High-heat and high-humidity region flexible resource random robust optimization regulation and control method and system

Abstract

The invention belongs to the technical field of supply and demand interaction in high-heat and high-humidity areas, and discloses a high-heat and high-humidity area flexible resource random robust optimization regulation and control method and system so as to solve the regulation and control problem of a regional energy system in a high-heat and high-humidity environment. The method comprises the following steps: collecting flexible resource data, power grid data and meteorological data; a deterministic optimization model is constructed, and minimization of the daily operation total cost of the micro-grid in the target region is taken as a target; on the basis of the deterministic optimization model, introducing an uncertainty set of photovoltaic output and load power and Monte Carlo simulation of real-time electricity price, constructing a random robust optimization model, and performing iterative solution through a column constraint generation algorithm to generate a collaborative regulation and control strategy; and issuing the cooperative regulation and control strategy to the building group, the electric vehicle and the energy storage equipment for execution, monitoring the execution effect in real time, and dynamically adjusting the strategy. According to the invention, the energy utilization efficiency is improved, the carbon emission is reduced, and the stable and reliable operation of the system in a high-heat and high-humidity environment is guaranteed.

Description

Technical Field

[0001] This invention belongs to the field of supply and demand interaction technology in high-heat and high-humidity areas, specifically relating to a flexible resource stochastic robust optimization regulation method and system in high-heat and high-humidity areas. Background Art

[0002] Currently, in the hot and humid summer environment, power system regulation still relies mainly on traditional centralized dispatch and single-resource response, lacking a coordinated response mechanism for diverse energy-consuming terminals, making it difficult to effectively address sudden supply-demand imbalances. Specifically, on the one hand, the concentrated activation of cooling loads such as air conditioners and dehumidifiers in building complexes leads to a significant increase in grid load, widening the peak-to-valley difference and continuously increasing the pressure on power supply security. On the other hand, with the rapid popularization of electric vehicles, their distributed access and user-driven charging behavior have disrupted the traditional temporal patterns of load, further exacerbating the randomness of grid load and the risk of superimposed peak loads.

[0003] It is worth noting that both buildings and electric vehicles possess a certain degree of flexible regulation capability. For example, air conditioning loads can be appropriately shifted, and electric vehicles have energy storage capabilities; both have significant potential to become demand response resources. However, at present, there is a lack of comprehensive regulation mechanisms for high-intensity weather scenarios and coordinated responses to multiple types of resources.

[0004] Existing building load regulation technologies primarily focus on building energy consumption modeling and air conditioning load optimization under the premise of ensuring comfort. Typical methods include predictive control based on thermal dynamic models (such as MPC) or energy-optimal control strategies based on regression models. While these methods emphasize modeling the temperature control patterns of the building itself, their adaptability to external disturbances (such as sudden temperature changes and electricity price fluctuations) is limited, and they lack scalability in the coordinated regulation of multiple building clusters. In addition, there has been some research progress in electric vehicle load regulation, mainly focusing on strategies such as orderly charging and V2G regulation, emphasizing the time-series power scheduling of individual or aggregated electric vehicle groups.

[0005] In terms of cross-resource coordination, some integrated energy systems or park-level dispatching schemes have attempted to introduce multi-energy coordinated control mechanisms, but these focus more on the coupling of multiple energy flows such as electricity, heat, cooling, and storage, with few studies involving the coordinated response of buildings and electric vehicles. Even those schemes that do cover building-vehicle coordination often adopt static control boundaries and deterministic scenario assumptions, exhibiting significant lack of robustness when facing uncertainties such as sudden weather changes, reduced photovoltaic output, and real-time electricity price changes.

[0006] In summary, the current regional energy system regulation under high heat and humidity environments suffers from the following shortcomings: First, there is a lack of multi-source uncertainty modeling capabilities. Distributed photovoltaic power is greatly affected by meteorological fluctuations, building air conditioning loads are complexly affected by temperature and humidity coupling, and electric vehicles are strongly driven by user behavior, all three exhibiting high randomness. Second, the collaborative regulation mechanism between buildings and electric vehicles is weak. Existing strategies mostly focus on single-type resources and lack aggregation mechanisms and real-time coordination logic oriented towards source-load coordination. Third, robustness and generalization capabilities are limited. Traditional regulation methods are mostly static or semi-static decisions, which are difficult to adapt to situations with frequent disturbances during operation. Fourth, there is a lack of closed-loop feedback capabilities and real-time dynamic optimization capabilities. Strategies mostly rely on predictive models or rule bases, making it difficult to achieve integrated closed-loop control of "perception-decision-feedback". Summary of the Invention

[0007] Based on the aforementioned shortcomings and deficiencies in the existing technology, one of the objectives of this invention is to at least solve one or more of the aforementioned problems in the existing technology. In other words, one of the objectives of this invention is to provide a flexible resource stochastic robust optimization control method and system for high-heat and high-humidity areas that meets one or more of the aforementioned requirements, so as to improve the adaptability of resource control schemes, reduce grid load peaks, improve energy utilization efficiency, reduce carbon emissions, and achieve the goal of ensuring the stable and reliable operation of the system in high-heat and high-humidity environments.

[0008] To achieve the above-mentioned objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides a method for the stochastic robust optimization and control of flexible resources in high-heat and high-humidity areas, comprising the following steps: S1, collecting flexible resource data, power grid data, and meteorological data within the target area through a cloud-edge-device architecture, wherein the flexible resource data includes real-time operational data of building clusters, electric vehicles, distributed photovoltaics, and decentralized energy storage, and the power grid data includes power grid dispatch data and real-time electricity prices; S2, constructing a deterministic optimization model based on the flexible resource data, power grid data, and meteorological data, with the objective of minimizing the total daily operating cost of microgrids within the target area, wherein the total cost includes microgrid... The costs of gas turbine power generation, distributed energy storage operation, demand response adjustable load regulation, resource and grid electricity purchase and sale costs, and carbon emission costs are considered. S3: Based on the deterministic optimization model constructed in step S2, an uncertain set of photovoltaic output and load power, along with Monte Carlo simulation of real-time electricity prices, are introduced to construct a stochastic robust optimization model. This model is then iteratively solved using a column constraint generation algorithm to generate a collaborative regulation strategy. S4: The collaborative regulation strategy is distributed to building clusters, electric vehicles, and energy storage devices via a cloud-edge-device architecture for execution. The execution effect is monitored in real time, and the strategy is dynamically adjusted.

[0009] As a preferred embodiment, the real-time operational data of the building complex includes real-time values ​​of building complex temperature, humidity, and air conditioning load; the real-time operational data of the electric vehicles includes real-time charging status, battery charge status, and user activity plans, charging needs, and charging preferences; the real-time operational data of the distributed photovoltaic system includes real-time photovoltaic power generation; the real-time operational data of the distributed energy storage system includes distributed energy storage SOC, real-time charging and discharging status, and power; and the meteorological data includes real-time temperature, humidity, and light intensity of the target area.

[0010] As a preferred approach, the objective function of the deterministic optimization model constructed in step S2 is:

[0011] In the formula, This indicates the power generation cost of a micro gas turbine. This indicates the operating cost of distributed energy storage devices. This represents the control cost of demand-response adjustable load. This indicates the cost of purchasing and selling electricity from the main grid. This indicates the cost of carbon emissions.

[0012] As a preferred embodiment, the carbon emission cost is calculated using an hourly carbon emission factor, and the calculation formula is as follows:

[0013] In the formula, express t The carbon emission factor coefficient at time. express The day-ahead trading price of the time-of-use distribution network, and They represent The power purchase and sale of electricity by the microgrid in the commercial park during a given time period.

[0014] As a preferred embodiment, the uncertainty set of the photovoltaic output and load power is defined by the following formula.

[0015] In the formula, and They represent the results after considering uncertainty. Uncertain variables in photovoltaic power generation and load power consumption during different time periods. and They represent Forecast values ​​of photovoltaic power generation and load power consumption during the time period. and These represent the maximum prediction deviations for photovoltaic power output and load power, respectively. The Monte Carlo simulation of the real-time electricity price generates random samples based on the probability distribution of historical electricity price data.

[0016] As a preferred embodiment, before step S3, which iteratively solves the problem using the column constraint generation algorithm, the method further includes the step of decomposing the problem of solving the stochastic robust optimization model into an outer problem and an inner problem with mixed integer linear form.

[0017] Secondly, this invention provides a flexible resource stochastic robust optimization control system for high-heat and high-humidity areas, used to implement the flexible resource stochastic robust optimization control method as described in the first aspect, comprising: a data acquisition module, used to acquire real-time operating data of building complexes, electric vehicles, distributed photovoltaics and energy storage through end-side devices, and to acquire grid data and meteorological data through a cloud-side interface; an edge computing module, used to aggregate the data acquired by the data acquisition module to construct a deterministic optimization module and generate short-term load forecasts; an optimization decision module, used to construct and solve a stochastic robust optimization model to generate a collaborative control strategy; and an execution feedback module, used to distribute the strategy to terminal devices and monitor the execution effect in real time.

[0018] As a preferred embodiment, the optimization decision module includes: a deterministic optimization unit for calculating a basic optimization scheme without considering uncertainties; a robust expansion unit for introducing uncertainties in photovoltaics, load, and electricity prices, simulating real-time electricity price distribution, and generating anti-interference strategies; and a carbon emission optimization unit for adjusting the power purchase and sale plan according to the carbon emission factor.

[0019] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention constructs a deterministic optimization model and introduces carbon emission factor parameters and unit control cost parameters (i.e., interactive incentives) for demand-response adjustable loads into the model's optimization objective. Based on this, it further considers the uncertainties of photovoltaic power generation and load, as well as the volatility of real-time electricity trading prices, to construct a stochastic robust optimization model. This model can fully leverage the synergistic interaction capabilities of multiple types of flexible resources on the user side, effectively addressing the supply-demand imbalance in high-heat and high-humidity regions, significantly improving the system's flexible regulation capabilities, substantially reducing system carbon emissions, effectively alleviating grid operation pressure, and comprehensively improving grid security and stability. Compared to traditional schemes that only target a single type of resource for regulation, this model offers a richer variety of resource types and a wider range of applications.

[0020] 2. This invention constructs an uncertainty set for photovoltaic (PV) and load output forecasting using PV and load forecasting data and forecast deviation range data. By employing linearization and strong duality theory, the problem is linearized, transforming the complex problem into a linear one. This successfully quantifies the uncertainty of PV and load forecasting into "uncertain adjustment parameters" for PV output and load power. This significantly enhances the robustness and resilience of flexible resource regulation schemes. Compared to traditional optimized regulation schemes, it can more readily cope with the impact of source-load randomness under high temperature and high humidity environments, demonstrating the adaptability and economic advantages of the regulation strategy.

[0021] 3. This invention employs the Monte Carlo method to simulate electricity price fluctuations. By deeply analyzing historical electricity price data, its fluctuation characteristics are accurately extracted and the fluctuation range is determined. Probabilistic statistical methods are used to meticulously model electricity price volatility, successfully fitting its probability distribution characteristics. Subsequently, random sampling is performed based on the constructed probability distribution to generate electricity price samples that conform to actual market characteristics. This process provides a more scientific and accurate basis for resource regulation, helping to coordinate regulation while ensuring the comfort of building complexes and the charging needs of electric vehicle users. This effectively meets users' actual needs, optimizes energy utilization, and significantly improves the overall user experience.

[0022] Further or more detailed beneficial effects will be described in conjunction with specific embodiments in the detailed implementation. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart illustrating the flexible resource stochastic robust optimization and control method according to an embodiment of the present invention.

[0025] Figure 2 This is a schematic diagram of the cloud-edge-device architecture described in an embodiment of the present invention.

[0026] Figure 3 This is a schematic diagram of the process of iteratively solving a stochastic robust optimization model using the column constraint generation algorithm described in an embodiment of the present invention. Detailed Implementation

[0027] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0028] In the following description, several embodiments of the present invention are provided. Different embodiments can be substituted or combined. Therefore, the present invention can also be considered to include all possible combinations of the same and / or different embodiments described. Thus, if one embodiment includes features A, B, and C, and another embodiment includes features B and D, then the present invention should also be considered to include embodiments containing one or more other possible combinations of A, B, C, and D, even if such embodiments are not explicitly described in the following text.

[0029] The following description provides examples and does not limit the scope, applicability, or examples set forth in the claims. Changes may be made to the function and arrangement of the described elements without departing from the scope of the invention. Various processes or components may be appropriately omitted, substituted, or added to the various examples. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Furthermore, features described with respect to some examples may be combined into other examples.

[0030] To facilitate a better understanding of the embodiments of the present invention, its application scenarios will be explained before providing a detailed explanation of the specific implementation methods.

[0031] The flexible resource stochastic robust optimization control method described in the embodiments of this specification is applied to energy management systems in high-heat and high-humidity areas, particularly for the coordinated optimization control of building complexes, electric vehicle charging networks, and distributed photovoltaic and energy storage systems. In these scenarios, the application of the flexible resource stochastic robust optimization control method aims to address supply and demand imbalances under extreme climate conditions, enhance the synergistic interaction capabilities of multiple types of flexible resources, strengthen robustness under uncertain environments, reduce system carbon emissions, and ensure user comfort and charging needs.

[0032] The following is a brief explanation of the high-temperature and high-humidity areas, cloud-edge-device architecture, flexible resources, carbon emission factors, Monte Carlo simulation, and column constraint generation algorithms involved in several embodiments of this specification: High-temperature and high-humidity regions typically refer to geographical areas with hot climates and high humidity. In my country, these regions are mainly concentrated in the middle and lower reaches of the Yangtze River (Hunan, Hubei, Jiangxi, Anhui, Jiangsu, and Zhejiang), the southeastern coastal areas (Fujian, Guangdong, and Hainan), and the Sichuan Basin (Sichuan and Chongqing). Under these conditions, the contradiction between power supply and demand is more pronounced. On the one hand, high temperatures lead to the extensive use of air conditioning and other cooling equipment, resulting in a sharp increase in electricity load; on the other hand, high humidity can affect the normal operation of power equipment, reducing equipment efficiency and increasing the risk of equipment failure. Simultaneously, high-temperature and high-humidity conditions also pose challenges to energy storage and transmission, such as affecting battery performance and lifespan. The flexible resource stochastic robust optimization control method of this invention is specifically designed for these special circumstances in high-temperature and high-humidity regions. By synergistically controlling multiple flexible resources, it effectively alleviates the supply-demand tension and ensures the stable operation of the power system.

[0033] The cloud-edge-device architecture is a novel computing and data processing architecture. "Cloud" refers to cloud computing centers, which possess powerful computing capabilities and massive data storage capacity, enabling centralized processing and analysis of large-scale data to provide decision support and global optimization strategies for the entire system. "Edge" refers to the edge computing layer, located close to the data source, such as substations and distributed energy sites. Edge computing can perform preliminary data processing and analysis locally, reducing data transmission latency, improving system response speed, and alleviating the burden on cloud computing centers. "Device" represents various terminal devices, such as smart meters, electric vehicle charging stations, and smart home devices; these devices are the source of data generation and the execution terminals for control. Through layered collaboration, the cloud-edge-device architecture achieves efficient data processing and optimized resource allocation, providing strong technical support for flexible resource-based, stochastic, and robust optimization and control.

[0034] Flexible resources refer to resources in a power system that possess flexible adjustment capabilities, enabling rapid and accurate response and adjustment based on system needs at different time and spatial scales. The flexible resources discussed in this specification mainly include building loads, distributed photovoltaics, electric vehicles, and distributed energy storage. Building loads can adjust the power consumption of equipment such as air conditioners and lighting through intelligent control systems, achieving flexible load adjustment; distributed photovoltaics can adjust power generation in real time according to sunlight intensity and can store and release energy through energy storage devices; electric vehicles, as mobile energy storage devices, can have their charging and discharging behavior rationally arranged according to grid demand; distributed energy storage systems can store electrical energy during off-peak hours and release it during peak hours, playing a role in peak shaving and valley filling. The synergistic interaction of these flexible resources can effectively improve the flexibility and reliability of the power system.

[0035] The carbon emission factor refers to the amount of carbon dioxide emitted per unit of energy during combustion or use. In power systems, different power generation methods and energy types have different carbon emission factors. For example, coal-fired power generation has a relatively high carbon emission factor, while renewable energy sources such as solar and wind power have almost zero carbon emission factors. The flexible resource stochastic robust optimization control method of this invention introduces the carbon emission factor parameter into the model optimization objective. By considering the carbon emissions of different energy sources, it guides the system to prioritize the use of low-carbon energy, optimizes the energy structure, and thereby reduces the carbon emissions of the entire power system, achieving green and sustainable development.

[0036] Monte Carlo simulation is a numerical computation method based on probability and statistics that simulates the behavior and characteristics of complex systems through random sampling and statistical experiments. In this invention, the Monte Carlo method is used to simulate electricity price fluctuations. Specifically, historical electricity price data is first analyzed to extract its fluctuation characteristics and determine the fluctuation range. Then, probabilistic and statistical methods are used to model the electricity price volatility and fit its probability distribution characteristics. Finally, random sampling is performed based on the constructed probability distribution to generate a large number of electricity price samples that conform to actual market characteristics. These samples can be used to evaluate the effectiveness of flexible resource regulation strategies under different electricity price scenarios, providing decision-makers with a scientific basis and improving the adaptability and economy of regulation strategies.

[0037] Column constraint generation algorithms are an effective method for solving large-scale optimization problems, especially those with complex constraints. In the flexible resource stochastic robust optimization and control method of this invention, numerous complex constraints may be involved, such as power constraints for photovoltaics and loads, charging and discharging constraints for electric vehicles, and capacity constraints for energy storage systems. Column constraint generation algorithms decompose the original optimization problem into a series of smaller subproblems by progressively generating and filtering column vectors that satisfy the constraints, thereby reducing the difficulty of solving the problem and improving efficiency. This method can quickly find the optimal solution that satisfies all constraints while ensuring solution accuracy, providing reliable decision support for the optimization and control of flexible resources.

[0038] Example 1: like Figure 1 As shown in the figure, this embodiment provides a flexible resource stochastic robust optimization control method for high-heat and high-humidity areas. The specific steps are as follows: Step S1 involves collecting flexible resource data, power grid data, and meteorological data within the target area through a cloud-edge-device architecture. The flexible resource data includes real-time operational data of building clusters, electric vehicles, distributed photovoltaics, and decentralized energy storage. The power grid data includes power grid dispatch data and real-time electricity prices.

[0039] Specifically, the real-time operational data of the building complex includes real-time temperature, humidity, and air conditioning load; the real-time operational data of the electric vehicles includes real-time charging status, battery charge status, user activity plans, charging needs, and charging preferences; the real-time operational data of the distributed photovoltaic system includes real-time photovoltaic power generation; the real-time operational data of the distributed energy storage system includes distributed energy storage SOC, real-time charging and discharging status, and power; and the meteorological data includes real-time temperature, humidity, and light intensity of the target area.

[0040] More specifically, this embodiment provides a method based on, for example, Figure 2 The preferred implementation method for data collection and monitoring using the cloud-edge-device architecture shown includes the following steps: 1. On the edge, data collection and monitoring terminals are used to collect on-site data and execution feedback from building complexes, electric vehicles, distributed photovoltaic systems, and decentralized energy storage systems: 1) Building Complex Data Acquisition: Temperature sensors, humidity sensors, and air conditioning load monitoring equipment are installed in various rooms and public areas of the building complex. The collected data is transmitted to the data processing center in real time via a wireless communication network. 2) Electric Vehicle Data Collection: Install charging pile monitoring equipment at electric vehicle charging stations to obtain real-time charging status and battery information of electric vehicles. Collect user travel plans and charging preference information through communication with electric vehicle users' mobile applications or in-vehicle systems; 3) Distributed photovoltaic data acquisition: Install a data acquisition terminal at the distributed photovoltaic controller to obtain photovoltaic power generation data; 4) Distributed energy storage data acquisition: Install a data acquisition and monitoring terminal at the distributed energy storage control cabinet to obtain data such as distributed energy storage SOC, charge and discharge status, and power.

[0041] 2. On the cloud side, data is acquired through interfaces with third-party systems such as power grids, trading systems, and meteorological systems: 1) Power Grid Data Acquisition: Establish a data interface with the power grid dispatch center to acquire real-time power grid operation status information, future load forecasts, and power generation plan data; 2) Energy trading data: real-time electricity price data, etc. 3) Meteorological data: temperature, humidity, light intensity, etc.

[0042] 3. Edge-side: Aggregates data from both the cloud and the client, performs data prediction based on historical data and user industry data, and monitors the strategy's operational status, providing real-time feedback on execution. 1) Building cluster energy demand forecasting: Using historical meteorological data and building cluster energy consumption data, a machine learning model is trained to forecast the energy demand of the building cluster in the future. 2) Electric vehicle charging demand forecasting: Utilizing users' travel plans and charging demand preferences, forecast electric vehicle charging demand over a period of time. 3) Execution feedback: Analyze data during the strategy execution period in real time and predict abnormal states throughout the entire "start-run-end" cycle of the strategy.

[0043] Step S2 involves constructing a deterministic optimization model based on the flexible resource data, power grid data, and meteorological data, with the objective of minimizing the total daily operating cost of the microgrid within the target area. The total cost includes the power generation cost of the micro gas turbine, the operating cost of the distributed energy storage device, the regulation cost of the demand-response adjustable load, the cost of purchasing and selling electricity from the main grid, and the cost of carbon emissions.

[0044] Specifically, the objective function of the deterministic optimization model constructed in step S2 is: (1) In equation (1), This indicates the power generation cost of a micro gas turbine. This indicates the operating cost of distributed energy storage devices. This represents the control cost of demand-response adjustable load. This indicates the cost of purchasing and selling electricity from the main grid. This indicates the cost of carbon emissions.

[0045] Understandably, micro gas turbines can not only provide stable power output as a distributed power source, but also achieve flexible response to fluctuations in renewable energy generation and load changes through rapid start-up, shutdown and output adjustment, thereby improving power supply reliability. Micro gas turbines are of great value in improving the energy utilization efficiency, economy and stability of microgrids.

[0046] More specifically, the cost of power generation by micro gas turbines in a microgrid can be expressed as: (2) In equation (2), express Output power of a micro gas turbine during a given period, measured in kW. and This represents the cost coefficient, expressed in yuan / kWh.

[0047] More specifically, the output power constraint of the micro gas turbine is (3) In equation (3), and These represent the upper and lower limits of the output of the micro gas turbine, respectively, in kW.

[0048] More specifically, the operating cost of distributed energy storage devices It can be represented as (4) In equation (4), This represents the converted charging and discharging cost of distributed energy storage devices, expressed in yuan / kWh. This indicates the charge and discharge efficiency of a distributed energy storage device. and They represent Discharge and charge power of time-distributed energy storage devices, in kW.

[0049] More specifically, distributed energy storage devices must meet constraints on energy storage charging and discharging power, energy storage state of charge, and charging and discharging balance during operation.

[0050] Energy storage charging and discharging power constraints are (5) In equation (5), This indicates the upper limit of the charging and discharging power of distributed energy storage, in kW. express The charging and discharging status of distributed energy storage during different time periods is represented by a value of 1 indicating charging and a value of 0 indicating discharging.

[0051] Energy storage state of charge constraints are (6) In equation (6), and These represent the upper and lower limits of the state of charge for distributed energy storage, in kWh. This represents the initial energy capacity of distributed energy storage, expressed in kWh. This indicates the charge and discharge efficiency of distributed energy storage. The time step is indicated; in this embodiment, the time step is 1 hour.

[0052] The charge / discharge balance constraint is (7) It is understandable that adjusting flexible loads will change electricity consumption plans and thus affect user comfort. Therefore, the cost of compensating for user comfort is expressed as... (8) In equation (8), This represents the unit control cost of demand-response adjustable load, expressed in yuan / kWh. express The actual power consumption of time-of-use demand response adjustable loads, in kW. express Forecast power consumption of time-dependent demand response adjustable load, in kW.

[0053] More specifically, the operational constraints of demand response load are: (9) (10) In equation (10), and These represent the adjustable power range of the demand response load, in kW.

[0054] Understandably, commercial park microgrids can supplement insufficient power through power purchase via tie lines or absorb surplus power through power sales, thereby optimizing operating costs and improving their operational stability. The tie lines between the microgrid and the distribution network must meet the following conditions. (11) In equation (11), express The day-ahead trading price for time-of-use distribution networks is expressed in yuan / kW. and They represent The electricity purchase and sale power of the commercial park microgrid during the specified time period, in kW.

[0055] Operational constraints of the tie lines between microgrids and distribution networks include (12) (13) In equation (12), express The typical electrical load power consumption of the commercial park microgrid during a given time period, expressed in kW. express Photovoltaic power generation during a given time period, expressed in kW.

[0056] In equation (13), This indicates the upper limit of the interaction power of the tie line between the microgrid and the distribution network, in kW. express The state variables of electricity purchase and sale in the commercial park microgrid during specific time periods. A value of 1 represents purchasing electricity. A value of 0 indicates electricity sales.

[0057] When the uncertainties of photovoltaic output and load power are not considered, a deterministic optimization model for the optimal operation of the park's microgrid can be obtained, and its compact form can be expressed as follows: (14) In equation (14), and To optimize the variables, the specific expression is as follows: (15) In equation (15), the day-ahead transaction price and real-time price of the microgrid and distribution network in the commercial park are the same in the deterministic optimization model, that is, the coefficient column vector in the objective function. This is a constant column vector. The first row of constraints represents the inequality constraints in the deterministic optimization model, corresponding to equations (3), (6), and (10); the second row represents the equality constraints in the deterministic optimization model, including equations (3), (7), and (12); the third row corresponds to equations (5) and (13); the fourth row represents the predicted values ​​of photovoltaic output and load power for each time period in the deterministic optimization model, i.e., ... (16) In equation (16), and They represent The predicted values ​​of photovoltaic power generation and load power consumption for the time period are in kW.

[0058] It is understandable that the energy sources for generating electricity differ at different times, resulting in varying carbon emission factors for electricity across different time periods. While incentivizing various flexible resources to participate in grid interaction and enhancing grid regulation capabilities, this embodiment also considers low-carbon requirements, calculating the carbon emission cost of the entire interactive system using hourly carbon emission factors. The calculation formula is as follows: (17) In equation (17), express t The carbon emission factor coefficient at time. express The day-ahead trading price for time-of-use distribution networks is expressed in yuan / kW. and They represent The power purchased and sold by the microgrid in the commercial park during a given time period, in kW.

[0059] Step S3 involves introducing the uncertainty set of photovoltaic output and load power, as well as Monte Carlo simulation of real-time electricity price, into the deterministic optimization model constructed in step S2 to build a stochastic robust optimization model. This model is then iteratively solved using a column constraint generation algorithm to generate a coordinated control strategy.

[0060] It is understandable that deterministic optimization methods rely on the accuracy of the forecast curves for the next day's photovoltaic power output and load power consumption when formulating control schemes. However, the microgrids upon which these resources depend are subject to various random factors in actual operation, making it difficult for microgrid decision-makers to achieve accurate forecasts and thus affecting actual economic benefits. The uncertainty in photovoltaic and load output forecasts within the microgrid mainly stems from weather changes, environmental factors, and the randomness of user electricity consumption behavior. Photovoltaic power generation is affected by meteorological factors such as solar irradiance, temperature, and cloud cover, resulting in significant fluctuations in output that are difficult to predict accurately. Load demand is affected by factors such as weather, season, time, human behavior, and economic activities, exhibiting randomness and dynamic changes, leading to forecasting errors. Furthermore, the volatility of real-time electricity prices is mainly determined by factors such as changes in the supply and demand relationship in the electricity market, fluctuations in renewable energy output, adjustments to dispatch strategies, and grid operation constraints. Due to the time-varying and complex nature of the electricity market, there may be discrepancies between the trading plans formulated by the microgrid in the day-ahead market and the actual operating conditions, further exacerbating the uncertainty of real-time market transactions and the volatility of electricity purchase and sale costs. Therefore, it is necessary to consider the uncertainties in photovoltaic and load output forecasts as well as the volatility of real-time electricity trading prices in the model.

[0061] To address the volatility of real-time electricity trading prices, considering that trading prices fluctuate around the theoretical price across different time periods and are influenced by various factors such as supply and demand, market competition, policy regulation, and fluctuations in renewable energy output, this embodiment employs the Monte Carlo method for simulation. By analyzing historical electricity trading price data, its volatility characteristics are extracted and the fluctuation range is determined. Probabilistic statistical methods are used to model the price volatility, fitting its probability distribution characteristics. Then, random sampling is performed based on the constructed probability distribution to generate electricity price samples that conform to actual market characteristics. This method effectively characterizes the randomness and uncertainty of real-time electricity trading prices, providing a more reliable decision-making basis for microgrid day-ahead regulation and economic dispatch. The coefficient column vector in the objective function is obtained through sampling using this method. Electricity trading price in Then, regarding the uncertainties in photovoltaic and load output forecasting, the uncertainty set is determined using equation (18). To describe, its expression is (18) In equation (18), and They represent the results after considering uncertainty. Uncertain variables in photovoltaic power generation and load power consumption during a given time period, in kW. and These represent the maximum prediction deviations for photovoltaic output and load power, respectively, in kW.

[0062] Specifically, the above processing on the deterministic optimization model results in a stochastic robust optimization model. To further eliminate the impact of prediction errors, the problem of solving the optimization model becomes finding the optimal variable that minimizes the objective function as the uncertain variable u approaches its range boundary value. The model simplifies to... (19) Indicates a given set Time optimization variables The feasible region, expressed as (20) The simplified stochastic robust optimization model (19) is decomposed into two layers of problems to be solved, inner and outer. For each set of problems in the inner layer... Each outer-layer problem has an optimal solution, which can be obtained through iterative solving. Therefore, the model is decomposed and expressed as follows: (twenty one) In equation (21), Indicates the current iteration number. Indicates the first The solution to the inner problem after the next iteration Indicates the first The uncertainty variables obtained after the second iteration under the worst-case scenario The value of .

[0063] Meanwhile, the inner problem is expressed as (twenty two) The objective function minimization problem in the decomposed inner layer is a linear problem. Based on strong duality theory, it can be transformed into a max problem and merged with the max problem in the outer layer. The merged problem is expressed as follows: (twenty three) Since equation (22) contains a bilinear term, the optimal solution corresponds to... For an uncertain set One extreme point, that is, when equation (22) reaches its maximum value, is the uncertainty variable. The value of should be the boundary of the uncertain interval described by equation (18).

[0064] Equation (18) can be rewritten in the following form: (twenty four) In the model, and As an "uncertainty adjustment parameter" for photovoltaic output and load power, its value ranges from 0 to 24, being an integer. When, it means that the uncertain variables in each time period are equal to the predicted values, which is equivalent to a deterministic optimization model; when When the uncertain variables in each time period reach the interval boundary value, that is, the case with the largest prediction deviation.

[0065] Based on this, by introducing auxiliary variables for linearization, the problem of solving the stochastic robust optimization model is decomposed into an outer problem (21) and an inner problem (25) with mixed integer linear form. (25) The solution is then obtained using a column constraint generation algorithm; see reference [link / reference]. Figure 3 The process is as follows: The first step is to give a set of uncertain variables. The value of is used as the initial worst-case scenario to set the lower bound of the operating cost corresponding to the final control plan. Upper Realm Number of iterations ; The second step is based on the worst-case scenario. Solving the main problem equation (20), we obtain the optimal solution. The objective function value of the main problem is used as the new lower bound.

[0066] The third step is to obtain the solution to the main problem. Substituting into equation (24), we solve the subproblem and obtain the objective function value of the subproblem. and corresponding worst-case scenario uncertainties The value of Update the upper boundary ; Fourth step, given the convergence threshold of the algorithm is ,like If the iteration stops, return the optimal solution. and Otherwise, add variables. and the following constraints: (26) make Proceed to the second step until the algorithm converges.

[0067] Step S4 involves distributing the collaborative control strategy to building clusters, electric vehicles, and energy storage devices via a cloud-edge-device architecture, monitoring the execution effect in real time, and dynamically adjusting the strategy.

[0068] Specifically, this embodiment establishes a collaborative control cloud service platform that integrates functional modules such as data processing, strategy formulation, and command transmission. The collaborative control cloud service platform receives data from building clusters, electric vehicles, and the power grid, and interfaces with systems such as scheduling, trading, and meteorology to acquire and analyze the data. Based on the established collaborative control strategy, it sends control commands to the edge control terminal, which then transmits the signals to the specific regulating equipment.

[0069] Specifically, during the coordinated control process, the operating status and energy demand changes of each subsystem are monitored in real time through terminals and the system. When meteorological conditions change suddenly (such as a sharp rise in temperature) or equipment malfunctions (such as air conditioning equipment failure), energy demand and supply capacity are reassessed in a timely manner, and control strategies are adjusted through the coordinated control platform. The final implementation results are uploaded to the cloud service platform through the terminals, serving as the basis for interactive effect analysis and subsidy distribution.

[0070] Example 2: This embodiment provides a flexible resource stochastic robust optimization control system for high-heat and high-humidity areas, used to implement the flexible resource stochastic robust optimization control method as described in Embodiment 1. The system includes: a data acquisition module for collecting real-time operational data from building clusters, electric vehicles, distributed photovoltaic systems, and energy storage via end-side devices, and acquiring grid data and meteorological data through a cloud-side interface; an edge computing module for aggregating the data collected by the data acquisition module to construct a deterministic optimization module and generate short-term load forecasts; an optimization decision module for constructing and solving a stochastic robust optimization model to generate a collaborative control strategy; and an execution feedback module for distributing the strategy to terminal devices and monitoring the execution effect in real time.

[0071] Specifically, this embodiment provides a preferred implementation, wherein the optimization decision module includes: a deterministic optimization unit for calculating a basic optimization scheme without considering uncertainties; a robust expansion unit for introducing uncertainties in photovoltaics, load, and electricity prices, simulating real-time electricity price distribution, and generating an anti-interference strategy; and a carbon emission optimization unit for adjusting the power purchase and sale plan according to the carbon emission factor.

[0072] It should be noted that for the aforementioned method embodiments, for simplicity of description, they are all expressed as a series of action combinations. However, those skilled in the art should be aware that the present invention is not limited by the order of the actions described, because according to the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0073] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0074] The above description is merely an exemplary embodiment of the present invention and should not be construed as limiting the scope of the invention. Any equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of embodiments of the invention upon considering the specification and practicing the disclosure herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not described herein. The specification and embodiments are to be considered exemplary only, and the scope and spirit of the invention are defined by the claims.

Claims

1. A flexible resource stochastic robust optimization and regulation method for high-heat and high-humidity areas, characterized in that, Including the following steps: S1. Collect flexible resource data, power grid data, and meteorological data within the target area through a cloud-edge-device architecture. The flexible resource data includes real-time operation data of building clusters, electric vehicles, distributed photovoltaics, and decentralized energy storage. The power grid data includes power grid dispatch data and real-time electricity prices. S2. Based on the aforementioned flexible resource data, power grid data, and meteorological data, a deterministic optimization model is constructed with the objective of minimizing the total daily operating cost of the microgrid within the target area. The total cost includes the power generation cost of the micro gas turbine, the operating cost of the distributed energy storage device, the regulation cost of the demand response adjustable load, the cost of purchasing and selling electricity from the main grid, and the cost of carbon emissions. S3. Based on the deterministic optimization model constructed in step S2, introduce the uncertainty set of photovoltaic output and load power and Monte Carlo simulation of real-time electricity price to construct a stochastic robust optimization model, and solve iteratively through column constraint generation algorithm to generate a coordinated control strategy. S4. The collaborative control strategy is distributed to building clusters, electric vehicles and energy storage devices through a cloud-edge-device architecture, and the execution effect is monitored in real time and the strategy is dynamically adjusted.

2. The flexible resource stochastic robust optimization and regulation method for high-heat and high-humidity areas according to claim 1, characterized in that: The real-time operating data of the building complex includes the real-time temperature, humidity, and air conditioning load of the building complex. The real-time operating data of the electric vehicle includes the real-time charging status of the electric vehicle, the real-time battery charge status of the electric vehicle, as well as the electric vehicle user's travel plans, charging needs, and charging preferences. The real-time operating data of the distributed photovoltaic system includes the real-time value of photovoltaic power generation. The real-time operating data of the distributed energy storage includes the distributed energy storage SOC, the real-time charging and discharging status of the distributed energy storage, and the real-time power value of the distributed energy storage. The meteorological data includes real-time temperature, real-time humidity, and real-time light intensity of the target area.

3. The flexible resource stochastic robust optimization and regulation method for high-heat and high-humidity areas according to claim 2, characterized in that: The objective function of the deterministic optimization model constructed in step S2 is: In the formula, This indicates the power generation cost of a micro gas turbine. This indicates the operating cost of distributed energy storage devices. This represents the control cost of demand-response adjustable load. This indicates the cost of purchasing and selling electricity from the main grid. This indicates the cost of carbon emissions.

4. The flexible resource stochastic robust optimization and regulation method for high-heat and high-humidity areas according to claim 3, characterized in that: The carbon emission cost is calculated using hourly carbon emission factors, and the calculation formula is as follows: In the formula, express t The carbon emission factor coefficient at time. express The day-ahead trading price of the time-of-use distribution network, and They represent The power purchase and sale of electricity by the microgrid in the commercial park during a given time period.

5. The flexible resource stochastic robust optimization and control method for high-heat and high-humidity areas according to claim 4, characterized in that: The uncertainty set of photovoltaic output and load power is defined by the following formula. In the formula, and They represent the results after considering uncertainty. Uncertain variables in photovoltaic power generation and load power consumption during different time periods. and They represent Forecast values ​​of photovoltaic power generation and load power consumption during the time period. and These represent the maximum prediction deviations for photovoltaic power output and load power, respectively. The Monte Carlo simulation of the real-time electricity price generates random samples based on the probability distribution of historical electricity price data.

6. The flexible resource stochastic robust optimization and control method for high-heat and high-humidity areas according to claim 5, characterized in that, Before step S3, which iteratively solves the problem using the column constraint generation algorithm, the following steps are also included: The problem of solving the stochastic robust optimization model is decomposed into an outer problem and an inner problem with mixed integer linear form.

7. A flexible resource stochastic robust optimization and control system for high-heat and high-humidity areas, characterized in that, The method for implementing the flexible resource stochastic robust optimization control method as described in any one of claims 1 to 6 includes: The data acquisition module is used to collect real-time operational data of building complexes, electric vehicles, distributed photovoltaics and energy storage through end-side devices, and to obtain grid data and meteorological data through cloud-side interfaces; The edge computing module is used to aggregate the data collected by the data acquisition module to construct a deterministic optimization module and generate short-term load forecasts; The optimization decision module is used to construct and solve stochastic robust optimization models and generate collaborative control strategies. The execution feedback module is used to send policies to terminal devices and monitor the execution effect in real time.

8. A flexible resource stochastic robust optimization and control system for high-heat and high-humidity areas according to claim 7, characterized in that, The optimization decision module includes: Deterministic optimization unit, used to calculate basic optimization schemes without considering uncertainties; Robust extension units are used to introduce uncertainties in photovoltaics, loads, and electricity prices, and to simulate real-time electricity price distribution to generate anti-interference strategies; The carbon emission optimization unit is used to adjust the power purchase and sale plan based on the carbon emission factor.

Images