Virtual power plant scheduling method and device oriented to power system and electronic equipment

Abstract

The invention provides a virtual power plant scheduling method and device oriented to a power system, and electronic equipment. The method comprises the following steps: obtaining prediction data, equipment parameters and market information of a virtual power plant in a predetermined scheduling period; constructing an economic dispatching optimization model of the virtual power plant; the economic dispatching optimization model is solved, candidate dispatching schemes in a dispatching period are generated, iterative optimization is carried out on the candidate dispatching schemes, and the iterative optimization comprises uniformly encoding time sequence values of various dispatching variables in the dispatching period into a dispatching variable set of a virtual power plant dispatching scheme; integrating a feasible solution repairing mechanism based on the physical constraint of the virtual power plant in the updating process; and when the iteration process meets a preset termination condition, taking the current optimal virtual power plant scheduling scheme as an optimal scheduling plan, generating a control instruction corresponding to the scheduling instruction, and executing the control instruction in a predetermined scheduling period. By adopting the technical scheme, the overall efficiency, reliability and economical efficiency of virtual power plant scheduling can be improved.

Description

Technical Field

[0001] This disclosure relates to the field of power dispatching technology, and in particular to a virtual power plant dispatching method, apparatus, and electronic equipment for power systems. Background Technology

[0002] With the development of energy transition and smart grids, Virtual Power Plants (VPPs), as an emerging power system optimization and dispatch solution, are gradually becoming an indispensable part of modern power networks. VPPs integrate various flexible adjustment resources such as distributed energy resources, energy storage devices, and adjustable loads, and interact with the power grid through information and communication technologies, thereby achieving centralized optimization control and diversified management of resources at the dispatch level. They not only enhance the flexibility and stability of the power system but also play a crucial role in promoting the consumption of new energy sources, optimizing power resource allocation, and improving system economics.

[0003] In the scheduling optimization of virtual power plants, a series of complex challenges are faced, including multi-objective optimization, handling of strong constraints, and system-wide coordinated scheduling. Existing traditional scheduling methods mostly rely on optimization algorithms such as linear programming, nonlinear programming, and dynamic programming. While these algorithms have achieved success in some simple scenarios, they often fail to meet practical needs when facing highly complex and dynamically changing power system environments. Summary of the Invention

[0004] In view of this, the present disclosure provides a virtual power plant dispatching method, apparatus and electronic equipment for power systems, which can improve the overall efficiency, reliability and economy of virtual power plant dispatching.

[0005] In a first aspect, embodiments of this disclosure provide a virtual power plant dispatching method for power systems, including: Obtain forecast data, equipment parameters, and market information for virtual power plants within a predetermined scheduling cycle; An economic dispatch optimization model for the virtual power plant is constructed. The economic dispatch optimization model uses the set values ​​of each adjustable resource in the equipment parameters at each time period as decision variables, the minimization of the comprehensive operating cost of the virtual power plant as the objective function, and power balance as a constraint. The economic dispatch optimization model is solved to generate candidate dispatch schemes within the dispatch period, and the candidate dispatch schemes are iteratively optimized. The iterative optimization includes uniformly encoding the time-series values ​​of various dispatch variables within the dispatch period into a set of dispatch variables for the virtual power plant dispatch scheme, and integrating a feasible solution repair mechanism based on the physical constraints of the virtual power plant during the update process. When the iteration process meets the preset termination condition, the current best virtual power plant scheduling scheme is taken as the optimal scheduling plan, the corresponding control instructions for the scheduling instructions are generated, and they are executed within the predetermined scheduling period.

[0006] Optionally, the forecast data includes load forecast, distributed generation output forecast and / or electricity price forecast; the equipment parameters include operating boundary parameters of distributed generation, energy storage devices and controllable loads; and the market information includes real-time electricity prices, peak-valley electricity prices and ancillary service prices in the electricity market. The overall operating cost is minimized as follows: ; The power balance: ; in, Indicates the scheduled scheduling period, Indicates in The cost of purchasing electricity at all times This indicates the operating cost of the power generation equipment within the virtual power plant. Indicates the charging and discharging cost of virtual power plant energy storage devices, This indicates that the power generation equipment in the virtual power plant is... Output power at any time Indicates that the energy storage device is in Charge and discharge power at any time Indicates in Total load demand at any given time This indicates the power purchased.

[0007] Optionally, solving the economic scheduling optimization model includes: Under the premise that the predetermined scheduling period and time resolution are determined, a unified scheduling variable expression structure is established for various types of equipment within the virtual power plant; Under the constraints of preset variable value range and equipment operation boundary, based on the scheduling variable expression structure, multiple initial candidate scheduling schemes for virtual power plants are randomly generated. Each candidate scheduling scheme is made feasible to ensure that it at least meets the upper and lower capacity limits of continuous equipment, the initial state of charge of energy storage equipment, and the minimum continuous operating time and minimum continuous downtime constraints of equipment with start-stop states. A multi-objective evaluation index system that considers both economic efficiency and operational performance is constructed for the virtual power plant, and a corresponding comprehensive evaluation index is formed through a multi-objective evaluation strategy. The comprehensive evaluation index, as well as the degree of violation of power balance constraints and equipment operation boundary constraints, are used as evaluation criteria. The multi-objective fitness value of each virtual power plant candidate scheduling scheme is calculated, and the fitness value is used as the basis for iterative improvement.

[0008] Optionally, the establishment of a unified scheduling variable expression structure for various types of equipment within the virtual power plant includes: encoding continuously adjustable quantities such as distributed power output, energy storage charging and discharging power, power exchange with the power system, and controllable load power level in a continuous numerical manner; encoding equipment with start-stop states in a discrete state sequence manner, so that the operating status and power settings of each equipment in each time period together constitute a complete virtual power plant scheduling scheme. The multi-objective evaluation indicators include at least two of the following: the comprehensive operating cost and / or comprehensive economic benefits of the virtual power plant during the dispatch cycle, the level of renewable energy output absorption, and the degree of deviation in user-side electricity comfort.

[0009] Optionally, the feasible solution repair mechanism includes at least two of the following repair mechanisms, which are executed in a preset order after each new candidate scheduling scheme is generated: The mechanism for repairing power balance in virtual power plants is as follows: For the current candidate scheduling scheme, the difference between the total power generation output, net energy storage output, controllable load adjustment amount and the sum of rigid load and baseline load of the virtual power plant is calculated in each scheduling period. This difference is taken as the power imbalance amount. According to the preset scheduling priority rules, the power exchange plan between the virtual power plant and the power system is adjusted first, and then the charging and discharging power of the energy storage device and / or the load level of the controllable load are adjusted. A repair mechanism for energy storage dynamics: During the scheduling period, the charging and discharging power of the energy storage devices in the current candidate scheduling scheme is accumulated in chronological order, and the trajectory of the state of charge change of each energy storage device is simulated and calculated. When the state of charge of a certain period exceeds the preset upper and lower limits or does not meet the predetermined beginning and end state requirements at the end of the scheduling period, the energy storage power sequence in the candidate scheduling scheme is modified by scaling, shifting or redistributing the charging and discharging power of the relevant period, so that the updated state of charge trajectory meets the capacity constraints and the beginning and end state constraints. Operation sequence repair mechanism for equipment with start-stop states: The start-stop sequences of generator sets, load devices or other equipment described by start-stop states in the current candidate scheduling scheme are detected. When there are operation segments that do not meet the minimum continuous running time and / or minimum continuous downtime constraints, the start-stop state sequence is locally adjusted by merging, extending or shortening adjacent start-stop segments so that the repaired start-stop sequence meets the corresponding running time constraints.

[0010] Optionally, in solving the economic dispatch optimization model of the virtual power plant, the candidate dispatch schemes for the virtual power plant and its internal equipment are iteratively optimized using the following evolutionary update mechanism: In each iteration round, when comparing and screening multiple candidate scheduling schemes for virtual power plants, the candidate scheduling schemes that satisfy the virtual power plant power balance constraints, distributed power output constraints, energy storage charge state constraints, and controllable load operation constraints are given priority. When two candidate scheduling schemes satisfy the above constraints, the candidate scheduling scheme with the better multi-objective fitness value is selected as the corresponding virtual power plant's historical best scheduling scheme and / or current global best scheduling scheme. When iteratively updating the scheduling variables of the equipment inside the virtual power plant, the scheduling variables representing the output of distributed power sources, the charging and discharging power of energy storage, the power exchange with the power system, and the continuous adjustment of the controllable load power level are adjusted using the numerical update rules of continuous space; the scheduling variables representing the switching states of generator sets, load devices, or other equipment with start-stop states are updated using the state flipping rules based on probability mapping. During the iterative evolution of the virtual power plant scheduling scheme, the weight parameters and guiding coefficients used to generate a new round of candidate scheduling schemes are linearly reduced based on the iteration process, the fitness distribution of each candidate scheduling scheme, and the constraint violation.

[0011] Optionally, during the iterative optimization of candidate scheduling schemes for virtual power plants, local search enhancement is periodically performed on the current optimal scheduling scheme for virtual power plants. This local search enhancement includes: Within a preset neighborhood, based on the current best scheduling scheme of the virtual power plant, a controlled-amplitude random disturbance is applied to the output curves, energy storage charging and discharging sequences and / or start-stop state sequences of some of the equipment to generate several new candidate scheduling schemes for the virtual power plant. The feasible solution repair mechanism is invoked for each of the new candidate scheduling schemes to correct the constraints on power balance, energy storage state of charge and equipment operating time, so as to obtain a repair scheme that meets the physical constraints of the virtual power plant, and the corresponding fitness value is calculated according to the multi-objective evaluation index. When any repaired new candidate scheduling scheme has a better multi-objective fitness value than the current best virtual power plant scheduling scheme, provided that all constraints are met, the new candidate scheduling scheme replaces the current best virtual power plant scheduling scheme and serves as the benchmark scheme to guide the evolution and update of other candidate scheduling schemes in subsequent iterations.

[0012] Optionally, the economic scheduling optimization model is solved using a particle swarm optimization algorithm. The particle swarm optimization algorithm encodes the temporal values ​​of various scheduling variables within the scheduling period into a set of scheduling variables for the virtual power plant scheduling scheme, and integrates a feasible solution repair mechanism based on the physical constraints of the virtual power plant during the particle position update process.

[0013] Secondly, embodiments of this disclosure provide a virtual power plant dispatching device for power systems, comprising: The acquisition unit is configured to acquire forecast data, equipment parameters, and market information of the virtual power plant within a predetermined scheduling period; The construction unit is configured to construct an economic dispatch optimization model for the virtual power plant. The economic dispatch optimization model uses the set values ​​of each adjustable resource in the equipment parameters at each time period as decision variables, the minimization of the overall operating cost of the virtual power plant as the objective function, and power balance as a constraint. The processing unit is configured to solve the economic scheduling optimization model, generate candidate scheduling schemes within the scheduling period, and iteratively optimize the candidate scheduling schemes. The iterative optimization includes uniformly encoding the time-series values ​​of various scheduling variables within the scheduling period into a set of scheduling variables for the virtual power plant scheduling scheme, and integrating a feasible solution repair mechanism based on the physical constraints of the virtual power plant during the update process. The scheduling unit is configured to take the current best virtual power plant scheduling scheme as the optimal scheduling plan when the iteration process meets the preset termination condition, generate the corresponding control instructions for the scheduling instructions, and execute them within the predetermined scheduling period.

[0014] Thirdly, embodiments of this disclosure provide an electronic device, including a memory and a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method as described in any of the foregoing embodiments.

[0015] Compared with the prior art, the technical solution of this application has the following advantages: The virtual power plant scheduling method for power systems provided in this disclosure acquires predicted data, equipment parameters, and market information of the virtual power plant to provide accurate input data for optimization calculations. Based on the constructed economic scheduling optimization model, scheduling variables of various equipment within the virtual power plant are uniformly encoded, and the time-series values ​​of each decision variable in the scheduling cycle are used as optimization objectives, thus forming a comprehensive and systematic optimization framework. This framework can reasonably optimize the overall operating cost and economic benefits of the virtual power plant under strict constraints such as power balance and equipment operating boundary constraints. Furthermore, a feasible solution repair mechanism is adopted during the iterative optimization process to ensure that the generated scheduling scheme remains physically feasible in each iteration, avoiding interference from infeasible solutions. This mechanism effectively guarantees the reliability of the optimized solution, thereby ensuring that the operation of the virtual power plant always meets various physical constraints. Thus, through preset termination conditions, candidate scheduling schemes can be continuously optimized until the optimal solution is reached, ensuring that the virtual power plant achieves optimal economic benefits within the predetermined scheduling cycle and achieves a balance in power exchange with the power system, thereby improving the overall operating efficiency and economy of the virtual power plant. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0017] Figure 1 A flowchart of a virtual power plant scheduling method for power systems according to an embodiment of this disclosure is shown; Figure 2 A flowchart of a hybrid encoding and initialization method according to an embodiment of this disclosure is shown; Figure 3 A schematic diagram of the structure of a virtual power plant dispatching device for power systems according to an embodiment of this disclosure is shown; Figure 4 This illustration shows a hierarchical structure diagram of a virtual power plant dispatching device for a power system according to an embodiment of the present disclosure; Figure 5 A schematic diagram of the hardware structure of an electronic device according to an embodiment of the present disclosure is shown. Detailed Implementation

[0018] As described in the background section, existing traditional scheduling methods largely rely on optimization algorithms such as linear programming, nonlinear programming, and dynamic programming. While these algorithms have achieved success in some simple scenarios, they often fail to meet practical needs when facing highly complex and dynamically changing power system environments. Specifically: Insufficient handling of strong constraints: The scheduling of virtual power plants not only requires optimizing operating costs, but also needs to consider multiple physical constraints such as power balance, equipment operating limitations, and timing constraints of energy storage charging and discharging. Traditional optimization methods are prone to getting trapped in local optima when dealing with these complex constraints, resulting in scheduling schemes that cannot meet all constraints, or even producing infeasible solutions.

[0019] Multi-objective optimization is challenging: Virtual power plant scheduling requires balancing multiple objectives, such as minimizing operating costs, maximizing renewable energy absorption, and ensuring system security and stability. Most existing scheduling algorithms focus on only a single objective and lack effective multi-objective optimization methods, making it difficult to simultaneously satisfy these objectives.

[0020] Poor dynamic adaptability of dispatching schemes: In actual power systems, factors such as load forecasting errors and uncertainties in renewable energy generation cause the dispatching environment to change constantly. Traditional dispatching methods often lack flexible dynamic adjustment capabilities and cannot respond to system changes in a timely manner, resulting in poor effectiveness and timeliness of dispatching results.

[0021] Particle Swarm Optimization (PSO), a swarm intelligence-based optimization algorithm, has been widely used in power system optimization, especially in the scheduling optimization of virtual power plants, in recent years due to its advantages such as global search capability, fast convergence speed, and strong adaptability. By simulating the process of cooperation and competition among particles in the solution space, PSO can efficiently find the global optimum of a problem, avoiding the local optima problems that may occur in traditional optimization methods.

[0022] However, despite the PSO algorithm's good global search capabilities, it still faces some challenges in practical applications, mainly in the following aspects: Constraint handling: Traditional PSO algorithms do not directly optimize for constraints, which can easily lead to solutions that do not conform to actual physical constraints. Therefore, a constraint repair mechanism needs to be introduced to ensure that all solutions generated during the search process meet the actual operating requirements of the virtual power plant.

[0023] The complexity of multi-objective optimization: Virtual power plant scheduling involves multiple optimization objectives (such as operating costs, renewable energy consumption, and system stability). However, the PSO algorithm often lacks a direct multi-objective optimization mechanism when handling multi-objective optimization, making it difficult to optimize all objectives simultaneously. To improve scheduling efficiency, PSO must be improved by adding multi-objective optimization strategies.

[0024] Adaptability in dynamic environments: The dynamic nature of virtual power plant dispatching environments requires dispatching algorithms to flexibly respond to changes such as load fluctuations and uncertainties in renewable energy generation. Traditional PSO algorithms often fail to adjust dispatching strategies in a timely manner when the environment changes drastically, necessitating the introduction of more dynamic adjustment mechanisms.

[0025] To address the aforementioned issues, this disclosure proposes a novel economic dispatch method for virtual power plants by combining an improved particle swarm optimization algorithm with hybrid coding, feasible solution repair, dynamic adjustment mechanisms, and multi-objective optimization strategies. This method can effectively optimize under multiple objectives and strong constraints, while ensuring good adaptability in dynamic environments. By improving the particle swarm optimization algorithm and considering the actual operational needs of virtual power plants, the accuracy, stability, and economy of the dispatch results can be enhanced, thus providing a more reliable dispatch scheme for the efficient operation of virtual power plants.

[0026] Specifically, by acquiring predicted data, equipment parameters, and market information from the virtual power plant, accurate input data is provided for optimization calculations. Based on the constructed economic dispatch optimization model, by uniformly encoding the dispatch variables of various equipment within the virtual power plant, the time-series values ​​of each decision variable in the dispatch cycle are used as optimization objectives, thus forming a comprehensive and systematic optimization framework. This framework can reasonably optimize the overall operating cost and economic benefits of the virtual power plant under strict constraints such as power balance and equipment operating boundary constraints. Furthermore, during the iterative optimization process, a feasible solution repair mechanism is adopted to ensure that the generated dispatch scheme remains physically feasible in each iteration, avoiding infeasible solutions interfering with the optimization process. This mechanism effectively guarantees the reliability of the optimized solution, thereby ensuring that the operation of the virtual power plant always meets various physical constraints. In this way, through preset termination conditions, candidate dispatch schemes can be continuously optimized until the optimal solution is reached, ensuring that the virtual power plant achieves optimal economic benefits within the predetermined dispatch cycle and achieves a balance in power exchange with the power system, thereby improving the overall operating efficiency and economy of the virtual power plant.

[0027] In other words, this embodiment of the disclosure constructs a full-chain scheduling optimization system from data acquisition to scheme generation, and from initialization settings to iterative updates, through closed-loop optimization management of "unified encoding of decision variables, physical constraint repair, multi-dimensional optimization of objectives, and iterative optimization updates." This method upgrades traditional scheduling schemes that rely on a single objective and static constraints to particle swarm optimization control based on multi-objective evaluation and dynamic constraint repair, forming a dynamic scheduling mechanism that can adapt to changes in power load, equipment status, and system demand. This significantly improves the overall efficiency, reliability, and economy of virtual power plant scheduling in complex power system environments.

[0028] To enable those skilled in the art to better understand and implement this disclosure, the following detailed description of the specific solutions, principles, advantages, and effects of this disclosure is provided with reference to the accompanying drawings and specific embodiments.

[0029] See Figure 1 , Figure 1 This is a flowchart of a virtual power plant dispatching method for a power system according to an embodiment of the present disclosure. The following steps can be performed: S101, obtain the predicted data, equipment parameters and market information of the virtual power plant within the predetermined scheduling cycle.

[0030] In some embodiments, in the economic dispatch of virtual power plants, it is necessary to obtain accurate forecast data, equipment parameters and market information to provide reliable input for the subsequent optimization process.

[0031] Furthermore, considering the variability of power grid implementation, this data is acquired within a predetermined scheduling cycle. This predetermined scheduling cycle can be one day, one week, or even one hour.

[0032] In some embodiments, the forecast data includes load forecasts, distributed generation output forecasts, and / or electricity price forecasts.

[0033] Load forecasting refers to electricity demand prediction data within the area where the virtual power plant is located. Load forecasting typically includes daily load, seasonal fluctuations, and short-term real-time load changes. These forecasts can be generated through historical data analysis and load forecasting models (such as time series models, machine learning models, etc.).

[0034] Distributed power generation output forecasting includes forecasting the output of all renewable energy sources (such as wind and solar power) within a virtual power plant. Because renewable energy generation is uncertain, future output is typically estimated using meteorological data, historical output data, and meteorological forecasting models.

[0035] Electricity price forecasting is based on market electricity price fluctuation patterns, combined with market rules and historical electricity price data, using forecasting algorithms (such as time series analysis, regression analysis, or artificial intelligence models) to predict future electricity price trends. Electricity price forecasting provides a basis for assessing the market economics of power dispatching decisions.

[0036] Equipment parameters include the operating boundary parameters of distributed power sources, energy storage devices, and controllable loads.

[0037] Distributed power parameters include the rated power, start / stop time, and operating limitations of various distributed power sources. For each device, its optimal operating range (such as minimum and maximum output power) also needs to be defined, taking into account its reliability and service life.

[0038] Energy storage device parameters include the total capacity of the energy storage device, charge and discharge efficiency, charge and discharge power limits, and state of storage (SOC) range. Energy storage devices not only affect the load regulation capability of virtual power plants, but also determine the time window for energy storage and release.

[0039] Controllable load parameters include the maximum response capacity, response speed, and minimum response time of adjustable load equipment. Controllable loads are an important resource for the flexible dispatching of virtual power plants and can be used to balance load fluctuations.

[0040] Market information includes real-time electricity prices, peak-valley electricity prices, and ancillary service prices in the electricity market.

[0041] In this way, by combining these predictive data with equipment parameters, an accurate and detailed virtual power plant operating environment can be formed, ensuring that subsequent dispatch models can find the optimal solution under changing market environments and technical constraints.

[0042] S102, construct the economic dispatch optimization model of the virtual power plant. The economic dispatch optimization model uses the set values ​​of each adjustable resource in the equipment parameters at each time period as the decision variable, the minimization of the comprehensive operating cost of the virtual power plant as the objective function, and power balance as the constraint.

[0043] In some embodiments, the scheduling optimization model for a virtual power plant aims to achieve optimal scheduling under multiple constraints based on the objective function of the virtual power plant. The model is specifically constructed as follows: Decision variables include time-series decision variables for adjustable resources and time-series decision variables for control commands. Specifically, the time-series decision variables for adjustable resources include the setpoints for all adjustable resources within the virtual power plant (such as the output of distributed power sources, the charging and discharging power of energy storage devices, and the adjustment amounts of controllable loads) for each time period. For example, for the charging power of energy storage devices, the specific charging or discharging amount needs to be determined for each time period within the scheduling cycle.

[0044] The timing decision variables for control commands include control commands that exchange power with the power system, such as power dispatch between virtual power plants and the power grid, and their specific dispatch periods.

[0045] The objective function includes minimizing operating costs and maximizing economic benefits. The core objective is to minimize the overall operating costs of the virtual power plant during the dispatch cycle, including fuel costs for distributed generation, charging costs for energy storage devices, and the cost of purchasing electricity from the grid. On the other hand, the virtual power plant also needs to maximize revenue from electricity sales or energy storage discharge, especially during peak electricity price periods for energy exchange to obtain higher profits.

[0046] By combining these two objectives, this optimization model achieves the comprehensive economic goal of the virtual power plant, namely, maximizing economic benefits and minimizing operating costs while ensuring power supply security and system stability.

[0047] The constraints include power balance constraints, state of energy storage (SOC) constraints, and equipment operation constraints. Among them, the power balance constraint means that the total power generation of the virtual power plant and the power load of the grid must meet the requirements at any time period, including the output of distributed generation, the charging and discharging of energy storage devices, and the adjustment of controllable loads.

[0048] Energy storage state constraints refer to the requirement that the state of charge (SOC) of energy storage devices must be kept within a reasonable range to ensure battery life and efficiency.

[0049] Equipment operation constraints refer to the limitations of each distributed power source's power output on its equipment capabilities (such as maximum output power limits), the charging and discharging power limits of energy storage devices, and the start-stop state limits.

[0050] By constructing this comprehensive economic dispatch optimization model, the virtual power plant can optimize resource allocation and dispatch based on real-time market information and equipment status under given constraints.

[0051] In one embodiment, the overall operating cost is minimized: ; The power balance: ; in, Indicates the scheduled scheduling period, Indicates in The cost of purchasing electricity at all times This indicates the operating cost of the power generation equipment within the virtual power plant. Indicates the charging and discharging cost of virtual power plant energy storage devices, This indicates that the power generation equipment in the virtual power plant is... Output power at any time Indicates that the energy storage device is in Charge and discharge power at any time Indicates in Total load demand at any given time This indicates the power purchased.

[0052] It should be noted that the parameters of the above formula may vary for different application scenarios, but all of them satisfy the balance.

[0053] S103, Solve the economic dispatch optimization model to generate candidate dispatch schemes within the dispatch period, and iteratively optimize the candidate dispatch schemes. The iterative optimization includes uniformly encoding the time-series values ​​of various dispatch variables within the dispatch period into a set of dispatch variables for the virtual power plant dispatch scheme, and integrating a feasible solution repair mechanism based on the physical constraints of the virtual power plant during the update process.

[0054] In one embodiment, the particle swarm optimization algorithm is used to solve the economic scheduling optimization model.

[0055] Accordingly, step S103 can be described as follows: the particle swarm optimization algorithm is used to solve the economic scheduling optimization model. The particle swarm optimization algorithm encodes the time-series values ​​of various scheduling variables in the scheduling period into a set of scheduling variables for the virtual power plant scheduling scheme, and integrates a feasible solution repair mechanism based on the physical constraints of the virtual power plant during the particle position update process.

[0056] The particle swarm optimization algorithm takes the decision variables as the optimization objects, encodes the time-series values ​​of various decision variables in the scheduling period into a set of scheduling variables for the virtual power plant scheduling scheme, and integrates a feasible solution repair mechanism based on the physical constraints of the virtual power plant during the scheduling scheme update process to ensure the physical feasibility or approximate feasibility of the candidate scheduling scheme in the iteration process.

[0057] The reason why the particle swarm optimization algorithm can determine the optimal scheduling scheme is as follows: Particle Swarm Optimization (PSO) searches the search space through a collaborative process, with each particle moving within the search space, mimicking the behavior of a flock of birds foraging. In each iteration, a particle relies not only on its own historical best solution but also on the collective historical best solution, thus enabling it to find a global optimum in complex, multi-modal search spaces. Furthermore, due to its widespread application, PSO has been proven to find optimal solutions in many optimization problems, especially strongly constrained, multi-objective problems.

[0058] Furthermore, while particle swarm optimization deals with a continuous solution space, virtual power plant scheduling involves numerous physical constraints (such as power balance, energy storage state of charge, and equipment start-up / shutdown states). By introducing a feasible solution repair mechanism, it can be ensured that after each particle position update, the generated candidate solutions satisfy the physical constraints of the virtual power plant (such as power balance, energy storage constraints, and start-up / shutdown constraints), avoiding unsuitable or unrealistic solutions. The repair process guarantees that even if infeasible solutions are generated during the search process, they can be transformed into suitable, physically feasible solutions through repair.

[0059] Furthermore, by constructing a multi-objective fitness function that includes multiple optimization goals (such as operating cost, carbon emissions, energy integration rate, and user comfort), the particle swarm optimization algorithm can balance and optimize among these goals. The final optimal scheduling scheme not only minimizes costs but also improves renewable energy integration, enhances user experience, and comprehensively improves the economics and sustainability of the virtual power plant.

[0060] Based on these characteristics, the improved particle swarm optimization algorithm can find the optimal scheduling scheme under the complex constraints and multi-objective requirements of virtual power plants through a reasonable search mechanism and effective repair methods.

[0061] In some embodiments, to simultaneously handle continuous decision variables (such as generator output and energy storage charging / discharging power) and discrete decision variables (such as generator start / stop status and controllable load response) of equipment in a virtual power plant, this invention employs a hybrid encoding strategy to uniformly encode the time-series values ​​of various decision variables in the time dimension. Real-number encoding is used for continuous variables, while binary encoding is used for discrete variables. This hybrid encoding method enables efficient scheduling of different types of equipment in the virtual power plant.

[0062] During population initialization, the initial position of each particle is feasiblely initialized based on the equipment operation boundary constraints of the virtual power plant, ensuring that the initial solution conforms to the physical constraints. This process provides a reasonable starting point for subsequent particle swarm optimization, avoids invalid searches, and improves the efficiency and feasibility of the algorithm.

[0063] See Figure 2 , Figure 2 This is a flowchart of a hybrid encoding and initialization method according to an embodiment of this disclosure, such as... Figure 2 As shown, it may include: S201, under the premise that the predetermined scheduling period and time resolution are determined, establish a unified scheduling variable expression structure for various types of equipment inside the virtual power plant.

[0064] The predetermined scheduling period refers to the time range within which the virtual power plant performs scheduling optimization.

[0065] For example, a virtual power plant might be scheduled over a day (24 hours), a week, or a month, and the scheduling cycle determines the time span of the optimization problem. The scheduling cycle is generally set based on actual needs, and is usually selected as a time period at the level of days, hours, or minutes.

[0066] For example, if a 24-hour scheduling cycle is chosen, then all equipment operation, power scheduling, energy storage management, etc. within this cycle will be optimized according to the demand changes within 24 hours.

[0067] Time resolution refers to the precision of each time unit (e.g., hour, minute) within a predetermined scheduling period.

[0068] In some cases, the time resolution may be on the order of hours, meaning each scheduling step is one hour, or it may be on the order of minutes, meaning the scheduling is calculated every 5 minutes or 1 minute. The choice of time resolution affects the granularity of the scheduling results and the computational complexity.

[0069] In a 24-hour scheduling cycle, if the time resolution is hourly, there are 24 scheduling periods per day; if the time resolution is 30 minutes, there are 48 scheduling periods per day. This means that the precision of each time unit has a significant impact on the granularity of scheduling optimization and the amount of computation.

[0070] In other words, in virtual power plant scheduling optimization, it is first necessary to determine the scheduling time span (scheduling period), and then select an appropriate time granularity (time resolution) as needed. The choice of time resolution affects the granularity of the scheduling problem and also affects the computational cost of the optimization algorithm.

[0071] In some embodiments, hybrid coding is employed to handle continuous and discrete decision variables (e.g., distributed power generation, energy storage charging and discharging, and equipment start-up and shutdown status) in a virtual power plant.

[0072] For example, continuous variables (such as power setpoints) use real number encoding; discrete variables (such as start / stop status) use binary encoding.

[0073] Hybrid encoding allows particles to be optimized simultaneously in different search spaces (continuous and discrete), thus enabling more flexible and extensive exploration.

[0074] In one embodiment, a unified scheduling variable expression structure is established for various types of equipment within the virtual power plant, including: encoding continuously adjustable quantities such as distributed power output, energy storage charging and discharging power, power exchange with the power system, and controllable load power level in a continuous numerical manner; encoding equipment with start-stop states in a discrete state sequence manner, so that the operating state and power settings of each equipment in each time period together constitute a complete virtual power plant scheduling scheme.

[0075] S202, under the preset variable value range and equipment operation boundary constraints, based on the scheduling variable expression structure, multiple initial candidate scheduling schemes for virtual power plants are randomly generated. Each candidate scheduling scheme is processed for feasibility to ensure that it at least meets the upper and lower capacity limits of continuous equipment, the initial state of charge constraints of energy storage equipment, and the minimum continuous operating time and minimum continuous downtime constraints of equipment with start-stop states.

[0076] In some embodiments, in the scheduling optimization problem, the preset variable range refers to the maximum and minimum values ​​that each decision variable (e.g., power output, energy storage capacity, start-stop status, etc.) can accept during the optimization process. The value range of each variable is a preset boundary based on the physical limitations of the equipment, operating conditions, or market demand.

[0077] For example, if the decision variable is the output power of a wind turbine, its value range may be from 0 (shutdown) to the maximum rated power.

[0078] In one example, if the maximum power of a wind turbine is 2MW, then its preset variable value range is [0, 2MW].

[0079] Equipment operating boundary constraints refer to the physical or operational limitations that each piece of equipment must adhere to during actual operation.

[0080] For example, generator sets, energy storage devices, or load equipment must meet certain constraints under different operating modes. These constraints may include the equipment's minimum / maximum output, start-up and shutdown times, temperature limits, etc.

[0081] For example, a coal-fired power generation unit may have a minimum operating time limit (e.g., the operating time must not be less than 4 hours) or a minimum downtime (e.g., it cannot be restarted immediately after shutdown and must be shut down for 8 hours).

[0082] In short, during the scheduling optimization process, decision variables (such as power generation and energy storage charging and discharging power) are subject to preset range limitations, while the physical and operating conditions of the equipment also provide additional operational boundary constraints.

[0083] In some embodiments, step S202 may specifically include: S2021 defines a scheduling variable expression structure for the scheduling problem of a virtual power plant. The structure includes decision variables such as the output of adjustable resources, charging and discharging power, and start-up and shutdown status in each time period. The decision variables include continuous variables and discrete variables.

[0084] When a continuous decision variable is randomly generated, the value range of the decision variable is the interval between the maximum and minimum output of the equipment, and the range is determined by the physical characteristics of the equipment.

[0085] When a discrete decision variable is randomly generated, the value of the decision variable is 0 or 1, which indicates the activation or deactivation status of the device in the corresponding time period.

[0086] S2022, randomly generate continuous decision variables such as distributed power output, energy storage device charging and discharging power, and controllable load power for each time period, with their values ​​within the corresponding physical operating boundaries; randomly generate discrete decision variables such as equipment start-up and shutdown status and load response for each time period, where the discrete decision variables are binary values ​​representing the start-up or shutdown status of the equipment.

[0087] S2023, perform feasibility repair on the generated candidate scheduling scheme to ensure that all decision variables comply with the physical constraints and operational limitations of the equipment. The constraints include, but are not limited to, power balance constraints, energy storage state of charge constraints, equipment start-up and shutdown time constraints, and maximum and minimum output constraints.

[0088] The feasibility repair includes power balance repair, energy storage state of charge repair, and equipment start-up and shutdown state repair, ensuring that all generated candidate solutions are physically feasible and meet the constraints.

[0089] S2024. Repeat steps S2022 and S2023 to generate multiple different initial candidate scheduling schemes for virtual power plants, which will serve as the initial population for subsequent optimization algorithms.

[0090] S203, construct a multi-objective evaluation index system for the virtual power plant that simultaneously considers economic efficiency and operational performance, and form corresponding comprehensive evaluation indexes through a multi-objective evaluation strategy.

[0091] Among them, the multi-objective evaluation indicators include at least two of the following: the comprehensive operating cost and / or comprehensive economic benefits of the virtual power plant during the dispatch cycle, the level of renewable energy output absorption, and the degree of deviation of user-side electricity comfort.

[0092] Comprehensive evaluation metrics are typically used to quantify the performance of multiple objectives or constraints in a scheduling scheme. These metrics take into account factors such as cost, efficiency, and environmental impact.

[0093] It should be noted that the goal of the economic dispatch optimization model is to optimize the dispatch scheme of the virtual power plant so that certain objectives (such as minimizing operating costs or maximizing revenue) are optimal; while the role of the multi-objective evaluation index is to comprehensively evaluate each candidate dispatch scheme, which usually includes the measurement of multiple objectives or indicators.

[0094] S204, using the comprehensive evaluation index and the degree of violation of power balance constraints and equipment operation boundary constraints as evaluation criteria, calculate the multi-objective fitness value of each virtual power plant candidate scheduling scheme, and use the fitness value as the basis for iterative improvement.

[0095] In some embodiments, the multi-objective fitness value is an indicator for evaluating the overall performance of a candidate scheduling scheme. The fitness value comprehensively considers the effects of different objectives (such as minimizing cost, maximizing revenue, and achieving power balance). A lower fitness value indicates a better scheme. By calculating the multi-objective fitness value, multiple scheduling objectives can be evaluated simultaneously.

[0096] During the optimization process, the fitness value serves as feedback information, guiding the algorithm's iterative process. By adjusting the virtual power plant's scheduling scheme (such as power generation and energy storage charging and discharging power), the fitness value is gradually optimized, thus gradually approaching the optimal solution.

[0097] Specifically, step S204 may include: S2041. Based on the scheduling scheme of the virtual power plant, calculate the comprehensive evaluation index. The comprehensive evaluation index includes at least the weighted sum of multiple objectives such as operating cost, equipment efficiency, and environmental impact, and serves as the overall performance evaluation of each candidate scheduling scheme for the virtual power plant.

[0098] S2042, Based on the candidate scheduling scheme of the virtual power plant, calculate the degree of violation of its power balance constraints and equipment operation boundary constraints. The degree of violation of the power balance constraints is the deviation between the power generation, energy storage power and load demand of the candidate scheme in each time period. The degree of violation of the equipment operation boundary constraints is the violation of the equipment's maximum / minimum output, start-up and shutdown time and other operating restrictions.

[0099] S2043. Calculate the multi-objective fitness value of the candidate scheduling scheme for the virtual power plant based on the comprehensive evaluation index and the degree of constraint violation. The fitness value is a comprehensive evaluation result obtained after comprehensively considering multiple objectives such as the scheme's operating cost, efficiency, and constraint violation. The smaller the fitness value, the better the candidate scheme.

[0100] S2044 uses the multi-objective fitness value as the basis for subsequent optimization algorithm iterations. The candidate scheduling scheme is adjusted according to the fitness value to achieve the optimization of the objective function until the preset iteration termination condition is met.

[0101] In other words, this disclosure constructs a complete process from initializing scheduling decision variables to evaluating optimization targets through a hybrid coding strategy, feasibility processing during initialization, and closed-loop management of multi-dimensional target evaluation. This method upgrades the traditional scheduling method, which relies on encoding a single scheduling variable, to a hybrid mode based on real number and binary encoding. Combined with a feasibility initialization mechanism, it ensures that the initial particle scheme meets the physical constraints of the virtual power plant, avoiding interference from invalid solutions.

[0102] Through this mechanism, various devices in the virtual power plant can obtain a scheduling scheme that meets the actual requirements in the initial stage, ensuring the efficiency and convergence of the optimization process, thereby significantly improving the startup efficiency and calculation accuracy of the entire virtual power plant scheduling process.

[0103] In practical applications, the virtual power plant scheduling scheme generated after each particle swarm update may violate certain physical constraints. To ensure the feasibility of the virtual power plant scheduling scheme, this disclosure introduces a feasible solution repair mechanism. This repair mechanism repairs aspects such as power balance, energy storage state of charge, and equipment start-up / shutdown sequences one by one, ensuring that candidate scheduling schemes strictly meet all physical constraints of the virtual power plant.

[0104] Each candidate scheduling scheme after the correction will undergo a fitness evaluation, with evaluation metrics including operating costs, renewable energy integration rate, and user comfort. This evaluation prioritizes candidate scheduling schemes with higher fitness values ​​that also meet physical constraints, ensuring the algorithm consistently finds a reasonable solution. More specifically, the feasible solution repair mechanism includes at least two of the following repair mechanisms, which are executed in a preset order after each new candidate scheduling scheme is generated: The mechanism for repairing power balance in virtual power plants is as follows: For the current candidate scheduling scheme, the difference between the total power generation output, net energy storage output, controllable load adjustment amount and the sum of rigid load and baseline load of the virtual power plant is calculated in each scheduling period. This difference is taken as the power imbalance amount. According to the preset scheduling priority rules, the power exchange plan between the virtual power plant and the power system is adjusted first, and then the charging and discharging power of the energy storage device and / or the load level of the controllable load are adjusted. A repair mechanism for energy storage dynamics: During the scheduling period, the charging and discharging power of the energy storage devices in the current candidate scheduling scheme is accumulated in chronological order, and the trajectory of the state of charge change of each energy storage device is simulated and calculated. When the state of charge of a certain period exceeds the preset upper and lower limits or does not meet the predetermined beginning and end state requirements at the end of the scheduling period, the energy storage power sequence in the candidate scheduling scheme is modified by scaling, shifting or redistributing the charging and discharging power of the relevant period, so that the updated state of charge trajectory meets the capacity constraints and the beginning and end state constraints. Operation sequence repair mechanism for equipment with start-stop states: The start-stop sequences of generator sets, load devices or other equipment described by start-stop states in the current candidate scheduling scheme are detected. When there are operation segments that do not meet the minimum continuous running time and / or minimum continuous downtime constraints, the start-stop state sequence is locally adjusted by merging, extending or shortening adjacent start-stop segments so that the repaired start-stop sequence meets the corresponding running time constraints.

[0105] The reasons for this approach are as follows: first, to restore power balance and ensure a balance between power supply and demand; second, to restore energy storage capacity and ensure that energy storage devices operate within feasible limits; and finally, to restore start-up and shutdown status and ensure that the equipment operates as required.

[0106] Thus, by implementing a closed-loop control system encompassing feasible solution repair, constraint verification, and real-time evaluation feedback, a comprehensive repair mechanism is constructed, covering the entire process from particle update to feasible solution generation. This method upgrades the constraint handling approach in traditional particle swarm optimization algorithms to include multiple repair methods such as power balance repair, energy storage state of charge repair, and start-stop state repair, ensuring that each candidate scheduling scheme maintains physical feasibility after each iteration. Through this repair mechanism, the algorithm can correct unconstrained schemes in real time under the complex constraints of a virtual power plant, preventing unsuitable solutions from entering the optimization process and continuously improving the effectiveness of the scheduling scheme in each round of optimization. This mechanism effectively enhances the feasibility guarantee of virtual power plant scheduling schemes, ensuring the reliability and executability of the optimization results.

[0107] In each iteration, based on the aforementioned improved virtual power plant scheduling scheme, the particle swarm optimization algorithm continuously improves the quality of candidate scheduling schemes through an evolutionary update mechanism. Specifically, a feasibility-first scheme selection mechanism is adopted, prioritizing the retention of scheduling schemes that meet the constraints and have better fitness. Simultaneously, the algorithm employs differentiated update strategies for different types of equipment: for continuously adjustable resources (such as generator output, energy storage charging and discharging), the standard particle swarm optimization update formula is used; for equipment with start-stop states, a state reversal rule based on probability mapping is adopted.

[0108] Furthermore, by dynamically adjusting parameters (such as inertia weights and learning factors) in each update, the algorithm can perform a broad global search in the early stages and focus on optimizing the high-quality solutions obtained in the later stages, thereby converging to the optimal solution in a shorter time.

[0109] More specifically, in solving the economic dispatch optimization model of the virtual power plant, the candidate dispatch schemes of the virtual power plant and its internal equipment are iteratively optimized using the following evolution update mechanism.

[0110] In each iteration round, when comparing and screening multiple candidate scheduling schemes for virtual power plants, candidate scheduling schemes that satisfy the virtual power plant power balance constraints, distributed power output constraints, energy storage state of charge constraints, and controllable load operation constraints are given priority. When two candidate scheduling schemes both satisfy the above constraints, the candidate scheduling scheme with the better multi-objective fitness value is selected as the corresponding virtual power plant's historical best scheduling scheme and / or current global best scheduling scheme.

[0111] When iteratively updating the scheduling variables of the equipment within the virtual power plant, the scheduling variables representing the output of distributed power sources, the charging and discharging power of energy storage, the power exchange with the power system, and the continuous adjustment of the controllable load power level are adjusted using numerical update rules in continuous space. For the scheduling variables representing the switching states of generator sets, load devices, or other equipment with start-stop states, state flipping rules based on probability mapping are used for updating. This achieves collaborative optimization of the operating states of continuous and discrete equipment within the virtual power plant under the same scheduling optimization framework.

[0112] During the iterative evolution of the virtual power plant scheduling scheme, the weight parameters and guiding coefficients used to generate a new round of candidate scheduling schemes are linearly reduced based on the iteration process, the fitness distribution of each candidate scheduling scheme, and the constraint violation.

[0113] In other words, this embodiment of the disclosure constructs a multi-level optimization framework from updating virtual power plant scheduling schemes to outputting the optimal solution through an evolutionary update mechanism, a dynamic adjustment strategy, and a closed-loop optimization control for scheme priority selection. This method upgrades the relatively static scheduling scheme update mechanism in traditional scheduling methods to a dynamic evolutionary process through a feasibility-priority selection mechanism, a differentiated update mechanism, and a dynamic parameter adjustment mechanism. This ensures that each virtual power plant scheduling scheme, under the premise of physical feasibility, can flexibly evolve in a multi-objective, strongly constrained environment, thereby gradually converging to the global optimal solution.

[0114] Through these technological innovations, this method can ensure that the scheme updates during the virtual power plant dispatching process can be carried out efficiently and robustly, especially in the face of a complex and ever-changing power market environment, and can continuously adapt to and output the optimal dispatching scheme.

[0115] Furthermore, during the iterative optimization of candidate scheduling schemes for virtual power plants, local search enhancements are periodically performed on the current optimal scheduling scheme for virtual power plants.

[0116] That is, based on particle swarm optimization, the scheduling scheme of the virtual power plant is further refined and optimized by perturbing and adjusting the current optimal solution to a certain extent.

[0117] Local search enhancement includes: Within a preset neighborhood, based on the current optimal scheduling scheme of the virtual power plant, a controlled-amplitude random perturbation is applied to the output curves, energy storage charging and discharging sequences, and / or start-stop state sequences of some of the equipment to generate several new candidate scheduling schemes for the virtual power plant.

[0118] The feasible solution repair mechanism is invoked for each of the new candidate scheduling schemes to correct the constraints on power balance, energy storage state of charge and equipment operating time, so as to obtain a repair scheme that meets the physical constraints of the virtual power plant, and the corresponding fitness value is calculated according to the multi-objective evaluation index.

[0119] When any repaired new candidate scheduling scheme has a better multi-objective fitness value than the current best virtual power plant scheduling scheme, provided that all constraints are met, the new candidate scheduling scheme replaces the current best virtual power plant scheduling scheme and serves as the benchmark scheme to guide the evolution and update of other candidate scheduling schemes in subsequent iterations.

[0120] Thus, by perturbing and adjusting the current optimal scheduling scheme, the particle swarm optimization algorithm can be effectively prevented from getting stuck in local optima, improving global search capabilities and ensuring that the virtual power plant can achieve a better economical scheduling scheme in a variable power system environment. This allows the virtual power plant scheduling not only to find a good solution in the initial stage but also to continuously refine it in subsequent optimization processes, ultimately obtaining an efficient, stable, and economical optimal scheduling scheme.

[0121] S104. When the iteration process meets the preset termination condition, the current best virtual power plant scheduling scheme is taken as the optimal scheduling plan, the corresponding control instructions for the scheduling instructions are generated, and they are executed within the predetermined scheduling period.

[0122] In some embodiments, by executing steps S101 to S103, when the iterative optimization process meets the preset termination condition, the current best virtual power plant scheduling scheme can be used as the optimal scheduling plan, and control instructions for distributed power sources, energy storage devices, controllable loads, and power exchange with the power system can be generated according to the optimal scheduling plan and executed within the scheduling cycle.

[0123] The virtual power plant dispatching method for power systems has been described in detail above through some embodiments. In order to enable those skilled in the art to better understand and implement it, the corresponding apparatus is also described in detail below through some embodiments.

[0124] See Figure 3 The diagram shown is a structural schematic of a virtual power plant dispatching device for power systems in an embodiment of this disclosure. Figure 3 As shown, the virtual power plant dispatching device 300 for power systems may include: The acquisition unit 310 is configured to acquire the predicted data, equipment parameters and market information of the virtual power plant within a predetermined scheduling period; The construction unit 320 is configured to construct an economic dispatch optimization model for the virtual power plant. The economic dispatch optimization model uses the set values ​​of each adjustable resource in the equipment parameters at each time period as decision variables, the minimization of the comprehensive operating cost of the virtual power plant as the objective function, and power balance as a constraint. The processing unit 330 is configured to solve the economic dispatch optimization model, generate candidate dispatch schemes within the dispatch period, and iteratively optimize the candidate dispatch schemes. The iterative optimization includes uniformly encoding the time-series values ​​of various dispatch variables within the dispatch period into a set of dispatch variables for the virtual power plant dispatch scheme, and integrating a feasible solution repair mechanism based on the physical constraints of the virtual power plant during the update process. The scheduling unit 340 is configured to take the current best virtual power plant scheduling scheme as the optimal scheduling plan when the iteration process meets the preset termination condition, generate the corresponding control instructions for the scheduling instructions, and execute them within the predetermined scheduling period.

[0125] For further details regarding the acquisition unit 310, the construction unit 320, the processing unit 330, and the scheduling unit 340, please refer to the aforementioned examples.

[0126] It is understandable that the above division of units is only a logical functional division. In actual implementation, they can be fully or partially integrated into a single physical entity, or they can be physically separated. Furthermore, the above units can be implemented by the processor calling software.

[0127] See Figure 4 The diagram shown below illustrates a hierarchical structure of a virtual power plant dispatching device for a power system, as described in this embodiment of the present disclosure. Figure 4 As shown, the entire scheduling system can be divided into three parts: virtual power plants, resource entities, and cluster agents.

[0128] The main resources include distributed photovoltaic (PV) systems, energy storage systems, adjustable loads, and electric vehicles. Correspondingly, the cluster agents include: power-side clusters, coupled to distributed PV and energy storage systems; energy storage-side clusters, coupled to energy storage systems; and load-side clusters, coupled to adjustable loads and electric vehicles.

[0129] In addition, the power generation cluster, energy storage cluster, and load cluster are all coupled to the virtual power plant, which is equipped with an economic dispatch optimization model, so that resource dispatch can be carried out together with the electricity market according to the scheme in the aforementioned example.

[0130] This disclosure also provides an electronic device for implementing a virtual power plant dispatching method for power systems.

[0131] The electronic device includes a memory and a processor, as well as a computer program stored in the memory and executable on the processor, wherein the processor, when executing the program, implements the method as described in any of the foregoing descriptions.

[0132] It should be noted that the computer system of the electronic device shown below is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0133] like Figure 5 The schematic diagram shown in this embodiment illustrates the hardware structure of an electronic device 500. The electronic device 500 includes a Central Processing Unit (CPU) 501, which can perform various appropriate actions and processes based on programs stored in Read-Only Memory (ROM) 502 or loaded from storage portion 508 into Random Access Memory (RAM) 503, such as executing the methods described in the above embodiments. The RAM 503 also stores various programs and data required for system operation. The CPU 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0134] The following components are connected to I / O interface 505: an input section 506 including a keyboard, mouse, etc.; an output section 507 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN (Local Area Network) card, modem, etc. The communication section 509 performs communication processing via a network such as the Internet. Drive 55 is also connected to I / O interface 505 as needed. Removable media 511, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 55 as needed so that computer programs read from them can be installed into storage section 508 as needed.

[0135] Specifically, according to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program including a computer program for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit (CPU), it performs various functions defined in the system of this application.

[0136] It should be noted that the computer-readable medium shown in the embodiments of this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0137] According to one aspect of this application, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods provided in the various alternative implementations described above.

[0138] In another aspect, this application also provides a computer-readable medium, which may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to perform the methods described in the above embodiments.

[0139] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0140] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0141] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0142] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0143] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0144] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A virtual power plant dispatching method for power systems, characterized in that, include: Obtain forecast data, equipment parameters, and market information for virtual power plants within a predetermined scheduling cycle; An economic dispatch optimization model for the virtual power plant is constructed. The economic dispatch optimization model uses the set values ​​of each adjustable resource in the equipment parameters at each time period as decision variables, the minimization of the comprehensive operating cost of the virtual power plant as the objective function, and power balance as a constraint. The economic dispatch optimization model is solved to generate candidate dispatch schemes within the dispatch period, and the candidate dispatch schemes are iteratively optimized. The iterative optimization includes uniformly encoding the time-series values ​​of various dispatch variables within the dispatch period into a set of dispatch variables for the virtual power plant dispatch scheme, and integrating a feasible solution repair mechanism based on the physical constraints of the virtual power plant during the update process. When the iteration process meets the preset termination condition, the current best virtual power plant scheduling scheme is taken as the optimal scheduling plan, the corresponding control instructions for the scheduling instructions are generated, and they are executed within the predetermined scheduling period.

2. The virtual power plant dispatching method for power systems according to claim 1, characterized in that, The forecast data includes load forecasts, distributed generation output forecasts, and / or electricity price forecasts; the equipment parameters include operating boundary parameters of distributed generation, energy storage devices, and controllable loads; and the market information includes real-time electricity prices, peak-valley electricity prices, and ancillary service prices in the electricity market. The overall operating cost is minimized as follows: ； The power balance: ； in, Indicates the scheduled scheduling period, Indicates in The cost of purchasing electricity at all times This indicates the operating cost of the power generation equipment within the virtual power plant. Indicates the charging and discharging cost of virtual power plant energy storage devices, This indicates that the power generation equipment in the virtual power plant is... Output power at any time Indicates that the energy storage device is in Charge and discharge power at any time Indicates in Total load demand at any given time This indicates the power purchased.

3. The virtual power plant dispatching method for power systems according to claim 1, characterized in that, Solving the aforementioned economic scheduling optimization model includes: Under the premise that the predetermined scheduling period and time resolution are determined, a unified scheduling variable expression structure is established for various types of equipment within the virtual power plant; Under the constraints of preset variable value range and equipment operation boundary, based on the scheduling variable expression structure, multiple initial candidate scheduling schemes for virtual power plants are randomly generated. Each candidate scheduling scheme is made feasible to ensure that it at least meets the upper and lower capacity limits of continuous equipment, the initial state of charge of energy storage equipment, and the minimum continuous operating time and minimum continuous downtime constraints of equipment with start-stop states. A multi-objective evaluation index system that considers both economic efficiency and operational performance is constructed for the virtual power plant, and a corresponding comprehensive evaluation index is formed through a multi-objective evaluation strategy. The comprehensive evaluation index, as well as the degree of violation of power balance constraints and equipment operation boundary constraints, are used as evaluation criteria. The multi-objective fitness value of each virtual power plant candidate scheduling scheme is calculated, and the fitness value is used as the basis for iterative improvement.

4. The virtual power plant dispatching method for power systems according to claim 3, characterized in that, The above establishes a unified scheduling variable expression structure for various equipment within the virtual power plant, including: encoding continuously adjustable quantities such as distributed power output, energy storage charging and discharging power, power exchange with the power system, and controllable load power level in a continuous numerical manner; encoding equipment with start-stop states in a discrete state sequence manner, so that the operating state and power settings of each equipment in each time period together constitute a complete virtual power plant scheduling scheme. The multi-objective evaluation indicators include at least two of the following: the comprehensive operating cost and / or comprehensive economic benefits of the virtual power plant during the dispatch cycle, the level of renewable energy output absorption, and the degree of deviation in user-side electricity comfort.

5. The virtual power plant dispatching method for power systems according to claim 1, characterized in that, The feasible solution repair mechanism includes at least two of the following repair mechanisms, which are executed in a preset order after each new candidate scheduling scheme is generated: The mechanism for repairing power balance in virtual power plants is as follows: For the current candidate scheduling scheme, the difference between the total power generation output, net energy storage output, controllable load adjustment amount and the sum of rigid load and baseline load of the virtual power plant is calculated in each scheduling period. This difference is taken as the power imbalance amount. According to the preset scheduling priority rules, the power exchange plan between the virtual power plant and the power system is adjusted first, and then the charging and discharging power of the energy storage device and / or the load level of the controllable load are adjusted. A repair mechanism for energy storage dynamics: During the scheduling period, the charging and discharging power of the energy storage devices in the current candidate scheduling scheme is accumulated in chronological order, and the trajectory of the state of charge change of each energy storage device is simulated and calculated. When the state of charge of a certain period exceeds the preset upper and lower limits or does not meet the predetermined beginning and end state requirements at the end of the scheduling period, the energy storage power sequence in the candidate scheduling scheme is modified by scaling, shifting or redistributing the charging and discharging power of the relevant period, so that the updated state of charge trajectory meets the capacity constraints and the beginning and end state constraints. Operation sequence repair mechanism for equipment with start-stop states: The start-stop sequences of generator sets, load devices or other equipment described by start-stop states in the current candidate scheduling scheme are detected. When there are operation segments that do not meet the minimum continuous running time and / or minimum continuous downtime constraints, the start-stop state sequence is locally adjusted by merging, extending or shortening adjacent start-stop segments so that the repaired start-stop sequence meets the corresponding running time constraints.

6. The virtual power plant dispatching method for power systems according to claim 1, characterized in that, In solving the economic dispatch optimization model for virtual power plants, the following evolutionary update mechanism is used for iterative optimization of candidate dispatch schemes for virtual power plants and their internal equipment: In each iteration round, when comparing and screening multiple candidate scheduling schemes for virtual power plants, the candidate scheduling schemes that satisfy the virtual power plant power balance constraints, distributed power output constraints, energy storage charge state constraints, and controllable load operation constraints are given priority. When two candidate scheduling schemes satisfy the above constraints, the candidate scheduling scheme with the better multi-objective fitness value is selected as the corresponding virtual power plant's historical best scheduling scheme and / or current global best scheduling scheme. When iteratively updating the scheduling variables of the equipment inside the virtual power plant, the scheduling variables representing the output of distributed power sources, the charging and discharging power of energy storage, the power exchange with the power system, and the continuous adjustment of the controllable load power level are adjusted using the numerical update rules of continuous space; the scheduling variables representing the switching states of generator sets, load devices, or other equipment with start-stop states are updated using the state flipping rules based on probability mapping. During the iterative evolution of the virtual power plant scheduling scheme, the weight parameters and guiding coefficients used to generate a new round of candidate scheduling schemes are linearly reduced based on the iteration process, the fitness distribution of each candidate scheduling scheme, and the constraint violation.

7. The virtual power plant dispatching method for power systems according to claim 6, characterized in that, During the iterative optimization of candidate scheduling schemes for virtual power plants, local search enhancement is also periodically performed on the current optimal scheduling scheme for virtual power plants. This local search enhancement includes: Within a preset neighborhood, based on the current best scheduling scheme of the virtual power plant, a controlled-amplitude random disturbance is applied to the output curves, energy storage charging and discharging sequences and / or start-stop state sequences of some of the equipment to generate several new candidate scheduling schemes for the virtual power plant. The feasible solution repair mechanism is invoked for each of the new candidate scheduling schemes to correct the constraints on power balance, energy storage state of charge and equipment operating time, so as to obtain a repair scheme that meets the physical constraints of the virtual power plant, and the corresponding fitness value is calculated according to the multi-objective evaluation index. When any repaired new candidate scheduling scheme has a better multi-objective fitness value than the current best virtual power plant scheduling scheme, provided that all constraints are met, the new candidate scheduling scheme replaces the current best virtual power plant scheduling scheme and serves as the benchmark scheme to guide the evolution and update of other candidate scheduling schemes in subsequent iterations.

8. The virtual power plant dispatching method for power systems according to any one of claims 3 to 7, characterized in that, The economic scheduling optimization model is solved using a particle swarm optimization algorithm. The particle swarm optimization algorithm encodes the time-series values ​​of various scheduling variables within the scheduling period into a set of scheduling variables for the virtual power plant scheduling scheme, and integrates a feasible solution repair mechanism based on the physical constraints of the virtual power plant during the particle position update process.

9. A virtual power plant dispatching device for power systems, characterized in that, include: The acquisition unit is configured to acquire forecast data, equipment parameters, and market information of the virtual power plant within a predetermined scheduling period; The construction unit is configured to construct an economic dispatch optimization model for the virtual power plant. The economic dispatch optimization model uses the set values ​​of each adjustable resource in the equipment parameters at each time period as decision variables, the minimization of the overall operating cost of the virtual power plant as the objective function, and power balance as a constraint. The processing unit is configured to solve the economic scheduling optimization model, generate candidate scheduling schemes within the scheduling period, and iteratively optimize the candidate scheduling schemes. The iterative optimization includes uniformly encoding the time-series values ​​of various scheduling variables within the scheduling period into a set of scheduling variables for the virtual power plant scheduling scheme, and integrating a feasible solution repair mechanism based on the physical constraints of the virtual power plant during the update process. The scheduling unit is configured to take the current best virtual power plant scheduling scheme as the optimal scheduling plan when the iteration process meets the preset termination condition, generate the corresponding control instructions for the scheduling instructions, and execute them within the predetermined scheduling period.

10. An electronic device comprising a memory and a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1-8.

Images