to a collaborative peak-shaving method, device, equipment and medium based on heterogeneous energy storage system. Background Technology

[0002] With the large-scale grid connection of renewable energy, the net load fluctuation of the power system has intensified, and the decline in the proportion of synchronous generators has led to a weakening of system inertia, posing a severe challenge to frequency stability. Although thermal power units are the main peak-shaving resource, their mechanical ramping constraints are strict, making it difficult to respond quickly to power disturbances on the order of seconds to minutes. Frequent and large-scale changes in operating conditions will also increase coal consumption and carbon emissions.

[0003] Among related technologies, single energy storage technologies cannot simultaneously meet the regulation needs of multiple time scales: power-type energy storage has a fast response but low energy density, energy-type energy storage is prone to aging and difficult to support long-term peak shaving, and large-capacity thermal storage has a slow response, which restricts the grid's effective mitigation of net load fluctuations. Summary of the Invention

[0004] This application provides a collaborative peak-shaving method based on heterogeneous energy storage systems to solve the problems in related technologies where a single type of energy storage cannot achieve full coverage of net load fluctuations across multiple time scales, resulting in difficulties in balancing high-frequency disturbances and long-term peak shaving.

[0005] The first aspect of this application provides a collaborative peak-shaving method based on a heterogeneous energy storage system, comprising the following steps: obtaining power system load forecasting results and wind power output forecasting results for a scheduling period; generating a net load time series based on the load forecasting results and wind power output forecasting results; determining thermal power ramping constraints based on the net load time series; determining the feasible region constraints of the heterogeneous energy storage system based on the physical operation constraints of the heterogeneous energy storage system and the thermal power ramping constraints; generating decision variables based on the energy storage output of the heterogeneous energy storage system at different scales corresponding to the target time, wherein the heterogeneous energy storage system includes... The system comprises flywheel energy storage, battery energy storage, and molten salt thermal energy storage. Objective functions are constructed based on total operating cost, total carbon emissions, and power system frequency stability. Multi-objective optimization of decision variables is performed according to the objective functions and feasible region constraints to generate a Pareto front solution set. The optimal scheduling scheme is determined based on the value of each solution in the Pareto front solution set on each objective function. The output commands of the flywheel energy storage, battery energy storage, and molten salt thermal energy storage systems at the target time are analytically obtained from the optimal scheduling scheme, enabling coordinated peak shaving of the heterogeneous energy storage systems based on these output commands.

[0006] Optionally, with the total operating cost as the objective, the corresponding objective function can be constructed as follows: ; in, This is the coal consumption coefficient. Let i be the unit power operation and maintenance factor of energy storage technology. The stored energy i outputs power at time t. Energy storage technologies include flywheel energy storage, battery energy storage, and molten salt thermal energy storage. Let t be the output of the thermal power unit at time t, and T be the dispatching cycle.

[0007] Optionally, with total carbon emissions as the target, the corresponding objective function can be constructed as follows: ; in, Let t be the output of the thermal power unit. This is the carbon emission intensity coefficient. is the carbon emission correction factor, and T is the scheduling period.

[0008] Optionally, with the system frequency stability index as the objective, the corresponding objective function can be constructed as follows: ; in, ; Let T be the variance of thermal power output, and T be the dispatch period. Let t be the output of the thermal power unit. This is the average net load. This represents the maximum output of thermal power. This represents the minimum output of thermal power. The output of the thermal power unit at time t-1 These are mean deviation weight, range weight, smoothing weight, and dynamic weight, respectively. , For cross-scale factors, for t The flywheel energy storage system outputs power at all times. for t The battery energy storage system is always providing power. for t The molten salt thermal storage system is constantly operating at full capacity. For energy storage systems i Response coefficient.

[0009] Optionally, the optimal scheduling scheme is determined based on the values ​​of each solution in the Pareto front solution set on each objective function, including: normalizing the values ​​of each solution in the Pareto front solution set on each objective function to generate processing results; calculating the first distance to the positive ideal solution and the second distance to the negative ideal solution for each solution in the processing results, where the positive ideal solution is the minimum point of each objective after normalization, and the negative ideal solution is the maximum point of each objective after normalization; calculating the relative fit of each solution based on the first distance and the second distance, and selecting the solution with the largest relative fit as the optimal scheduling scheme.

[0010] Optionally, after controlling the heterogeneous energy storage system to coordinate peak shaving according to the output command, the process includes: identifying the energy storage output sequence corresponding to the output command; calculating the equivalent inertial constant based on the energy storage output sequence; solving the frequency deviation variation curve over time based on the equivalent inertial constant; determining the frequency drop depth and recovery time of the heterogeneous energy storage system after disturbance based on the variation curve; if the frequency drop depth is greater than or equal to a preset threshold and the recovery time is greater than or equal to a preset time, adjusting the weight coefficient of the frequency stability index, and re-performing multi-objective optimization based on the adjusted objective functions until the frequency drop depth is less than the preset threshold and the recovery time is less than the preset time.

[0011] A second aspect of this application provides a collaborative peak-shaving device based on a heterogeneous energy storage system, comprising: an acquisition module for acquiring power system load forecast results and wind power output forecast results for a scheduling period; a first generation module for generating a net load time series based on the load forecast results and wind power output forecast results, determining thermal power ramping constraints based on the net load time series, and determining the feasible region constraints of the heterogeneous energy storage system based on the physical operation constraints of the heterogeneous energy storage system and the thermal power ramping constraints; and a second generation module for generating decision variables based on the energy storage output of the heterogeneous energy storage system at different scales corresponding to the target time, wherein the heterogeneous energy storage system... The system includes a flywheel energy storage system, a battery energy storage system, and a molten salt thermal energy storage system. The processing module constructs corresponding objective functions based on total operating cost, total carbon emissions, and power system frequency stability indicators. It then performs multi-objective optimization on decision variables based on the objective functions and feasible region constraints, generating a Pareto front solution set. The control module determines the optimal scheduling scheme based on the values ​​of each solution in the Pareto front solution set on each objective function. It then analyzes the optimal scheduling scheme to obtain the output commands of the flywheel energy storage system, battery energy storage system, and molten salt thermal energy storage system at the target time, and controls the heterogeneous energy storage systems to coordinate peak shaving according to the output commands.

[0012] A third aspect of this application provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the program to perform the collaborative peak shaving method based on a heterogeneous energy storage system as described in the above embodiments.

[0013] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which is executed by a processor to perform the collaborative peak shaving method based on a heterogeneous energy storage system as described in the above embodiments.

[0014] The fifth aspect of this application provides a computer program product, including a computer program or instructions, which, when executed, implement the collaborative peak-shaving method based on a heterogeneous energy storage system as described in the above embodiments.

[0015] Therefore, this application has at least the following beneficial effects: This application's embodiments can generate net load time series by acquiring load and wind power forecasts, and then integrate thermal power ramping constraints and energy storage physical constraints into a feasible region. The tiered output of three types of heterogeneous energy storage—flywheel, battery, and molten salt thermal storage—is used as decision variables, corresponding to high-frequency disturbances, mid-frequency smoothing, and long-term peak shaving, respectively. A multi-objective function is constructed with total operating cost, total carbon emissions, and system frequency stability as objectives. The Pareto front solution set is generated by optimizing the decision variables using the objective function and feasible region. Based on the value of each solution in the Pareto front solution set on each objective function, the optimal scheduling scheme is determined, and the output commands of the three types of energy storage at each moment are analytically obtained. This achieves cross-scale coordination of heterogeneous energy storage on time scales from seconds to hours. It ensures the physical safety of thermal power units through linearized ramping constraints and quantifies dynamic frequency response capabilities at the static optimization level using dispatchable frequency stability indicators. Thus, without relying on subjective weights, it objectively balances economy, environmental protection, and frequency robustness, significantly improving the peak shaving capability and frequency stability of high-proportion new energy power systems, while reducing operating costs and carbon emissions.

[0016] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0017] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart of a collaborative peak-shaving method based on a heterogeneous energy storage system provided in an embodiment of this application; Figure 2 This is a flowchart illustrating the system frequency stability index analysis provided in the embodiments of this application; Figure 3 This is a system load and wind power output diagram provided according to an embodiment of this application; Figure 4 The following are system frequency curves for different energy storage schemes provided according to embodiments of this application; Figure 5 The diagram shows the total output of thermal power plants according to different energy storage schemes provided in the embodiments of this application; Figure 6 This is a total cost diagram of different energy storage solutions provided according to the embodiments of this application; Figure 7 A total carbon emission map of different energy storage schemes provided according to embodiments of this application; Figure 8 This is a block diagram of a collaborative peak-shaving device based on a heterogeneous energy storage system provided according to an embodiment of this application; Figure 9 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application. Detailed Implementation

[0018] The embodiments of this application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0019] Currently, renewable energy sources, such as wind and solar power, account for a significantly increased proportion of installed capacity in the power system. However, these energy sources have inherent randomness, volatility, and intermittency, resulting in a more pronounced sawtooth pattern in the system load curve.

[0020] Traditional power systems rely on the rotating rotors of synchronous generators to provide rotational inertia, while renewable energy equipment, connected via power electronic converters, lacks physical rotational inertia. This results in extremely rapid frequency drops and weak recovery capabilities when facing sudden load fluctuations or fault disturbances, exposing the system to a severe low rotational inertia crisis and frequency stability risks. Thermal power units, currently the primary source of flexibility in the power system, bear the main responsibility for peak shaving and valley filling. However, thermal power units are subject to strict ramp-up rate constraints due to their mechanical and thermal stresses and combustion efficiency. When the net load (total load minus renewable energy output) changes drastically, relying solely on the regulation of thermal power units not only fails to meet real-time power balance requirements on a second-to-minute basis but also leads to frequent large-scale variable-output operation. This drastic change in operating conditions leads to increased coal consumption, accelerated unit wear, and consequently, significantly increased power generation costs and carbon emissions.

[0021] To address these contradictions, energy storage systems have been introduced into peak-shaving systems. However, most current research tends to focus on the application of single types of energy storage, which is insufficient to meet the scheduling needs across multiple time scales. Specifically, power-type energy storage (such as flywheel energy storage), while offering fast response and long cycle life, has low energy density and cannot provide long-term peak-shaving support; energy-type energy storage (such as battery energy storage) has moderate energy density, but its lifespan deteriorates rapidly under high-frequency charge-discharge cycles and it struggles to meet peak-shaving demands spanning hours; and large-capacity energy storage (such as molten salt thermal energy storage systems) has enormous and stable capacity, but its response characteristics are slow. Scientifically coordinating these heterogeneous energy storage systems with vastly different physical characteristics and time scales ranging from seconds to hours is a key scientific challenge for achieving stable grid operation.

[0022] Most existing scheduling models are based on quasi-steady-state power balance and lack a quantitative description of transient frequency stability. Typically, frequency stability analysis requires time-domain simulations using complex systems of differential-algebraic equations, which seriously conflicts with the computational efficiency of linear / quadratic programming models at the scheduling level.

[0023] Therefore, this application proposes a collaborative peak-shaving method, device, equipment, and medium based on heterogeneous energy storage systems. It generates a net load time series by acquiring load and wind power forecasts, then integrates thermal power ramping constraints and energy storage physical constraints into a feasible region. The application uses the stratified output of three types of heterogeneous energy storage—flywheel, battery, and molten salt thermal energy storage—as decision variables, corresponding to high-frequency disturbances, mid-frequency smoothing, and long-term peak shaving, respectively. A multi-objective function is constructed with total operating cost, total carbon emissions, and system frequency stability as objectives. The Pareto front solution set is generated by optimizing the decision variables using the objective function and feasible region. Finally, the Pareto front solution set is used to... By taking the value of each solution on each objective function, the optimal scheduling scheme is determined, and the output commands of the three types of energy storage at each time are obtained analytically. This realizes cross-scale coordination of heterogeneous energy storage on time scales from seconds to hours. It not only ensures the physical safety of thermal power units through linearized ramp constraints, but also quantifies the dynamic frequency response capability at the static optimization level by using dispatchable frequency stability indicators. Thus, without relying on subjective weights, it objectively balances economy, environmental protection and frequency robustness, significantly improves the peak-shaving capability and frequency stability of high-proportion new energy power systems, and reduces operating costs and carbon emissions.

[0024] The following description, with reference to the accompanying drawings, describes a collaborative peak-shaving method, apparatus, device, and medium based on heterogeneous energy storage systems according to embodiments of this application.

[0025] Specifically, Figure 1 This is a flowchart illustrating a collaborative peak-shaving method based on a heterogeneous energy storage system provided in an embodiment of this application.

[0026] like Figure 1 As shown, the collaborative peak-shaving method based on heterogeneous energy storage systems includes the following steps: In step S101, the power system load forecast results and wind power output forecast results for the scheduling cycle are obtained. It is understood that the embodiments of this application can obtain load forecasting and wind power output forecasting results for the scheduling cycle, providing an accurate data foundation for subsequent net load calculation, thermal power ramping constraint conversion and heterogeneous energy storage collaborative optimization, thereby ensuring that the generated scheduling scheme can truly reflect the actual operating boundary of the power grid, and improving the reliability and engineering practicality of cross-scale peak shaving and frequency stability optimization from the source.

[0027] It should be noted that the dispatch cycle usually refers to a future period of time, which can be 24 hours (day-ahead dispatch), without specific limitations. Therefore, the power system load forecast result can be an estimate of the total power consumption in the next 24 hours. The load changes with time (e.g., high during the day and low at night, with morning and evening peaks). The forecast value is usually derived based on historical data, weather, holidays, and other factors. The wind power output forecast result can be an estimate of the power generation of wind farms in the next 24 hours. Wind power has randomness and volatility, and the forecast value may change drastically over time (due to wind speed).

[0028] In step S102, a net load time series is generated based on the load forecast results and wind power output forecast results. The thermal power ramping constraint is determined based on the net load time series. The feasible region constraint of the heterogeneous energy storage system is determined based on the physical operation constraint of the heterogeneous energy storage system and the thermal power ramping constraint. It is understood that the embodiments of this application can convert load forecasting and wind power output forecasting into net load time series, and quantitatively construct the ramping constraints of thermal power units based on this. Then, it can be integrated with physical constraints such as power limits and capacity limits of energy storage systems to form a complete feasible domain of heterogeneous energy storage systems. This feasible domain can eliminate all infeasible schemes that violate the mechanical ramping limits of thermal power units or the safety boundaries of energy storage in the early stage of scheduling optimization, ensuring that the subsequent optimization process is carried out only within the physically achievable range. This significantly improves the engineering feasibility and safety of the scheduling scheme and avoids the risk of command failure due to constraint omissions or approximate processing.

[0029] It should be noted that the actual load demand of the power system already includes a portion of the power provided by wind power. Therefore, wind power output needs to be deducted from the total load to obtain the net power demand that thermal power units and energy storage systems must share. The generation of the net load time series allows subsequent scheduling to focus on the target curve that controllable equipment needs to track, while the degree of fluctuation of the net load over time and the difference between adjacent times directly reflect the system's requirement for the regulation rate.

[0030] After obtaining the net load time series, the determination of the thermal power ramping constraint utilizes the change in net load and the mechanical limits of the thermal power unit itself. Due to the limitations imposed by thermal stress and combustion stability, the output of a thermal power unit cannot vary beyond a safe threshold between adjacent time points. This threshold is obtained by multiplying the unit's rated power by the maximum ramp rate, and is called the maximum ramping capacity. To transform this constraint into a limitation on energy storage output, the power balance relationship is needed: thermal power output equals net load minus the sum of all energy storage output. Therefore, the change in thermal power output between adjacent time points can be decomposed into the change in net load minus the change in total energy storage output. Requiring the absolute value of the thermal power output change not to exceed the maximum ramping capacity is equivalent to applying two inequalities to the change in total energy storage output between adjacent time points: one restricting a rapid decrease in total energy storage output (corresponding to an excessively steep increase in thermal power load), and the other restricting a rapid increase in total energy storage output (corresponding to an excessively steep decrease in thermal power load). The constant terms in both inequalities are calculated from the maximum ramping capacity and the net load change, thus realizing a linearized expression of the thermal power ramping constraint with respect to the energy storage decision variables.

[0031] Combining the linearized thermal power ramping constraints with the physical operational constraints of the energy storage system itself constitutes the feasible region constraints of the heterogeneous energy storage system. The physical operational constraints of the heterogeneous energy storage system include: the charging and discharging power at any given time must not exceed its rated maximum power; the remaining energy storage capacity must not exceed the minimum and maximum allowable values; and the energy balance relationship where capacity changes with the integral of charging and discharging power. These three types of constraints collectively limit the range of output values ​​for flywheel energy storage, battery energy storage, and molten salt thermal energy storage at each time point, as well as their coupling relationships across time points. By combining these constraints with the thermal power ramping constraints, a multidimensional feasible region can be obtained, within which each energy storage output scheme simultaneously satisfies the safe ramping boundary of the thermal power unit and the equipment safety boundaries of all energy storage devices.

[0032] Specifically, such as Figure 2 As shown, Figure 2 This is a flowchart for analyzing the system frequency stability index.

[0033] (1) Power system operation and power balance model.

[0034] First, the core task of the power system is to maintain a real-time balance between power generation and power load. However, due to the random fluctuations in wind power output, it cannot be directly incorporated into dispatch as a stable power source. Therefore, a net load model is introduced: ; in, for t Net load at all times; for t Constant load; for tWind power output at any given time, all in MW.

[0035] The model subtracts the wind power output forecast from the load forecast to obtain the power demand that must be shared by controllable units (thermal power) and energy storage systems after deducting the contribution of wind power. This is the net load time series. The fluctuation of the net load over time reflects the target curve that thermal power and energy storage need to track.

[0036] Secondly, given the known net load, the power balance model further refines the power allocation relationship between thermal power units and three types of heterogeneous energy storage (FES (Flywheel Energy Storage), BESS (Battery Energy Storage System), and TES (Thermal Energy Storage)). The power balance model ensures that at any given time... t The output of thermal power units It equals the net load minus the sum of the total output of the three types of heterogeneous energy storage (flywheel energy storage, battery energy storage, and molten salt thermal energy storage), specifically: ; in, for t The thermal power unit output at all times; Energy storage technologies, including FES, BESS, and TES. for t Energy storage at all times i contribute, for t Net load at all times.

[0037] The equation stipulates that at any given moment, the output of a thermal power unit is equal to the net load minus the algebraic sum of all energy storage outputs. The energy storage output can be positive (discharging, injecting power into the grid) or negative (charging, absorbing power from the grid). When the net load is higher than the economic output of thermal power or a rapid response is required, energy storage discharges to fill the gap; when the net load is lower than the minimum output of thermal power or there is a surplus of new energy sources, energy storage charges to absorb the surplus.

[0038] (2) Linearization construction of hard constraints for climbing slope of thermal power units.

[0039] The hard constraint on the ramp-up capability of thermal power units is crucial to ensuring the mechanical safety of the equipment. The absolute value of the change in thermal power output between adjacent moments cannot exceed the maximum ramp-up capability, i.e., the change in thermal power output: ; in, This represents the maximum ramp rate for thermal power plants. for t The thermal power unit output at all times. fort -1 moment thermal power unit output This refers to the rated power of the thermal power unit.

[0040] To incorporate this constraint into an optimization model with energy storage output as the decision variable, a linear transformation using the power balance equation is required. The power balance equation is as follows: ; in, for t Net load at all times for t Net load at time -1 for t Energy storage at all times i The total output is the sum of the output, for t -1 hour energy storage i The total output is the sum of the output, This represents the maximum ramp rate for thermal power plants. This refers to the rated power of the thermal power unit.

[0041] After simplification, the absolute value represents the change in net load minus the change in total energy storage output. Removing the absolute value sign yields inequalities in two directions: When thermal power output increases, the corresponding ramp-up limit constraint is: ; When the output of thermal power plants decreases, the corresponding lower limit constraint for ramping is: ; in, for t Energy storage at all times i The total output is the sum of the output, for t -1 hour energy storage i The total output is the sum of the output, for t Energy storage at all times i Contribute,, Energy storage technology, including , This represents the maximum ramp rate for thermal power plants. Rated power of thermal power units for t Net load at all times for t Net load at time -1.

[0042] The left side of these two linear inequalities represents the difference in total energy storage output at adjacent times (in reverse order), while the right side is composed of the maximum ramping capacity of thermal power plants and the change in net load. Thus, the nonlinear absolute value constraint that originally directly affected the output of thermal power plants is precisely transformed into a linear constraint on the energy storage decision variables.

[0043] (3) Physical operational constraints of heterogeneous energy storage systems.

[0044] The physical operating constraints of energy storage systems aim to ensure that the charging and discharging behavior of energy storage in optimized scheduling schemes does not exceed the physical limits of the equipment itself, thereby guaranteeing the engineering feasibility of the scheme. Power limit constraints stipulate that the charging and discharging power of an energy storage system at any given time cannot exceed its rated maximum power. This is because energy storage devices (whether flywheels, batteries, or molten salt thermal storage) are limited by the rated capacity of their converters, motors, or heat exchangers; exceeding the maximum power will lead to overcurrent, overheating, or mechanical damage.

[0045] Among them, the power limit is: ; in, For energy storage i Maximum output.

[0046] Capacity constraints ensure that the remaining energy in energy storage is always within a safe range. The capacity of energy storage (i.e., the stored energy, measured in MWh) varies over time and cannot exceed its maximum design capacity because overcharging can cause safety risks, nor can it fall below the minimum allowable capacity because deep over-discharge can severely shorten its lifespan.

[0047] The capacity constraint is as follows: ; in, for t Real-time energy storage capacity , Energy storage technology i Maximum and minimum capacity.

[0048] The energy balance constraint establishes a link between the decision variable (power) and the state variable (capacity), used to describe the cumulative change in energy storage capacity with charge and discharge power. The energy balance constraint includes: ; in, For energy storage i The initial capacity is generally set to Half of For energy storage i Design capacity, For energy storage i Energy conversion efficiency, for t Energy storage at all times i contribute, T represents the scheduling time step and the scheduling period.

[0049] The remaining capacity at the current moment depends on the initial capacity minus the cumulative net charge and discharge energy (the equivalent value after efficiency conversion). Since this constraint involves the power accumulation at all moments of the entire cycle, it introduces a time-dependent coupling relationship, meaning that the storage cannot discharge without limit at any moment, but must ensure that there was sufficient charging energy stored beforehand; similarly, it cannot be charged beyond the upper limit for an extended period.

[0050] These three constraints work together to ensure that the output command of the energy storage device does not exceed the power limit at any time, while its remaining capacity is always within a safe range, and the capacity change is strictly consistent with the charging and discharging history.

[0051] In this embodiment of the application, determining the thermal power ramping constraint based on the net load time series includes: obtaining the output change, rated ramping power, and maximum ramping rate of the thermal power unit at adjacent times; calculating the maximum ramping capacity based on the maximum ramping rate and rated power; calculating the net load change corresponding to adjacent times in the net load time series; and determining the thermal power ramping constraint based on the output change of the thermal power unit at adjacent times, the net load change corresponding to adjacent times in the net load time series, the rated ramping power, and the maximum ramping rate.

[0052] It is understood that the embodiments of this application can transform the original absolute value constraint of the ramp rate into a linearized expression based on the net load time series by obtaining the parameters of the thermal power unit and the net load change. This makes the change in thermal power output no longer directly used as a decision variable, but indirectly limits the adjustment range of energy storage output through the maximum ramp capacity and the net load change. This avoids the feasibility risks brought about by traditional penalty functions or ex-post corrections, and ensures that the thermal power units in the generated scheduling scheme always operate within the safe range allowed by mechanical thermal stress and combustion efficiency, thereby significantly improving the engineering reliability and physical executability of the collaborative peak shaving method.

[0053] It should be noted that thermal power units have a physical ramp rate limitation, meaning that the output change between adjacent moments cannot exceed a safety threshold, which is obtained by multiplying the unit's rated power by the maximum ramp rate. This constraint aims to prevent excessive thermal stress or combustion instability caused by rapid load changes. However, in dispatch optimization, thermal power output is not an independent decision variable but is implicitly determined by the power balance equation (net load minus total energy storage output). Therefore, thermal power output cannot be directly constrained; instead, this constraint needs to be transformed into a limitation on energy storage output.

[0054] Specifically, such as Figure 2As shown, given the net load time series, the change in net load between adjacent times can be calculated. The change in net load is determined by both the change in thermal power output and the change in total energy storage output: the change in net load equals the change in thermal power output plus the change in total energy storage output (considering direction, in reality, the change in thermal power output equals the change in net load minus the change in total energy storage output). Therefore, the change in thermal power output can be expressed as the difference between the change in net load and the change in total energy storage output. Substituting this expression into the absolute value inequality for thermal power load ramping, and rearranging, yields two independent linear inequalities: one restricts the total energy storage output from decreasing too much when thermal power load increases too rapidly, and the other restricts the total energy storage output from increasing too much when thermal power load decreases too rapidly.

[0055] In step S103, decision variables are generated based on the energy storage output of the heterogeneous energy storage system at different scales at the target time. The heterogeneous energy storage system includes a flywheel energy storage system, a battery energy storage system, and a molten salt thermal energy storage system. It is understood that the embodiments of this application can use the output of three types of heterogeneous energy storage systems—flywheel energy storage, battery energy storage, and molten salt thermal energy storage—at different time scales as independent decision variables, corresponding to high-frequency disturbance smoothing, mid-frequency smoothing adjustment, and long-term energy transfer, respectively. This fully preserves the differences in physical characteristics of each energy storage system, enabling the optimization algorithm to automatically allocate the charging and discharging power of different energy storage systems according to the spectral distribution of the net load. This achieves multi-time scale collaborative scheduling from the second level to the hour level, effectively overcoming the technical defects of single energy storage or simple hybrid modes that cannot simultaneously achieve rapid response and large-capacity support.

[0056] Specifically, such as Figure 2 As shown, the specific process of multi-scale heterogeneous energy storage variable modeling is as follows: Decision variables are defined based on physical characteristics. X , representing three different scales of energy storage output.

[0057] ; in, for t The FES output is constantly updated, with a high response frequency, capable of handling fluctuations ranging from seconds to minutes. for t BESS provides real-time output, offers a medium-scale response, and can handle fluctuations ranging from minutes to hours. for t TES operates at all times, with a relatively slow response but huge capacity, undertaking hourly and cross-time peak shaving.

[0058] It should be noted that this application utilizes the heterogeneity of physical characteristics to achieve efficient coverage of net load fluctuations. Specifically, the short-scale response layer (corresponding to FES) leverages second-level response characteristics to prioritize handling high-frequency random disturbances in the net load, smoothing instantaneous frequency drops. The mid-scale regulation layer (corresponding to BESS) compensates for minute-level load fluctuations, alleviating the frequency regulation pressure on thermal power units. The long-scale peak-shaving layer (corresponding to TES) utilizes its enormous energy capacity to perform large-scale energy transfer on an hourly scale, achieving peak shaving and valley filling, and optimizing the system's baseload distribution.

[0059] In step S104, objective functions are constructed with total operating cost, total carbon emissions, and power system frequency stability index as objectives, respectively. Based on the objective functions and feasible region constraints, multi-objective optimization is performed on the decision variables to generate Pareto front solution set.

[0060] It is understood that the embodiments of this application can construct three conflicting objective functions with total operating cost, total carbon emissions, and power system frequency stability index as optimization objectives, respectively. Under the constraints of the feasible region, multi-objective optimization is performed on the energy storage decision variables to generate a Pareto front solution set. Each solution in this solution set is a non-dominated solution, that is, improvement on any objective will inevitably lead to degradation of at least one other objective. This fully reveals the inherent trade-off between economy, environmental protection and frequency stability. Decision-makers can obtain a series of feasible dispatch schemes with different performances without subjective weighting, providing an objective and quantitative scientific basis for subsequent flexible decision-making based on actual grid preferences.

[0061] It should be noted that NSGA II treats energy storage output sequences as individuals and evolves the population within the feasible region through genetic operations such as crossover and mutation. It also employs non-dominated sorting and crowding distance comparison to retain non-dominated solutions in each generation that cannot be simultaneously superior to other solutions on all objectives. After multiple iterations, a Pareto front solution set is finally obtained, with each solution having its own strengths and weaknesses on the three objectives. Each solution in this set is non-dominated, meaning no other solution is better than it on all three objectives. This solves the problem of solution space exploration under multi-objective conflict, but does not provide a unique answer.

[0062] In multi-objective optimization problems, decision variables refer to adjustable unknowns. In this application, these are the output of flywheel energy storage, battery energy storage, and molten salt thermal energy storage at various times. The feasible region constraints include power limits, capacity limits, energy balance constraints of the heterogeneous energy storage system, and upper and lower limits for thermal power ramping obtained through net load time series linearization. These constraints collectively define the range of values ​​for the decision variables; any solution exceeding this range is physically infeasible.

[0063] There are three objective functions: total operating cost, total carbon emissions, and system frequency stability index. These three are conflicting and cannot be optimized simultaneously. The goal of multi-objective optimization is not to find a single optimal solution, but to find a set of Pareto optimal solutions. That is, among these solutions, it is impossible to improve one objective without worsening at least one objective. The set of all these solutions is called the Pareto front solution set.

[0064] Specifically, such as Figure 2 As shown, this application uses a non-dominated sorting genetic algorithm with an elite strategy to achieve this optimization process. The algorithm first randomly initializes a group of individuals that satisfy the feasible region constraint (each individual represents an energy storage output scheme). Then it iteratively evolves through the following operations: (1) Initialize the population: randomly generate a group of individuals, and the possible schemes with different outputs in the system; (2) Multi-objective evaluation: calculate f1, f2, f3 corresponding to each scheme; (3) Non-dominated sorting: compare all schemes, find the individual that no other scheme can completely surpass it, mark it L1, and find the optimal among the remaining individuals, mark it L2, and so on; (4) Crowding calculation: calculate the distance between each individual and its neighboring individuals. The larger the distance, the rarer the individual is, and the less likely it is to fall into a local optimum; (5) Evolutionary operation: generate a new generation through selection, crossover and mutation, and repeat the above (2)-(4); (6) When the set number of iterations is reached, the outermost L1 is the Pareto solution set.

[0065] In this embodiment of the application, the objective function is constructed with the total operating cost as the objective: ; in, This is the coal consumption coefficient. Let i be the unit power operation and maintenance factor of energy storage technology. The stored energy i outputs power at time t. Energy storage technologies include flywheel energy storage, battery energy storage, and molten salt thermal energy storage. Let t be the output of the thermal power unit at time t, and T be the dispatching cycle.

[0066] It is understood that the embodiments of this application can construct a total operating cost objective function that includes the fuel cost of thermal power units and the operation and maintenance costs of three types of heterogeneous energy storage. The thermal power output is implicitly determined by the power balance equation, and the energy storage operation and maintenance cost reflects the bidirectional losses of charging and discharging in absolute value form. In this way, the optimization process can automatically weigh the economic differences between coal consumption costs and the unit power operation and maintenance costs of different energy storage (flywheel, battery, molten salt thermal storage). The objective function guides the optimization algorithm to prioritize scheduling long-term energy storage with lower operation and maintenance coefficients to undertake base load transfer, and limit the overdraft of battery life by high-frequency charging and discharging, thereby minimizing the comprehensive operating cost of the entire cycle under the premise of meeting peak shaving requirements.

[0067] In this embodiment of the application, the objective function is constructed with total carbon emissions as the target: ; in, Let t be the output of the thermal power unit. This is the carbon emission intensity coefficient. is the carbon emission correction factor, and T is the scheduling period.

[0068] It is understood that the embodiments of this application can construct a total carbon emission objective function that includes the square term of thermal power output, so that the unit carbon emission intensity in the high load range increases nonlinearly with the increase of output. In this way, the optimization process can actively suppress the thermal power units from operating in the high emission area for a long time, and guide the dispatch scheme to give priority to the use of heterogeneous energy storage for peak shaving and load tracking. The quadratic function form more accurately reflects the carbon emission growth characteristics of thermal power units in actual operation than the linear model, and effectively reduces the total carbon emissions of the entire dispatch cycle while ensuring power balance.

[0069] In this embodiment of the application, the objective function is constructed with the system frequency stability index as the target: ; in, ; Let T be the variance of thermal power output, and T be the dispatch period. Let t be the output of the thermal power unit. This is the average net load. This represents the maximum output of thermal power. This represents the minimum output of thermal power. The output of the thermal power unit at time t-1 These are mean deviation weight, range weight, smoothing weight, and dynamic weight, respectively. , For cross-scale factors, for t The flywheel energy storage system outputs power at all times. for t The battery energy storage system is always providing power. for t The molten salt thermal storage system is constantly operating at full capacity. For energy storage systems i Response coefficient.

[0070] It is understood that the embodiments of this application can achieve a quantitative characterization of dynamic frequency response characteristics at the static scheduling level by constructing a system frequency stability index objective function that includes a penalty term for the variance, range, and rate of change of thermal power output, as well as a reward term for the rapid energy storage action. Among them, the variance, range, and rate of change penalty term effectively suppress the drastic fluctuations and frequent ramp-ups of thermal power unit output, reducing the risk of steady-state frequency deviation. Meanwhile, the cross-scale factor indirectly improves the virtual inertia and damping support capability of the system under disturbances by weighted rewarding the rapid energy storage actions such as flywheels and batteries. This objective function can be embedded into a multi-objective optimization model without complex time-domain simulation, guiding the scheduling scheme to spontaneously improve the frequency drop depth and recovery time while reducing operating costs and carbon emissions, thereby enhancing the frequency robustness of high-proportion renewable energy power systems.

[0071] In step S105, the optimal scheduling scheme is determined based on the value of each solution in the Pareto front solution set on each objective function. The output commands of the flywheel energy storage system, battery energy storage system and molten salt thermal energy storage system at the target time are obtained by analyzing the optimal scheduling scheme, so as to control the heterogeneous energy storage system to coordinate peak shaving according to the output commands.

[0072] It is understood that the embodiments of this application can normalize the values ​​of each solution in the Pareto front solution set in terms of total operating cost, total carbon emissions, and system frequency stability index, calculate the weighted distance and relative proximity of each solution to the positive ideal solution and the negative ideal solution, and automatically select the solution with the largest proximity as the optimal scheduling scheme. This avoids the decision bias caused by subjective weighted summation. Then, the output commands of flywheel energy storage, battery energy storage, and molten salt thermal energy storage at each moment are extracted from the optimal scheduling scheme to guide the heterogeneous energy storage system to participate in peak shaving in a coordinated manner according to the cross-scale hierarchical strategy. This achieves a seamless conversion from multi-objective optimization results to engineering executable commands, ensuring that the scheduling scheme has the best trade-off between economy, environmental protection and frequency stability while also being physically feasible.

[0073] In this embodiment, determining the optimal scheduling scheme based on the value of each solution in the Pareto front solution set on each objective function includes: normalizing the value of each solution in the Pareto front solution set on each objective function to generate a processing result; calculating the first distance to the positive ideal solution and the second distance to the negative ideal solution for each solution in the processing result, wherein the positive ideal solution is the minimum point of each objective function after normalization, and the negative ideal solution is the maximum point of each objective function after normalization; calculating the relative fit of each solution based on the first distance and the second distance, and selecting the solution with the largest relative fit as the optimal scheduling scheme.

[0074] It is understood that the embodiments of this application can eliminate the dimensional differences between operating costs, carbon emissions, and frequency stability indicators by normalizing the range of the three objective function values ​​of each solution in the Pareto front solution set, making the objectives comparable on the same numerical scale. Then, the weighted Euclidean distances from each normalized solution to the positive ideal point (the optimal value point of each objective) and the negative ideal point (the worst value point of each objective) are calculated respectively, and all solutions are sorted according to the relative fit (i.e., the ratio of the negative ideal distance to the sum of the positive and negative ideal distances). The solution with the largest fit is automatically selected as the optimal scheduling scheme. Based on the objective distribution of the Pareto front, there is no need for decision-makers to subjectively set the objective weights, avoiding the bias interference in the weighted summation method, and ensuring that the optimal scheme is geometrically closest to the ideal state and furthest from the worst state. Thus, a scientific and reproducible trade-off decision is achieved between economy, environmental protection, and frequency stability.

[0075] Specifically, such as Figure 2 As shown, since the Pareto front typically contains dozens of candidate solutions, further decision-making is required. TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) transforms multiple objectives into a single ranking index through the following steps. Since the three objectives have different dimensions (cost is measured in yuan, carbon emissions in kilograms, and frequency is dimensionless), they cannot be directly compared. Therefore, range normalization can be used to map each objective value to the [0,1] interval. Specifically: ; in, For normalized values, For the original target value, the first i The individual in the first j The original calculated values ​​for each target. For the target value, the th in the population j The value of the objective function, Let j be the minimum value of the objective function in the Pareto front solution set. Let be the maximum value of the j-th objective function in the Pareto front solution set.

[0076] A positive ideal solution takes the minimum value of each objective after normalization (i.e., all 0s), representing the virtual perfect solution with the lowest cost, least emissions, and most stable frequency; a negative ideal solution takes the maximum value of each objective after normalization (i.e., all 1s), representing the worst solution.

[0077] For each solution, calculate its distance to the positive ideal solution and its distance to the negative ideal solution, specifically: ; ; in, To be the ideal distance, The negative ideal distance For decision weighting coefficients, For a positive ideal value, It is a negative ideal value. This is a normalized value.

[0078] Combining the two distances, we obtain the closeness of each solution: ; in, For the first i The relative fit of each individual To be the ideal distance, It is a negative ideal distance.

[0079] Finally select The largest solution is taken as the optimal scheduling scheme.

[0080] In this embodiment of the application, after controlling the heterogeneous energy storage system to coordinate peak shaving according to the output command, the process includes: identifying the energy storage output sequence corresponding to the output command; calculating the equivalent inertial constant according to the energy storage output sequence; solving the frequency deviation change curve with time according to the equivalent inertial constant; determining the frequency drop depth and recovery time of the heterogeneous energy storage system after being disturbed according to the change curve; if the frequency drop depth is greater than or equal to a preset threshold and the recovery time is greater than or equal to a preset time, adjusting the weight coefficient of the frequency stability index, and re-performing multi-objective optimization according to the adjusted objective functions until the frequency drop depth is less than the preset threshold and the recovery time is less than the preset time.

[0081] The preset threshold and preset duration can be set according to actual needs without specific limitations.

[0082] It is understood that the embodiments of this application can identify the energy storage output sequence after the output command is executed, calculate the equivalent inertial constant and solve the frequency deviation time-domain curve to obtain the frequency drop depth and recovery time of the heterogeneous energy storage system after disturbance, and then compare it with the preset threshold. If the frequency drop depth or recovery time exceeds the standard, the weight coefficient in the frequency stability index is automatically adjusted and multi-objective optimization is performed again until the frequency safety requirements are met. This realizes iterative correction from static command to dynamic response, ensuring that the final scheduling scheme is not only theoretically optimal in terms of cost, carbon emissions and stability indicators, but also meets the frequency safety margin under actual physical disturbances. This significantly improves the adaptability and robustness of the method to the real operating environment of the power grid.

[0083] Specifically, such as Figure 2As shown, when a power deficit occurs in the power grid (such as generator tripping or a sudden load surge), the rate of change of the system frequency is inversely proportional to the total inertia constant. The inertia of a traditional synchronous generator is fixed, while new energy equipment (wind power, photovoltaic) connected to the grid via converters cannot provide physical inertia. Energy storage systems can simulate inertial behavior, i.e., virtual inertia, through rapid power response; however, the magnitude of this virtual inertia depends on the actual level of activity of the energy storage system in regulation.

[0084] Among them, the equivalent inertial response model is: ; ; in, For the system's equivalent inertia, The reference inertia represents the inherent synchronous inertial constant of the power grid. For energy storage i Inertial contribution gain coefficient For energy storage i Utilization factor The stored energy i outputs power at time t. Energy storage technologies include flywheel energy storage, battery energy storage, and molten salt thermal energy storage. Let be the maximum output power of the i-th type of energy storage system.

[0085] The more active the energy storage output (higher average power) in the scheduling scheme, the larger the utilization factor. After weighting by the gain coefficient, the higher the equivalent inertia of the system. The higher the equivalent inertia of the system, the slower the frequency decreases and the higher the minimum point.

[0086] After obtaining the equivalent inertia, substitute it into the time-domain equation of frequency deviation: ; in, For system frequency deviation, , The disturbance response coefficient is... For the damping ratio, The attenuation coefficient is... The natural oscillation frequency, For damped oscillation frequency, This is the equivalent inertia of the system.

[0087] This equation is a simplified second-order model of the frequency change after a power system is disturbed. The first term describes the frequency oscillation decay process. The larger the equivalent inertia, the smaller the oscillation amplitude. The second term describes the steady-state deviation recovery process of the frequency.

[0088] By solving this equation, a complete curve of frequency variation over time can be obtained, from which two key indicators can be extracted: frequency drop depth (the difference between the lowest point of the curve and the rated frequency) and recovery time (the time from the occurrence of the disturbance to the frequency recovering to the allowable range). These two indicators respectively reflect the effectiveness of the virtual inertia (damped oscillation) and primary frequency regulation capability (steady-state recovery) provided by energy storage.

[0089] In summary, to verify the effectiveness of the method of this invention, a thermal power unit with energy storage was used, and a typical daily calculation was conducted to implement and analyze the method. In the calculation, the power of the thermal power unit was set to 600 MW, and the energy storage capacities of FES, BESS, and TES in the heterogeneous energy storage system were set to 300, 800, and 3000 MWh, respectively. The time step was set to 60 min. The system and energy storage parameter settings in the calculation are shown in Tables 1 and 2, respectively. In addition, the load and wind power output curves are shown in... Figure 3 .

[0090] Table 1 System Basic Parameters

[0091] Table 2 Basic Parameters of Energy Storage System

[0092] This application employs a hierarchical management strategy using heterogeneous energy storage systems to suppress frequent peak-shaving by thermal power units, transforming the power output curve from drastic fluctuations to a smoother trend, thereby significantly reducing unit physical losses and carbon emissions. Simultaneously, the method utilizes the virtual inertial feedback of the heterogeneous energy storage system to effectively increase the frequency drop depth and eliminate oscillations, ensuring stable grid operation when facing the random impacts of wind power. Figure 4 It can be seen that, Figure 4 The system frequency curves for different energy storage schemes show that using the method described in this application reduces system fluctuations and improves system frequency stability. Simultaneously, from... Figure 5 It can be seen that, Figure 5 The thermal power output curves for different energy storage schemes are shown in the diagrams. Compared with the thermal power output curves for single energy storage schemes, the thermal power output curves produced by the method in this application are smoother. Figure 6 and Figure 7 Further evidence shows that using the method of this application results in a reduction in thermal power output, leading to lower total costs and lower total carbon emissions. Figure 6 A total cost diagram for different energy storage solutions. Figure 7 A graph showing the total carbon emissions of different energy storage solutions.

[0093] According to the collaborative peak-shaving method based on heterogeneous energy storage systems proposed in this application, a net load time series is generated by obtaining load and wind power forecasts. Then, the thermal power ramping constraint and energy storage physical constraint are integrated into a feasible region. The stratified output of three types of heterogeneous energy storage—flywheel, battery, and molten salt thermal storage—is used as decision variables, corresponding to high-frequency disturbances, mid-frequency smoothing, and long-term peak shaving, respectively. A multi-objective function is constructed with total operating cost, total carbon emissions, and system frequency stability as objectives. The Pareto front solution set is generated by optimizing the decision variables using the objective function and feasible region. Finally, based on each solution in the Pareto front solution set… By taking values ​​for each objective function, the optimal scheduling scheme is determined, and the output commands of the three types of energy storage at each moment are obtained analytically. This realizes cross-scale coordination of heterogeneous energy storage on time scales ranging from seconds to hours. It not only ensures the physical safety of thermal power units through linearized ramp constraints, but also quantifies the dynamic frequency response capability at the static optimization level by using dispatchable frequency stability indicators. Thus, without relying on subjective weights, it objectively balances economy, environmental protection and frequency robustness, significantly improves the peak-shaving capability and frequency stability of high-proportion renewable energy power systems, and reduces operating costs and carbon emissions.

[0094] Next, referring to the accompanying drawings, a collaborative peak-shaving device based on a heterogeneous energy storage system proposed in accordance with the embodiments of this application is described.

[0095] Figure 8 This is a block diagram of a collaborative peak-shaving device based on a heterogeneous energy storage system according to an embodiment of this application.

[0096] like Figure 8 As shown, the collaborative peak-shaving device 10 based on a heterogeneous energy storage system includes: an acquisition module 100, a first generation module 200, a second generation module 300, a processing module 400, and a control module 500.

[0097] The acquisition module 100 is used to acquire the power system load forecast results and wind power output forecast results for the scheduling cycle; the first generation module 200 is used to generate a net load time series based on the load forecast results and wind power output forecast results, determine the thermal power ramping constraints based on the net load time series, and determine the feasible region constraints of the heterogeneous energy storage system based on the physical operation constraints of the heterogeneous energy storage system and the thermal power ramping constraints; the second generation module 300 is used to generate decision variables based on the energy storage output of the heterogeneous energy storage system at different scales corresponding to the target time, wherein the heterogeneous energy storage system includes flywheel energy storage systems, battery energy storage systems, etc. The system integrates a molten salt thermal energy storage system; the processing module 400 is used to construct corresponding objective functions with total operating cost, total carbon emissions, and power system frequency stability index as objectives, respectively, and to perform multi-objective optimization on decision variables based on the objective functions and feasible region constraints to generate a Pareto front solution set; the control module 500 is used to determine the optimal scheduling scheme based on the value of each solution in the Pareto front solution set on each objective function, and to obtain the output commands of the flywheel energy storage system, battery energy storage system, and molten salt thermal energy storage system at the target time based on the optimal scheduling scheme, so as to control the heterogeneous energy storage system to coordinate peak shaving according to the output commands.

[0098] According to the collaborative peak-shaving device based on heterogeneous energy storage system proposed in this application, the net load time series is generated by acquiring load and wind power forecasts. Then, the thermal power ramping constraint and energy storage physical constraint are integrated into a feasible region. The stratified output of three types of heterogeneous energy storage—flywheel, battery, and molten salt thermal energy storage—is used as decision variables, corresponding to high-frequency disturbance, mid-frequency smoothing, and long-term peak shaving, respectively. A multi-objective function is constructed with total operating cost, total carbon emissions, and system frequency stability as objectives. The Pareto front solution set is generated by optimizing the decision variables using the objective function and feasible region. Then, based on each solution in the Pareto front solution set… By taking values ​​for each objective function, the optimal scheduling scheme is determined, and the output commands of the three types of energy storage at each moment are obtained analytically. This realizes cross-scale coordination of heterogeneous energy storage on time scales ranging from seconds to hours. It not only ensures the physical safety of thermal power units through linearized ramp constraints, but also quantifies the dynamic frequency response capability at the static optimization level by using dispatchable frequency stability indicators. Thus, without relying on subjective weights, it objectively balances economy, environmental protection and frequency robustness, significantly improves the peak-shaving capability and frequency stability of high-proportion renewable energy power systems, and reduces operating costs and carbon emissions.

[0099] Figure 9 A schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include: The memory 901, the processor 902, and the computer program stored on the memory 901 and capable of running on the processor 902.

[0100] When the processor 902 executes the program, it implements the collaborative peak shaving method based on heterogeneous energy storage system provided in the above embodiments.

[0101] Furthermore, electronic devices also include: Communication interface 903 is used for communication between memory 901 and processor 902.

[0102] The memory 901 is used to store computer programs that can run on the processor 902.

[0103] The memory 901 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0104] If the memory 901, processor 902, and communication interface 903 are implemented independently, then the communication interface 903, memory 901, and processor 902 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized into address buses, data buses, control buses, etc. For ease of representation, Figure 9 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0105] Optionally, in a specific implementation, if the memory 901, processor 902, and communication interface 903 are integrated on a single chip, then the memory 901, processor 902, and communication interface 903 can communicate with each other through an internal interface.

[0106] The processor 902 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0107] This application also provides a computer-readable storage medium storing a computer program or instructions thereon, which, when executed by a processor, implements the above-described collaborative peak-shaving method based on a heterogeneous energy storage system.

[0108] This application also provides a computer program product, including a computer program or instructions, which, when executed, implement the above-described collaborative peak shaving method based on a heterogeneous energy storage system.

[0109] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0110] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0111] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0112] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or more of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0113] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

Claims

1. A collaborative peak-shaving method based on heterogeneous energy storage systems, characterized in that, Includes the following steps: Obtain power system load forecasting results and wind power output forecasting results for the scheduling cycle; A net load time series is generated based on the load forecast results and the wind power output forecast results. The thermal power ramping constraint is determined based on the net load time series. The feasible region constraint of the heterogeneous energy storage system is determined based on the physical operation constraint of the heterogeneous energy storage system and the thermal power ramping constraint. Decision variables are generated based on the energy storage output of the heterogeneous energy storage system at different scales at the target time, wherein the heterogeneous energy storage system includes a flywheel energy storage system, a battery energy storage system, and a molten salt thermal energy storage system. Objective functions are constructed with total operating cost, total carbon emissions and power system frequency stability as objectives, respectively. Based on the objective functions and the feasible region constraints, multi-objective optimization is performed on the decision variables to generate Pareto front solution set. Based on the value of each solution in the Pareto front solution set on each objective function, the optimal scheduling scheme is determined. The output commands of the flywheel energy storage system, battery energy storage system and molten salt thermal energy storage system at the target time are obtained by parsing the optimal scheduling scheme, so as to control the heterogeneous energy storage system to coordinate peak shaving according to the output commands.

2. The collaborative peak-shaving method based on heterogeneous energy storage systems according to claim 1, characterized in that, With total operating cost as the objective, the corresponding objective function is constructed as follows: ； in, This is the coal consumption coefficient. Let i be the unit power operation and maintenance factor of energy storage technology. The stored energy i outputs power at time t. This refers to energy storage technologies, which include flywheel energy storage, battery energy storage, and molten salt thermal energy storage. Let t be the output of the thermal power unit at time t, and T be the dispatching cycle.

3. The collaborative peak-shaving method based on heterogeneous energy storage systems according to claim 2, characterized in that, With total carbon emissions as the target, the corresponding objective function is constructed as follows: ； in, Let t be the output of the thermal power unit. This is the carbon emission intensity coefficient. is the carbon emission correction factor, and T is the scheduling period.

4. The collaborative peak-shaving method based on heterogeneous energy storage systems according to claim 3, characterized in that, With the system frequency stability index as the objective, the corresponding objective function is constructed as follows: ； in, ; Let T be the variance of thermal power output, and T be the dispatch period. Let t be the output of the thermal power unit. This is the average net load. This represents the maximum output of thermal power. This represents the minimum output of thermal power. The output of the thermal power unit at time t-1 These are mean deviation weight, range weight, smoothing weight, and dynamic weight, respectively. , For cross-scale factors, for t The flywheel energy storage system outputs power at all times. for t The battery energy storage system is always providing power. for t The molten salt thermal storage system is constantly operating at full capacity. For energy storage systems i Response coefficient.

5. The collaborative peak-shaving method based on heterogeneous energy storage systems according to claim 1, characterized in that, The step of determining the optimal scheduling scheme based on the value of each solution in the Pareto front solution set on each objective function includes: The values ​​of each solution in the Pareto front solution set are normalized on each objective function to generate the processing result; Calculate the first distance to the positive ideal solution and the second distance to the negative ideal solution for each solution in the processing result. The positive ideal solution is the minimum point of each objective after normalization, and the negative ideal solution is the maximum point of each objective after normalization. The relative fit of each solution is calculated based on the first distance and the second distance, and the solution with the highest relative fit is selected as the optimal scheduling scheme.

6. The collaborative peak-shaving method based on heterogeneous energy storage systems according to claim 1, characterized in that, After controlling the heterogeneous energy storage system to coordinate peak shaving according to the output command, the following is included: Identify the energy storage output sequence corresponding to the output command; The equivalent inertial constant is calculated based on the energy storage output sequence. The frequency deviation variation curve over time is obtained based on the equivalent inertial constant. The frequency drop depth and recovery time of the heterogeneous energy storage system after disturbance are determined based on the variation curve. If the frequency drop depth is greater than or equal to a preset threshold and the recovery time is greater than or equal to a preset duration, the weight coefficient of the frequency stability index is adjusted, and multi-objective optimization is performed again according to the adjusted objective functions until the frequency drop depth is less than the preset threshold and the recovery time is less than the preset duration.

7. A collaborative peak-shaving device based on a heterogeneous energy storage system, characterized in that, include: The acquisition module is used to acquire the power system load forecast results and wind power output forecast results for the scheduling cycle; The first generation module is used to generate a net load time series based on the load forecast results and the wind power output forecast results, determine the thermal power ramping constraints based on the net load time series, and determine the feasible domain constraints of the heterogeneous energy storage system based on the physical operation constraints of the heterogeneous energy storage system and the thermal power ramping constraints. The second generation module is used to generate decision variables based on the energy storage output of the heterogeneous energy storage system at different scales at the target time. The heterogeneous energy storage system includes a flywheel energy storage system, a battery energy storage system, and a molten salt thermal energy storage system. The processing module is used to construct corresponding objective functions with total operating cost, total carbon emissions and power system frequency stability index as objectives, respectively, and to perform multi-objective optimization on decision variables based on the objective functions and the feasible region constraints to generate Pareto front solution set; The control module is used to determine the optimal scheduling scheme based on the value of each solution in the Pareto front solution set on each objective function, and to obtain the output commands of the flywheel energy storage system, battery energy storage system and molten salt thermal energy storage system at the target time based on the optimal scheduling scheme, so as to control the heterogeneous energy storage system to coordinate peak shaving according to the output commands.

8. An electronic device, characterized in that, include: A memory, a processor, and a computer program stored in the memory and executable on the processor, the processor executing the program to implement the collaborative peak-shaving method based on a heterogeneous energy storage system as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by the processor, they are used to implement the collaborative peak shaving method based on heterogeneous energy storage systems as described in any one of claims 1-6.

10. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed, they implement the collaborative peak-shaving method based on heterogeneous energy storage systems as described in any one of claims 1-6.

Images