Smart community demand response rolling optimization scheduling method considering electric-thermal response difference

Abstract

The application discloses a smart community demand response rolling optimization scheduling method considering electric-thermal response difference, which comprises the following steps: collecting and processing smart community operation data; constructing an electric-thermal comprehensive energy system model of the smart community; dividing electric-thermal response resources and establishing response rate constraints; constructing a day-ahead-intra-day rolling optimization scheduling model; executing scheduling instructions and updating system state feedback; and analyzing a smart community demand response scheduling result based on the rolling optimization model. The method has clear process, can coordinate electric-side fast response resources and thermal-side slow regulation resources, reduce the influence of source and load prediction error on the scheduling result, improve demand response instruction tracking accuracy, suppress tie-line power fluctuation, improve renewable energy consumption capacity, and take into account user thermal comfort.

Description

Technical Field

[0001] This invention relates to the field of power system optimization scheduling technology, specifically to a smart community demand response rolling optimization scheduling method that takes into account the difference between electrical and thermal responses. Background Technology

[0002] Against the backdrop of new power system construction, smart communities are gradually integrating various resources such as photovoltaics, wind power, electric energy storage, heat pumps, gas boilers, thermal energy storage, flexible electricity, and thermal loads. Through coordinated electricity-heat dispatch and demand response, the capacity for renewable energy absorption and energy utilization efficiency can be improved. However, in actual operation, there are significant forecasting errors in photovoltaic and wind power output, as well as electricity and heat loads. Relying solely on day-ahead dispatch plans is insufficient to adapt to the random changes in intraday operating conditions in real time. This can easily lead to problems such as deviations in demand response command execution, increased power fluctuations in tie lines, and decreased user thermal comfort.

[0003] For the energy dispatch problem in smart communities, existing studies have proposed a variety of optimization methods. For example, reference [1] discloses an operation method for an integrated electric-thermal energy system based on day-ahead economic dispatch, which constructs a mixed integer linear programming model with the goal of minimizing system operating costs to determine the output plan of each device for the next day. However, this method is a static open-loop dispatch, which only makes decisions based on predicted data on the day-ahead time scale and fails to introduce an intraday rolling correction mechanism. When photovoltaic and wind power output or load demand fluctuates in real time, a large tracking deviation will occur between the tie-line power and the demand response target, which is difficult to meet the requirements of smart communities for response accuracy.

[0004] To address the uncertainty of forecasts, existing technologies have introduced a rolling optimization scheduling method based on model predictive control. This method continuously updates ultra-short-term forecast information and continuously optimizes the operation plan for future periods to reduce the impact of forecast errors on scheduling decisions. Reference [2] proposes a two-stage rolling scheduling architecture for active distribution networks, with the day-ahead and intraday phases. The intraday phase uses rolling time-domain optimization to perform closed-loop correction of the day-ahead plan, which improves the power balance capability under source-load forecast errors to a certain extent. However, this method treats electricity and heat resources as homogeneous adjustable units to participate in optimization without considering the essential differences in response characteristics between electricity and heat resources: electricity storage and flexible electricity loads have fast response speeds and are suitable for undertaking short-term power deviation correction and demand response tracking tasks; heat pumps, heat storage, adjustable heat loads, and building thermal inertia are constrained by the heat transfer process, resulting in slow response speeds and large inertia, making them more suitable for heat load balance and smooth maintenance of indoor temperature over longer time scales. Because the difference between the electric and thermal response rates is not distinguished, slow resources on the thermal side are often frequently used for power point tracking on a short time scale in actual dispatching. This makes it difficult for dispatching commands to be executed accurately by physical equipment, and it also increases the number of cycles of thermal energy storage devices and causes indoor temperatures to exceed limits. This reduces the feasibility of the dispatching scheme and also damages the user's energy comfort.

[0005] Reference [3] discloses a microgrid rolling optimization method that considers tie-line power smoothing. By introducing a penalty term for tie-line power change in adjacent time periods into the objective function, it effectively suppresses high-frequency fluctuations in the power interaction between the microgrid and the main grid. However, this method focuses on the rapid adjustment of power-side resources and tie-line smoothing control, ignoring the heat storage characteristics of adjustable resources on the thermal side, especially the thermal inertia of buildings and the constraints on user indoor thermal comfort. During the adjustment process, it cannot guarantee that the indoor temperature always meets the comfort range requirements. In practical engineering applications, it is difficult to balance grid-side friendliness and user-side comfort.

[0006] In summary, current rolling optimization scheduling methods for smart communities often fail to integrate and coordinate objectives such as demand response tracking accuracy, tie-line power smoothness, and thermal comfort. In particular, they lack differentiated response rate constraints for fast-response electrical resources and slow-regulation thermal resources, thus failing to effectively suppress tie-line power fluctuations and ensure user thermal comfort while improving demand response command tracking accuracy. Therefore, there is an urgent need to propose a rolling optimization scheduling method for smart community demand response that considers the differences between electrical and thermal responses. This method would address the shortcomings of existing scheduling methods by setting differentiated response rate constraints for fast-response electrical resources and slow-regulation thermal resources, and by constructing a day-ahead-intraday rolling optimization framework.

[0007] References: Reference [1]: Wang Chengshan, Li Peng, Yu Hao. Research on day-ahead economic dispatch method for integrated energy system [J]. Proceedings of the CSEE, 2018, 38(12): 3490-3502. Reference [2]: Liu Dong, Chen Fei, Zhang Peichao. Day-to-day multi-timescale coordinated optimization scheduling of active distribution network based on model predictive control [J]. Automation of Electric Power Systems, 2019, 43(15): 60-69. Reference [3]: Sun Hongbin, Guo Qinglai, Xin Shujun. Rolling optimization scheduling method for microgrids that takes into account the power smoothing of tie lines [J]. Automation of Electric Power Systems, 2020, 44(5): 85-93. Summary of the Invention To address the aforementioned technical problems, this invention provides a smart community demand response rolling optimization scheduling method that takes into account the differences between electricity and heat response. This method can coordinate the rapid response on the electricity side with the slow adjustment on the heat side, avoiding frequent adjustments of heat-side resources in a short period of time; improving the accuracy of demand response tracking and reducing tie-line power fluctuations; thereby enhancing the renewable energy absorption capacity and ensuring user thermal comfort.

[0008] The technical solution adopted in this invention is as follows: A rolling optimization scheduling method for demand response in smart communities that takes into account the difference between electrical and thermal responses includes the following steps: S1: Collect and preprocess the operation data of the smart community's integrated electric-thermal energy system to obtain input data for day-ahead scheduling and intraday rolling optimization; S2: Based on the equipment configuration and load characteristics in the input data, construct an integrated electric-thermal energy system model. This model includes electric power balance constraints, thermal power balance constraints, equipment operation constraints, and indoor thermal comfort constraints. S3: Based on the response speed of various adjustable resources, adjustable resources are divided into fast-response resources on the electrical side and slow-response resources on the thermal side, and electrical-side response rate constraints and thermal-side response rate constraints are established to limit the frequent use of slow-response resources on the thermal side within a short time scale. S4: Based on the integrated electric-thermal energy system model and the response rate constraints of the electric and thermal sides, a day-ahead and intraday rolling optimization scheduling model is constructed; in the day-ahead stage, the objective is to minimize the daily comprehensive operating cost, and a day-ahead basic scheduling plan is generated. The intraday phase uses a fixed-cycle rolling approach. Based on intraday ultra-short-term forecast data and real-time system status feedback, the intraday basic scheduling plan is dynamically modified with the goal of minimizing the weighted sum of demand response deviation, day-ahead plan tracking deviation, scheduling smoothness and thermal comfort deviation, resulting in an intraday rolling optimized scheduling scheme. S5: During each daily rolling optimization period, only the scheduling instructions for the current time period are executed. After the scheduling instructions for the current time period are completed, the actual operating status of the system is collected and fed back to the next rolling optimization cycle to form a closed-loop scheduling.

[0009] It also includes S6: Based on the actual operating data obtained from closed-loop scheduling, it calculates the demand response tracking error, day-ahead planning tracking error, tie-line power fluctuation, renewable energy absorption rate, and thermal comfort deviation index, and analyzes the scheduling effect.

[0010] S1 includes: S1.1: Collect day-ahead forecast data, intraday ultra-short-term forecast data, equipment operating parameters, electricity price information, demand response target power, and real-time system status data of the smart community's integrated electric-heat energy system; The day-ahead forecast data includes forecasts for photovoltaic power output, wind power output, electrical load, heat load, and outdoor temperature. The intraday ultra-short-term forecast data includes each forecast value updated within the current rolling optimization window; The system's real-time status data includes electrical energy storage state of charge (SOC), thermal energy storage state, indoor temperature, and tie-line power. S1.2: Perform time scale unification, abnormal data correction, missing data completion and normalization on the collected data to obtain 1-hour time scale data used in the day-ahead scheduling phase and 15-minute time scale data used in the intraday rolling optimization phase, and use 4 hours as the rolling optimization window length.

[0011] S2 includes: 2.1: Establish an integrated electric-thermal energy system model comprising energy supply units, conversion and storage units, load units, and external interaction units. The energy supply units include photovoltaic, wind power, upstream power grid, and natural gas; the conversion and storage units include electric energy storage, heat pumps, gas-fired boilers, and thermal energy storage; the load units include rigid electrical loads, flexible electrical loads, rigid thermal loads, and adjustable thermal loads; and the external interaction units are the interconnection lines between the community and the upstream power grid. 2.2: A first-order thermal inertia model is used to describe the relationship between indoor temperature changes in adjacent scheduling periods. The first-order thermal inertia model is shown in formula (1): (1); In formula (1): For time period The indoor temperature; Outdoor temperature; The effective heating power entering the building interior; a, b, and c are building thermal dynamic parameters, where: a represents the indoor temperature inertia maintenance capacity, b represents the regulating effect of heating power on indoor temperature, and c represents the influence of outdoor temperature disturbance on indoor temperature.

[0012] 2.3: Establish indoor thermal comfort constraints to keep the indoor temperature within the preset comfortable temperature range, as shown in formula (2): (2); In formula (2): and These are the lower and upper limits of the user's comfortable temperature, respectively.

[0013] 2.4: Establish power balance constraints on the electric side and the thermal side, and set capacity constraints, upper and lower power limits, state update constraints, and ramp constraints for electric energy storage, thermal energy storage, heat pumps, gas boilers, flexible electric loads, adjustable thermal loads, and tie-line power, respectively. These constraints can be uniformly expressed as... ,in: For a given time period, the set of system decision variables. This is the feasible operating domain for the device.

[0014] In section 2.4, the power balance constraint on the electric side is as shown in formula (3): (3); In formula (3): The power exchange between the community and the higher-level power grid; and The outputs are solar power and wind power, respectively. and These are the energy storage discharge and charging power, respectively. Total community load; This refers to the power consumption of the heat pump.

[0015] The heat-side power balance constraints are shown in equations (4) and (5): (4); (5); In the above formula: The heat pump outputs heat power; This is to output heat power for the gas-fired boiler; and These are thermal energy storage heat release and thermal storage power, respectively. For the community's heat load; The coefficient of performance (COP) of a heat pump; S3 includes: S3.1: Electric energy storage and flexible electric loads are classified as fast-response resources on the electric side, used to correct short-timescale power deviations and track demand response target power in intraday rolling optimization; heat pumps, thermal energy storage, adjustable heat loads and building thermal inertia are classified as slow-regulation resources on the thermal side, used to balance heat loads and perform slow-smooth regulation under the conditions of satisfying thermal power balance constraints and indoor thermal comfort constraints. S3.2: Define the net regulation power on the electric side Net regulating power on the heat side As shown in formulas (7) and (8) respectively: (7); (8); In the above formula: This refers to the net regulating power on the electric side. The power regulation for flexible electrical loads includes a combined regulation amount for transferable and load reduction. Net regulating power on the heat side; This refers to the adjustment amount of the heat pump's output heat power. This allows for adjustable heat load reduction or power transfer.

[0016] S3.3: Based on the net regulation power on the electric side and the net regulation power on the thermal side, establish the electric side response rate constraint and the thermal side response rate constraint, as shown in formula (9) and formula (10), and satisfy: As shown in formula (11): (9); (10); (11); In the above formula: and These represent the maximum response rates on the electrical and thermal sides, respectively.

[0017] Through the above constraints, the fast response resources on the electrical side are given priority to undertake the task of fast power correction, while the slow regulation resources on the thermal side are mainly responsible for the task of slow smooth regulation.

[0018] In S4, the day-ahead phase aims to minimize the daily comprehensive operating cost, which includes the cost of purchasing and selling electricity, the operating loss cost of electric energy storage, the gas consumption cost of gas boilers, the demand response compensation cost, and the cost of wind and solar curtailment penalties, as shown in formula (12): (12); In equation (12): This represents the total cost of dispatching as of today. This is the set of scheduling periods for the day before. To cover the cost of purchasing and selling electricity, For the operating losses and costs of energy storage, For the gas consumption cost of gas-fired boilers, To compensate for the cost of demand response, The cost of penalties for abandoning wind and solar power.

[0019] In S4, the intraday phase aims to minimize the weighted sum of the demand response deviation, day-ahead plan tracking, scheduling smoothing, and thermal comfort penalty within the current rolling optimization window, as shown in formula (13): (13); In equation (13): For the first The objective function for a scrolling window, , , , These are the weighting coefficients. For the first A set of time periods within a scrolling window; where the demand response deviation item... Describe the actual net load power of the community Demand response target power The deviation between them is shown in formula (14): (14); The project is currently planned to be tracked. Limit intraday tie-line power Compared to the planned power of the connecting line The deviation is shown in formula (15): (15); Scheduling smoothing term The variation range of net regulation power on the electric side and net regulation power on the thermal side in adjacent time periods is limited, as shown in formula (16): (16); In equation (16): This refers to the total adjustable power of the electrical side resources. This refers to the total adjustable power of the hot-side resources. Thermal comfort penalty A penalty is imposed when the indoor temperature exceeds the limit, as shown in formula (17): (17); Furthermore, each sub-objective item is normalized using the benchmark maximum value before weighting.

[0020] In equation (17): Indoor temperature; and These are the lower and upper limits of indoor comfort temperature, respectively. When the temperature is within the comfortable temperature range, the thermal comfort penalty is 0.

[0021] S5 includes: S5.1: In the During the daily rolling optimization period, the ultra-short-term forecast data within the current rolling optimization window is updated. The ultra-short-term forecast data includes the predicted values ​​of photovoltaic, wind power, electrical load, thermal load, and outdoor temperature. At the same time, the current real-time operating status of the system is obtained, including the state of charge (SOC) of the electrical energy storage, the state of the thermal energy storage, the indoor temperature, and the tie-line power. S5.2: Based on the updated ultra-short-term forecast data and real-time operating status, solve for the... A daily rolling optimization model is used to obtain the sequence of scheduling instructions within the current rolling optimization window; S5.3: Only issue and execute scheduling instructions for the current 15-minute period. The scheduling instructions include electric energy storage charging and discharging instructions, flexible electric load adjustment instructions, heat pump output instructions, gas boiler output instructions, thermal energy storage heat storage and release instructions, adjustable heat load adjustment instructions, and tie line power interaction instructions. S5.4: Collect the actual operating results after the execution of the command in the current time period, update the state of charge (SOC) of electrical energy storage, state of thermal energy storage, indoor temperature, actual net load and tie-line power, and use the updated system state as the initial state for the next rolling optimization cycle.

[0022] In S6, the demand response tracking error As shown in formula (18): (18); In equation (18): This represents the number of scheduling periods within the statistical period. This represents the actual net load power of the community. To meet the target power of demand response, The smaller the value, the better the tracking effect of the demand response instructions; The current plan is to track errors. As shown in formula (19): (19); In equation (19): and These are the intraday tie line power and the day-ahead planned tie line power, respectively. The smaller the value, the less the intraday optimization affects the day-ahead plan; Connection line power fluctuations As shown in formula (20): (20); In equation (20): The smaller the value, the smoother the change in the power of the connection line, and the better the community's user-friendliness. Renewable energy consumption rate As shown in formula (21): (twenty one); In equation (21): and These represent the actual electricity consumption of photovoltaic and wind power, respectively. and These are the available power generation capacities from photovoltaic and wind power, respectively. The larger the value, the stronger the capacity to absorb renewable energy. Thermal comfort deviation As shown in formula (22): (twenty two); In equation (22): Indoor temperature, and The upper and lower limits are set for the user's comfortable temperature.

[0023] This invention discloses a rolling optimization scheduling method for demand response in smart communities that takes into account the difference between electrical and thermal responses. The technical effects are as follows: 1) Improve the tracking accuracy of demand response commands and reduce execution deviations caused by source load prediction errors: This invention constructs a day-ahead-intraday rolling optimization scheduling model, which updates ultra-short-term forecast data in 15-minute intervals during the intraday phase, dynamically correcting the day-ahead basic scheduling plan in a closed-loop manner. Simultaneously, a demand response deviation term is introduced into the intraday rolling optimization objective function. Based directly on the actual net load power of the community Demand response target power The square of the deviation between the parameters is used as one of the optimization objectives. Through the above-mentioned rolling correction and deviation penalty mechanism, the transmission impact of photovoltaic, wind power output and load forecasting errors on dispatching decisions is effectively weakened, and the tracking accuracy of demand response commands is significantly improved.

[0024] 2) Avoid frequently using slow-speed resources on the hot side for short-timescale power point tracking to ensure user thermal comfort: This invention divides adjustable resources into fast-response electrical-side resources and slow-response thermal-side resources according to their response speed, and establishes electrical-side response rate constraints for each. and thermal response rate constraints And make Meanwhile, a scheduling smoothing term is introduced into the intraday rolling optimization objective function. and thermal comfort penalty The former suppresses drastic fluctuations in net regulation power between the electrical and thermal sides in adjacent time periods, while the latter controls indoor temperature. Beyond the comfort zone Penalties are imposed in a timely manner. As a result, fast resources such as electric energy storage and flexible electrical loads undertake the task of correcting short-term power deviations, while slow resources such as heat pumps, thermal energy storage, and building thermal inertia are mainly used for heat load balancing and temperature smoothing, avoiding frequent and large-scale use of heat-side resources, thereby ensuring that the indoor temperature is always within the comfortable range, taking into account both the demand response performance and user thermal comfort.

[0025] 3) Suppress power fluctuations in the tie line and improve community friendliness: This invention generates tie-line power plans with the objective of minimizing overall operating costs during the day-ahead basic scheduling phase, and introduces a day-ahead plan tracking term into the intraday rolling optimization objective function. By penalizing intraday communication line power Power of the planned connecting line The deviation between the two limits excessive deviation of intraday dispatch from the day-ahead plan. Simultaneously, the constraint of the difference in electrical and thermal response rates ensures that rapid power correction tasks are prioritized by electrical resources, reducing the need for frequent power fluctuation mitigation via tie lines due to thermal response lag. The synergistic effect of these mechanisms makes power changes on the tie lines between the community and the upper-level grid smoother, effectively suppressing power fluctuations and improving the grid-friendliness of the smart community.

[0026] 4) Enhance the capacity for renewable energy absorption and promote the local utilization of distributed energy resources: This invention incorporates the cost of wind and solar power curtailment penalties into the day-ahead scheduling objective function. Economic penalties are imposed on unabsorbed wind and solar power output to encourage dispatch schemes to maximize the utilization of photovoltaic and wind power. During the intraday rolling optimization phase, power generation plans are further revised based on updated ultra-short-term forecast data, utilizing various adjustable resources such as energy storage charging and discharging, flexible load regulation, and heat pump output adjustment to achieve dynamic supply and demand matching. Through the combination of day-ahead planning guidance and intraday real-time correction, the absorption rate of wind and solar power output within the community is effectively improved, reducing backfeeding to the upper-level grid or wind and solar curtailment.

[0027] 5) Enhance the system's adaptability to intraday random fluctuations through closed-loop rolling optimization and state feedback: This invention employs a closed-loop scheduling framework of "predictive update—rolling optimization—instruction execution—state feedback." Each intraday rolling optimization moment is based on the latest real-time system status data, including the state of charge of electrical storage, the state of thermal storage, indoor temperature, and tie-line power, as well as ultra-short-term forecast information, to make optimization decisions. Only the scheduling instructions for the current 15-minute period are executed, and the actual operating state after execution is fed back to the next rolling optimization cycle as the initial state. This closed-loop feedback mechanism ensures that each optimization decision reflects the current true state of the system, effectively avoiding control deviations caused by error accumulation in open-loop scheduling, and significantly enhancing the system's adaptability to intraday random fluctuations in photovoltaic and wind power output and load. Attached Figure Description

[0028] The present invention will be further described below with reference to the accompanying drawings and examples; Figure 1 A schematic diagram illustrating the supply and demand principle of electric and thermal energy in a smart community.

[0029] Figure 2 This is a two-stage rolling optimization scheduling framework diagram for smart communities.

[0030] Figure 3 This is a comparison chart of demand response targets and actual net load under different scenarios.

[0031] Figure 4 This is a comparison chart of demand response tracking errors under different schemes.

[0032] Figure 5 The graph shows the power variation curves of the tie line under different schemes.

[0033] Figure 6 This is a comparison chart of the power adjustment on the electrical and thermal sides.

[0034] Figure 7 This is a diagram showing the state of energy storage.

[0035] Figure 8 This chart compares the renewable energy absorption rates under different schemes.

[0036] Figure 9(a) shows the optimized operation results of the electric side of the method of the present invention; Figure 9(b) shows the results of the hot-side optimization operation of the method of the present invention.

[0037] Figure 10 This is a graph showing the changes in indoor temperature under different scenarios.

[0038] Figure 11 This is a graph showing the sensitivity analysis of the thermal side response rate. Detailed Implementation

[0039] 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.

[0040] This invention provides a rolling optimization scheduling method for demand response in smart communities that takes into account the differences between electricity and heat response. This method achieves optimized demand response scheduling in smart communities through steps such as smart community operation data acquisition and processing, integrated electricity-heat energy system modeling, electricity-heat response resource allocation and response rate constraint establishment, day-ahead-intraday rolling optimization scheduling model construction, scheduling command execution and system status feedback updates, and scheduling result analysis based on the rolling optimization model. This method can coordinate fast-response resources on the electricity side with slow-regulation resources on the heat side, avoiding frequent adjustments to heat-side resources within a short timescale, improving the accuracy of demand response command tracking, reducing tie-line power fluctuations, enhancing renewable energy absorption capacity, and ensuring user thermal comfort.

[0041] See Figure 1 and Figure 2 The method of the present invention specifically includes the following steps: Step 1: Smart Community Operation Data Collection and Processing: The data on the operation of the smart community are collected and preprocessed to obtain the input datasets for the day-ahead scheduling model and the intraday rolling optimization model.

[0042] The collected data includes day-ahead forecasts, intraday ultra-short-term forecasts, equipment operating parameters, electricity price information, demand response target power, and real-time system status data. Specifically, day-ahead forecasts include predicted photovoltaic (PV) output, wind power output, electrical load, thermal load, and outdoor temperature; intraday ultra-short-term forecasts include updated PV output, wind power output, electrical load, and thermal load forecasts within the current rolling optimization window; equipment operating parameters include limits on electrical energy storage capacity and power, limits on thermal energy storage capacity and power, rated power of heat pumps, coefficient of performance (COP) of heat pumps, rated thermal power of gas-fired boilers, adjustable capacity of flexible electrical load, and adjustable thermal load capacity; and real-time system status data includes state of charge (SOC) of electrical energy storage, state of thermal energy storage, indoor temperature, actual net load, and tie-line power.

[0043] The collected data underwent time-scale unification, outlier correction, missing data completion, and dimensional unification. Specifically, during the day-ahead scheduling phase, the data was processed to a 1-hour time scale; during the intraday rolling optimization phase, the data was processed to a 15-minute time scale, with a 4-hour rolling optimization window length. The demand response deviation term, day-ahead plan tracking term, scheduling smoothing term, and thermal comfort penalty term in the intraday rolling optimization objective function were normalized to obtain the input quantities for the weighted summation objective function.

[0044] In one embodiment of the present invention, the main equipment scale and operating parameters of the smart community are shown in Table 1, and the time-of-use electricity price and demand response incentive price are shown in Table 2.

[0045]

[0046]

[0047] Tables 1 and 2 can be used to determine the capacity of smart community equipment, the range of energy storage SOC, the range of indoor comfortable temperature, the maximum response rate on the electricity side, the maximum response rate on the heat side, as well as the purchase and sale price of electricity and the demand response incentive price under different electricity consumption periods, providing a parameter basis for day-ahead optimization and intraday rolling optimization.

[0048] Step 2: Construction of the Smart Community Electricity-Heat Integrated Energy System Model: Step 2.1: Establish a structural model of the intelligent community's integrated electricity-heat energy system. For example... Figure 1 As shown, the model includes energy supply units, conversion and storage units, load units, and external interaction units. Energy supply units include photovoltaic (PV), wind power, upstream grid power, and natural gas; conversion and storage units include electric energy storage, heat pumps, gas-fired boilers, and thermal energy storage; load units include rigid electrical loads, flexible electrical loads, rigid thermal loads, and adjustable thermal loads; external interaction units include the tie lines between the community and the upstream grid. Using PV output, wind power output, electric energy storage charging and discharging power, heat pump power consumption, heat pump heating power, gas-fired boiler heating power, thermal energy storage heat storage and release power, flexible electrical load adjustment, adjustable thermal load adjustment, and tie line power as scheduling variables, electrical power balance constraints, thermal power balance constraints, equipment operation constraints, and tie line power constraints are established.

[0049] Step 2.2: Establishing a building thermal inertia model: To reflect the lag and slow variation characteristics of the heat transfer process, this invention uses a first-order thermal inertia model to describe the relationship between indoor temperature changes in adjacent scheduling periods. The building thermal inertia model is used to constrain the dynamic relationship between indoor temperature, outdoor temperature, and effective heating power entering the building, as shown in formula (1): (1); In the formula, For time period The indoor temperature; Outdoor temperature; The effective heating capacity entering the building interior; , , Here are the building thermal dynamic parameters, where: Indicates the ability to maintain indoor temperature inertia. This indicates the effect of heating power on regulating indoor temperature. This indicates the impact of outdoor temperature disturbances on indoor temperature.

[0050] Step 2.3: Establish indoor thermal comfort constraints: To ensure user thermal comfort, the indoor temperature should be maintained within the preset comfort temperature range, as shown in formula (2): (2); In the formula, and These are the lower and upper limits of the user's comfortable temperature, respectively.

[0051] Step 2.4: Establish power balance constraints on the electric side and power balance constraints on the thermal side, and apply upper and lower limit constraints to the operating parameters of each device; The power balance relationship on the community's electrical side is shown in formula (3): (3); In the formula, The power exchange between the community and the higher-level power grid; and The outputs are solar power and wind power, respectively. and These are the energy storage discharge and charging power, respectively. Total community load; This refers to the power consumption of the heat pump.

[0052] The power balance relationship on the community heat side is shown in formulas (4) and (5): (4); (5); In the above formula: The heat pump outputs heat power; This is to output heat power for the gas-fired boiler; and These are thermal energy storage heat release and thermal storage power, respectively. For the community's heat load; The coefficient of performance (COP) of a heat pump; To ensure the safe operation of various equipment, the system also needs to meet the conventional operating constraints of renewable energy output, electric energy storage, thermal energy storage, flexible loads, heat pumps, and gas boilers, which can be uniformly expressed as the equipment operation feasible domain form shown in formula (6): (6); In the formula, For a given time period, the set of system decision variables. The feasible operating domain for the equipment includes the following constraints: photovoltaic and wind power output does not exceed the predicted available power; the charging and discharging power of electric energy storage meets the upper and lower limit constraints; the state of charge (SOC) of electric energy storage meets the constraints of dynamic updates, upper and lower capacity limits, and charging and discharging mutual exclusion; the heat storage power of thermal energy storage meets the upper and lower limit constraints; the state of thermal energy storage meets the constraints of dynamic updates and capacity boundaries; flexible electrical loads meet the constraints of adjustable power limits and the conservation of transferable load power; the heat load that can be reduced meets the constraint of maximum reduction capacity; the output power of heat pumps and gas boilers meets the constraints of the rated capacity of the equipment; at the same time, all types of adjustable resources should also meet the corresponding ramp rate constraints to avoid excessively rapid power changes of a single device in adjacent time periods.

[0053] Step 3: Allocation of electro-thermal response resources and establishment of response rate constraints: Step 3.1: Based on the response speed and operating characteristics of various adjustable resources, the adjustable resources in the smart community are divided into fast-response electrical resources and slow-response thermal resources. Electric energy storage and flexible electrical loads have fast response speeds and high execution accuracy, and are configured as fast-response electrical resources. They are used to correct short-term power deviations and track demand response target power during intraday rolling optimization. Heat pumps, thermal energy storage, adjustable heat loads, and building thermal inertia are affected by the heat transfer process and have relatively slow responses. They are configured as slow-response thermal resources, used for heat load balancing and slow, smooth adjustment under the conditions of meeting thermal power balance constraints and indoor thermal comfort constraints.

[0054] Step 3.2: Define the net regulation power on the electrical side. The net regulation power on the electrical side is determined by the discharge power of the electrical storage, the charging power of the electrical storage, and the regulation power of the flexible electrical load, as shown in formula (7): (7); In the formula, This refers to the net regulating power on the electric side. It is a power regulation function for flexible electrical loads, including a combined regulation of transferable and load reduction.

[0055] Step 3.3: Define the net heat-side regulation power. The net heat-side regulation power is jointly determined by the heat pump output heat power regulation, the heat release power of the thermal energy storage, the heat storage power of the thermal energy storage, and the adjustable heat load regulation, as shown in formula (8): (8); In the formula, Net regulating power on the heat side; This refers to the adjustment amount of the heat pump's output heat power. This allows for adjustable heat load reduction or power transfer.

[0056] Step 3.4: Establish electrical and thermal response rate constraints. To avoid excessive use of slow thermal resources in a short timescale, set the maximum response rate for the electrical side and the maximum response rate for the thermal side, as shown in formulas (9) and (10), and satisfy the relationship between them as shown in formula (11): (9); (10); (11); In the formula, and These represent the maximum response rate on the electrical side and the maximum response rate on the thermal side, respectively. Through these constraints, the fast-response resources on the electrical side are prioritized for rapid power correction tasks, while the slow-response resources on the thermal side primarily undertake slow, smooth regulation tasks, thus achieving coordinated complementarity between electrical and thermal resources on the time scale.

[0057] Step 4: Construction of the day-to-day rolling optimization scheduling model: Step 4.1: Constructing the Day-ahead Basic Scheduling Model: The day-ahead stage uses a 1-hour scheduling step size. Based on the day-ahead forecast data, the equipment output plan, energy storage charging and discharging plan, flexible load adjustment plan, heat pump operation plan, gas boiler output plan, thermal energy storage heat release plan, and tie-line power plan are determined within the scheduling cycle. The day-ahead basic scheduling model aims to minimize the overall operating cost, which includes the cost of purchasing and selling electricity, the operating loss cost of electric energy storage, the gas consumption cost of gas boilers, the demand response compensation cost, and the cost of wind and solar curtailment penalties, as shown in formula (12): (12); In the formula, This represents the total cost of day-ahead dispatching; This is the set of scheduling periods for the day before the current day; Cost of purchasing and selling electricity; Cost of operating losses for energy storage; The cost of gas consumed by the gas boiler; To compensate for costs in response to demand; The cost of penalties for abandoning wind and solar power.

[0058] Step 4.2: Constructing an Intraday Rolling Optimization Model: During the intraday phase, a rolling step of 15 minutes and an optimization window of 4 hours are used. Based on intraday ultra-short-term forecast data and real-time system status feedback, the daily basic scheduling plan is revised. The intelligent community's two-stage rolling optimization scheduling framework (daily-intraday phase) is as follows: Figure 2 As shown in the figure. The intraday rolling optimization model takes the current rolling optimization window as the optimization range, and uses the weighted sum of the demand response deviation term, the day-ahead plan tracking term, the scheduling smoothing term, and the thermal comfort penalty term as the objective function, as shown in formula (13): (13); In the formula, For the first The objective function within a scrolling window; , , , These are the weighting coefficients; This is the demand response deviation item; This is a tracked item for the current period; For scheduling smoothing terms; This represents a penalty term for thermal comfort. In the actual solution process, each sub-objective term is normalized using its baseline maximum value to eliminate the influence of different dimensions on the weighted optimization results.

[0059] Step 4.3: Establish a demand response deviation term to describe the deviation between the actual net load power of the community and the target demand response power, as shown in formula (14): (14); In the formula, This represents the actual net load power of the community. This is the target power for demand response. This parameter ensures that the actual net load power closely tracks the demand response command by minimizing the squared deviation.

[0060] Step 4.4: Establish a day-ahead plan tracking item to limit excessive deviation of intraday tie-line power from the day-ahead plan, as shown in formula (15): (15); In the formula, The power of the connecting line obtained through intraday optimization; This refers to the day-ahead tie-line power. This penalty helps prevent excessive adjustments to the day-ahead tie-line plan during intraday dispatch, thus contributing to smoother interaction with the upstream power grid.

[0061] Step 4.5: Establish a scheduling smoothing term to limit the variation range of net regulation power on the electric side and net regulation power on the thermal side between adjacent scheduling periods, as shown in formula (16): (16); In the formula, and These are the net regulation power on the electrical side and the net regulation power on the thermal side, respectively. This smoothing term effectively suppresses frequent fluctuations in adjustable resources, and is especially helpful in protecting slow-speed equipment on the thermal side.

[0062] Step 4.6: Establish a thermal comfort penalty term to generate a penalty value when the indoor temperature is below the lower limit of the comfort temperature or above the upper limit of the comfort temperature, as shown in formula (17): (17); In the formula, Indoor temperature; and These are the lower and upper limits of indoor comfort temperature, respectively. When When the temperature is within the comfortable temperature range, the thermal comfort penalty is 0; once the limit is exceeded, the penalty value increases with the square of the deviation.

[0063] Step 5: Execution of scheduling instructions and update of system status feedback: Step 5.1: Within the day At each rolling moment, the ultra-short-term forecast information within the current optimization window is first updated, including data such as photovoltaic, wind power, electrical load, thermal load, and outdoor temperature; at the same time, the current operating status of the system is read, such as the SOC of electrical storage, the status of thermal storage, indoor temperature, and tie-line power, to provide input for this round of rolling optimization.

[0064] Step 5.2: Based on the updated prediction data and current state information, solve for the... A rolling optimization model is used to obtain the scheduling arrangements for each device and load within the optimization window.

[0065] Step 5.3: During the scheduling instruction execution phase, only scheduling instructions for the current 15-minute time period are issued and executed. The scheduling results for the remaining time periods within the current rolling optimization window are used as reference values ​​for the next rolling cycle. The scheduling instructions for the current time period include electric energy storage charging and discharging instructions, flexible electric load adjustment instructions, heat pump output instructions, gas boiler output instructions, thermal energy storage heat release instructions, adjustable heat load adjustment instructions, and tie-line power interaction instructions.

[0066] Step 5.4: After the current time period command is completed, collect the actual operating results and update the electrical energy storage SOC, thermal energy storage status, indoor temperature, actual net load and tie line power.

[0067] Step 5.5: Determine whether the current time has reached the end of the scheduling cycle. If the end of the scheduling cycle has not been reached, use the system state data updated in Step 5.4 as the initial state for the next rolling optimization cycle and return to Step 5.1; if the end of the scheduling cycle has been reached, end the current round of scheduling.

[0068] Through the closed-loop mechanism of "prediction update - state estimation - rolling optimization - instruction execution - state feedback" mentioned above, each optimization decision reflects the current real operating state of the system, effectively avoiding control deviations caused by error accumulation in open-loop scheduling, and significantly enhancing the system's ability to adapt to real-time fluctuations in photovoltaic and wind power output and load.

[0069] Step 6: Analysis of Intelligent Community Demand Response Scheduling Results Based on Rolling Optimization Model: Step 6.1: Based on the day-ahead-intraday rolling optimization scheduling model that takes into account the differences in electricity and heat response constructed in the above steps, the demand response scheduling results of the smart community are analyzed. A typical smart community integrated electricity and heat energy system is selected as the analysis object. The system includes photovoltaic, wind power, electric energy storage, heat pump, gas boiler, thermal energy storage, flexible electrical load and adjustable heat load, and interacts with the upper-level power grid through tie lines.

[0070] Step 6.2: To compare the impact of different constraints on scheduling performance, multiple comparison schemes are designed. The baseline scheme (Scheme 1) only performs day-ahead optimization; Scheme 2 introduces conventional day-ahead-intraday rolling optimization; Scheme 3, based on Scheme 2, distinguishes between electrical and thermal response rate differences; Scheme 4, based on Scheme 2, adds tie-line power smoothing constraints; Scheme 5 incorporates both types of constraints, i.e., the method proposed in this invention. The results of each scheme are compared under scenarios where source-load prediction has biases.

[0071]

[0072] Step 6.3: The dispatching effectiveness is evaluated from six dimensions: overall operating cost, demand response tracking, day-ahead plan deviation, tie-line power fluctuation, renewable energy consumption, and thermal comfort. The definitions of each evaluation indicator are as follows: Demand response tracking error reflects the degree of deviation between the actual net load and the demand response target, as shown in formula (18): (18); In the formula, This represents the number of scheduling periods within the statistical period. The smaller the value, the better the demand response instruction tracking effect.

[0073] The tracking error of the day-ahead plan describes the magnitude of the change in intraday adjustment relative to the day-ahead plan, as shown in formula (19): (19); In the formula, The smaller the value, the less the intraday optimization will affect the day-ahead plan.

[0074] The tie-line power fluctuation index reflects the stability of the power exchange between the community and the upper-level power grid, as shown in formula (20): (20); In the formula, For time period The power of the communication line interaction; The smaller the value, the smoother the change in power of the connection line, and the better the community's user-friendliness.

[0075] The renewable energy integration rate measures the utilization level of wind and solar power within a community, as shown in formula (21): (twenty one); In the formula, and These represent the actual power absorbed by photovoltaic and wind power, respectively. and The predicted available power for photovoltaic and wind power, respectively; The larger the value, the stronger the capacity to absorb renewable energy.

[0076] Thermal comfort deviation is represented by the cumulative value of indoor temperature exceedance penalty, as shown in formula (22): (twenty two); In the formula, The smaller the value, the closer the indoor temperature is to the comfortable range. When the indoor temperature meets the comfort requirements throughout the entire scheduling cycle.

[0077] The various dispatch schemes were evaluated using indicators such as comprehensive operating cost, demand response tracking error, day-ahead planning tracking error, tie-line power fluctuation, renewable energy absorption rate, heat-side regulation power fluctuation, and thermal comfort deviation. The simulation results are shown in Table 4.

[0078]

[0079] Step 6.4: Analyze the scheduling effect of the method of the present invention based on the comparison results of each scheme.

[0080] Figure 3 and Figure 4 The tracking results of demand response targets and actual net load for each scheme are presented. Scheme 1 showed significant deviations during periods of severe fluctuations in photovoltaic output and peak-valley load transitions. This is because the daily plan is a static open-loop scheduling mechanism, lacking the ability to quickly correct for intraday random fluctuations. Scheme 2 improved tracking performance by introducing rolling optimization, but because it did not differentiate between the response rates of electrical and thermal resources, some short-term power deviations were borne by thermal resources, and the inertia of the thermal system caused a misalignment between scheduling commands and actual output. In contrast, the method proposed in this invention (Scheme 5) had the lowest tracking error, mainly due to the frequency-division coordination of heterogeneous resources: electric energy storage and flexible electrical loads have fast response speeds and are used for rapid deviation correction; heat pumps, thermal energy storage, and building thermal inertia have slow but persistent responses and are responsible for smoothing slowly varying components. The execution accuracy of demand response commands is thus significantly improved.

[0081] Figure 5The power variation curves of each tie-line scheme are presented. Conventional rolling optimization (Scheme 2) exhibits significant tie-line power fluctuations in certain periods because, without considering electro-thermal allometric responses, it tends to quickly smooth out short-term deviations via tie-lines. Scheme 4 and the method of this invention both introduce tie-line smoothing constraints, significantly reducing power fluctuations. The method of this invention further incorporates electro-thermal response rate constraints, resulting in the smoothest tie-line power changes and effectively suppressing power disturbances from the community to the upper-level power grid.

[0082] Figure 6 The regulation power on the electrical and thermal sides was compared under Scheme 2, Scheme 4, and the method of this invention. Scheme 2 and Scheme 4 did not distinguish the difference in electrical-thermal response rates, resulting in drastic changes in thermal-side regulation power during certain periods. Heat pumps, thermal storage, and adjustable heat loads were actually used for short-term power tracking. The heat transfer process has a lag, making it difficult to fully execute such high-frequency commands at the physical level, leading to indoor temperature fluctuations and execution deviations. This invention, by setting response rate constraints on both the electrical and thermal sides, allows electrical resources to handle rapid deviation correction, while the thermal side participates in heat load smoothing and comfort maintenance at a slower rate. This significantly reduces thermal-side regulation power fluctuations, validating the hierarchical scheduling concept of rapid electrical response and slow thermal adjustment.

[0083] Figure 7 and Figure 8 The paper demonstrates the energy storage status and renewable energy consumption. In the method of this invention, the changes in the State of Charge (SOC) of both electrical energy storage and thermal energy storage are relatively smooth: electrical energy storage handles short-term power correction, while thermal energy storage focuses on peak shaving and valley filling of heat load and slow smoothing. Compared with Scheme 1, the renewable energy consumption rate increases from 94.91% to 98.32%, and the intraday rolling correction and electricity-thermal coordinated regulation significantly improve the local consumption capacity of distributed energy.

[0084] Figures 9(a) and 9(b) illustrate the optimized operation results of the power and heat sides of the smart community under the method of this invention. In Figure 9(a), positive values ​​represent power supply or energy storage discharge, while negative values ​​represent electrical load, heat pump power consumption, and energy storage charging. During off-peak hours at night, the system purchases electricity through the tie line to meet the basic load, while simultaneously arranging energy storage charging to reserve adjustment capacity for subsequent peak periods. During periods of high photovoltaic output, distributed energy prioritizes supplying electricity to the community, and excess electricity is consumed locally through energy storage charging and heat pump operation. During midday and evening peak load periods, energy storage switches to discharge, sharing the electrical load with wind and solar power, and the power purchased through the tie line decreases accordingly. In Figure 9(b), heat pumps, gas boilers, and thermal energy storage work together to meet the community's heat load. When electricity prices are low or renewable energy is abundant, heat pumps undertake the main heating; when the heat load is high or heat pumps are insufficient, gas boilers provide compensation; thermal energy storage smooths out power fluctuations on the heat side through heat storage and release. Compared to the electrical side, the power changes of the heating side equipment are more gradual, reflecting the slow response characteristics of thermal energy storage and building thermal inertia. Overall, electrical energy storage and flexible electrical loads are responsible for rapid power correction, while heat pumps, thermal energy storage, and building thermal inertia are responsible for slow heat load balancing, thus achieving coordinated operation of rapid response on the electrical side and slow adjustment on the heating side.

[0085] Figure 10 Indoor temperatures were compared across different schemes. If thermal resources are frequently and significantly utilized within a short timescale, indoor temperatures inevitably fluctuate, leading to a decrease in user thermal comfort. The method of this invention incorporates building thermal inertia and thermal comfort constraints into the rolling optimization process, while simultaneously limiting the thermal adjustment rate, thus making the thermal adjustment process smoother. Figure 10 It can be seen that the indoor temperature is basically maintained within the set comfort range, achieving a balance between demand response performance and user-side thermal comfort.

[0086] To examine the impact of the maximum response rate on the hot side on the scheduling effect, three scenarios of 30, 50, and 70 kW / (15 min) were set up for comparison. The results are shown in Table 5. Figure 11 When the maximum response rate on the thermal side is too low, the space for thermal side resources to participate in intraday correction is limited; conversely, if the rate is set too high, thermal side resources are easily over-utilized, and the deviation in thermal comfort increases accordingly. In this embodiment, the setting of 50kW / (15min) achieves a relative balance between demand response accuracy, connection line smoothness, and thermal comfort.

[0087]

[0088] In summary, this invention primarily utilizes fast-response resources such as electric energy storage and flexible electrical loads for power deviation correction on a 15-minute timescale, while allocating more resources such as heat pumps, thermal energy storage, and building thermal inertia to heat load balancing and indoor temperature maintenance. This leverages the rapid response of electric resources while avoiding the frequent use of thermal resources for short-term power point tracking. Comparative results demonstrate that this method significantly reduces overall operating costs and demand response tracking errors, smooths tie-line power, improves local wind and solar power integration, and ensures indoor thermal comfort. The method has a clear process, effectively coordinating fast-response electric resources with slow-regulation thermal resources, reducing the impact of source-load prediction errors on scheduling results, improving demand response command tracking accuracy, suppressing tie-line power fluctuations, enhancing renewable energy integration capacity, and simultaneously ensuring user thermal comfort.

Claims

1. A rolling optimization scheduling method for demand response in smart communities that takes into account the difference between electrical and thermal responses, characterized in that... Includes the following steps: S1: Collect and preprocess the operation data of the smart community's integrated electric-thermal energy system to obtain input data for day-ahead scheduling and intraday rolling optimization; S2: Based on the equipment configuration and load characteristics in the input data, construct an integrated electric-thermal energy system model. This model includes electric power balance constraints, thermal power balance constraints, equipment operation constraints, and indoor thermal comfort constraints. S3: Based on the response speed of various adjustable resources, adjustable resources are divided into fast-response resources on the electrical side and slow-response resources on the thermal side, and electrical-side response rate constraints and thermal-side response rate constraints are established to limit the frequent use of slow-response resources on the thermal side within a short time scale. S4: Based on the integrated electric-thermal energy system model and the response rate constraints of the electric and thermal sides, a day-ahead and intraday rolling optimization scheduling model is constructed; in the day-ahead stage, the objective is to minimize the daily comprehensive operating cost, and a day-ahead basic scheduling plan is generated. The intraday phase uses a fixed-cycle rolling approach. Based on intraday ultra-short-term forecast data and real-time system status feedback, the intraday basic scheduling plan is dynamically modified with the goal of minimizing the weighted sum of demand response deviation, day-ahead plan tracking deviation, scheduling smoothness and thermal comfort deviation, resulting in an intraday rolling optimized scheduling scheme. S5: During each daily rolling optimization period, only the scheduling instructions for the current time period are executed. After the scheduling instructions for the current time period are completed, the actual operating status of the system is collected and fed back to the next rolling optimization cycle to form a closed-loop scheduling.

2. The intelligent community demand response rolling optimization scheduling method considering the difference in electrical and thermal response as described in claim 1, characterized in that: It also includes S6: Based on the actual operating data obtained from closed-loop scheduling, it calculates the demand response tracking error, day-ahead planning tracking error, tie-line power fluctuation, renewable energy absorption rate, and thermal comfort deviation index, and analyzes the scheduling effect.

3. The intelligent community demand response rolling optimization scheduling method considering the difference in electrical and thermal response as described in claim 2, characterized in that: S1 includes: S1.1: Collect day-ahead forecast data, intraday ultra-short-term forecast data, equipment operating parameters, electricity price information, demand response target power, and real-time system status data of the smart community's integrated electric-heat energy system; The day-ahead forecast data includes forecasts for photovoltaic power output, wind power output, electrical load, heat load, and outdoor temperature. The intraday ultra-short-term forecast data includes each forecast value updated within the current rolling optimization window; The system's real-time status data includes the state of charge (SOC) of electrical energy storage, the state of thermal energy storage, indoor temperature, and tie-line power. S1.2: Perform time scale unification, abnormal data correction, missing data completion and normalization on the collected data to obtain 1-hour time scale data used in the day-ahead scheduling phase and 15-minute time scale data used in the intraday rolling optimization phase, and use 4 hours as the rolling optimization window length.

4. The intelligent community demand response rolling optimization scheduling method considering the difference in electrical and thermal response as described in claim 3, characterized in that: S2 includes: 2.1: Establish an integrated electric-thermal energy system model comprising energy supply units, conversion and storage units, load units, and external interaction units. The energy supply units include photovoltaic, wind power, upstream power grid, and natural gas; the conversion and storage units include electric energy storage, heat pumps, gas-fired boilers, and thermal energy storage; the load units include rigid electrical loads, flexible electrical loads, rigid thermal loads, and adjustable thermal loads; and the external interaction units are the interconnection lines between the community and the upstream power grid. 2.2: A first-order thermal inertia model is used to describe the relationship between indoor temperature changes in adjacent scheduling periods. The first-order thermal inertia model is shown in formula (1): （1）； In formula (1): For time period The indoor temperature; Outdoor temperature; The effective heating power entering the building interior; a, b, and c are building thermal dynamic parameters, where: a represents the indoor temperature inertia maintenance capacity, b represents the regulating effect of heating power on indoor temperature, and c represents the influence of outdoor temperature disturbance on indoor temperature. 2.3: Establish indoor thermal comfort constraints to keep the indoor temperature within the preset comfortable temperature range, as shown in formula (2): （2）； In formula (2): and These are the lower and upper limits of the user's comfortable temperature, respectively; 2.4: Establish power balance constraints on the electric side and the thermal side, and set capacity constraints, upper and lower power limits, state update constraints, and ramp constraints for electric energy storage, thermal energy storage, heat pumps, gas boilers, flexible electric loads, adjustable thermal loads, and tie-line power, respectively. These constraints can be uniformly expressed as... ,in: For a given time period, the set of system decision variables. This is the feasible operating domain for the device.

5. The intelligent community demand response rolling optimization scheduling method considering the difference in electrical and thermal response as described in claim 4, characterized in that: In section 2.4, the power balance constraint on the electric side is as shown in formula (3): （3）； In formula (3): The power exchange between the community and the higher-level power grid; and The outputs are solar power and wind power, respectively. and These are the energy storage discharge and charging power, respectively. Total community load; This refers to the power consumption of the heat pump; The heat-side power balance constraints are shown in equations (4) and (5): （4）； （5）； In the above formula: The heat pump outputs heat power; This is to output heat power for the gas-fired boiler; and These are thermal energy storage heat release and thermal storage power, respectively. For the community's heat load; This is the coefficient of performance (COP) of the heat pump.

6. The intelligent community demand response rolling optimization scheduling method considering the difference in electrical-thermal response as described in claim 5, characterized in that: S3 includes: S3.1: Electric energy storage and flexible electric loads are classified as fast-response resources on the electric side, used to correct short-timescale power deviations and track demand response target power in intraday rolling optimization; heat pumps, thermal energy storage, adjustable heat loads and building thermal inertia are classified as slow-regulation resources on the thermal side, used to balance heat loads and perform slow-smooth regulation under the conditions of satisfying thermal power balance constraints and indoor thermal comfort constraints. S3.2: Define the net regulation power on the electric side Net regulating power on the heat side As shown in formulas (7) and (8) respectively: （7）； （8）； In the above formula: This refers to the net regulating power on the electric side. The power regulation for flexible electrical loads includes a combined regulation amount for transferable and load reduction. Net regulating power on the heat side; This refers to the adjustment amount of the heat pump's output heat power. To reduce or transfer power for adjustable heat load; S3.3: Based on the net regulation power on the electric side and the net regulation power on the thermal side, establish the electric side response rate constraint and the thermal side response rate constraint, as shown in formula (9) and formula (10), and satisfy: As shown in formula (11): （9）； （10）； （11）； In the above formula: and These are the maximum response rates on the electrical and thermal sides, respectively. Through the above constraints, the fast response resources on the electrical side are given priority to undertake the task of rapid power correction, while the slow regulation resources on the thermal side are given priority to undertake the task of slow and smooth regulation.

7. The intelligent community demand response rolling optimization scheduling method considering the difference in electrical-thermal response according to claim 6, characterized in that: In S4, the day-ahead phase aims to minimize the daily comprehensive operating cost, which includes the cost of purchasing and selling electricity, the operating loss cost of electric energy storage, the gas consumption cost of gas boilers, the demand response compensation cost, and the cost of wind and solar curtailment penalties, as shown in formula (12): （12）； In equation (12): The total cost of dispatching as of today. This is the set of scheduling periods for the day before. To cover the cost of purchasing and selling electricity, For the operating losses and costs of energy storage, For the gas consumption cost of gas-fired boilers, To compensate for the cost of demand response, The cost of penalties for abandoning wind and solar power.

8. The intelligent community demand response rolling optimization scheduling method considering the difference in electrical-thermal response according to claim 7, characterized in that: In S4, the intraday phase aims to minimize the weighted sum of the demand response deviation, day-ahead plan tracking, scheduling smoothing, and thermal comfort penalty within the current rolling optimization window, as shown in formula (13): （13）； In equation (13): For the first The objective function for a scrolling window, , , , These are the weighting coefficients. For the first A set of time periods within a scrolling window; where the demand response deviation item... Describe the actual net load power of the community Demand response target power The deviation between them is shown in formula (14): （14）； The project is currently under tracking. Limit intraday tie-line power Compared to the planned power of the connecting line The deviation is shown in formula (15): （15）； Scheduling smoothing term The variation range of net regulation power on the electric side and net regulation power on the thermal side in adjacent time periods is limited, as shown in formula (16): （16）； In equation (16): This refers to the total adjustable power of the electrical side resources. This refers to the total adjustable power of the hot-side resources. Thermal comfort penalty A penalty is imposed when the indoor temperature exceeds the limit, as shown in formula (17): （17）； Furthermore, each sub-objective item is normalized using the benchmark maximum value before weighting; In equation (17): Indoor temperature; and These are the lower and upper limits of indoor comfort temperature, respectively. When the temperature is within the comfortable temperature range, the thermal comfort penalty is 0.

9. The intelligent community demand response rolling optimization scheduling method considering the difference in electrical and thermal response as described in claim 8, characterized in that: S5 includes: S5.1: In the During the daily rolling optimization period, the ultra-short-term forecast data within the current rolling optimization window is updated. The ultra-short-term forecast data includes the predicted values ​​of photovoltaic, wind power, electrical load, thermal load, and outdoor temperature. At the same time, the current real-time operating status of the system is obtained, including the state of charge (SOC) of the electrical energy storage, the state of the thermal energy storage, the indoor temperature, and the tie-line power. S5.2: Based on the updated ultra-short-term forecast data and real-time operating status, solve for the... A daily rolling optimization model is used to obtain the sequence of scheduling instructions within the current rolling optimization window; S5.3: Only issue and execute scheduling instructions for the current 15-minute period. The scheduling instructions include electric energy storage charging and discharging instructions, flexible electric load adjustment instructions, heat pump output instructions, gas boiler output instructions, thermal energy storage heat storage and release instructions, adjustable heat load adjustment instructions, and tie line power interaction instructions. S5.4: Collect the actual operating results after the execution of the command in the current time period, update the state of charge (SOC) of electrical energy storage, state of thermal energy storage, indoor temperature, actual net load and tie-line power, and use the updated system state as the initial state for the next rolling optimization cycle.

10. The intelligent community demand response rolling optimization scheduling method considering the difference in electrical and thermal response as described in claim 2, characterized in that: In S6, the demand response tracking error As shown in formula (18): （18）； In formula (18): This represents the number of scheduling periods within the statistical period. This represents the actual net load power of the community. To meet the target power of demand response, The smaller the value, the better the tracking effect of the demand response instructions; The current plan is to track errors. As shown in formula (19): （19）； In equation (19): and These are the intraday tie line power and the day-ahead planned tie line power, respectively. The smaller the value, the less the intraday optimization affects the day-ahead plan; Connection line power fluctuations As shown in formula (20): （20）； In equation (20): The smaller the value, the smoother the change in the power of the connection line, and the better the community's user-friendliness. Renewable energy consumption rate As shown in formula (21): （21）； In equation (21): and These represent the actual electricity consumption of photovoltaic and wind power, respectively. and These are the available power generation capacities from photovoltaic and wind power, respectively. The larger the value, the stronger the capacity to absorb renewable energy. Thermal comfort deviation As shown in formula (22): （22）； In equation (22): Indoor temperature, and The upper and lower limits are set for the user's comfortable temperature.

Images