Building flexible load scheduling method and device based on PPD guided two-stage

Abstract

Embodiments of the present application relate to the technical field of building flexible load scheduling, and particularly relate to a building flexible load scheduling method and device based on PPD guided two-stage. The method comprises: determining a control variable candidate domain and a time period set with human occupation; under a preset constraint condition, obtaining an optimal fixed temperature set point and an optimal fixed air supply level of each room with the minimum total energy consumption of a heating ventilation air conditioning system as an optimization target; taking the optimal fixed air supply level as an air volume output value of each room, and obtaining an optimal hour-by-hour temperature set point sequence with the optimal fixed temperature set point as a reference value under the constraint of the preset constraint condition and the optimization target; and taking the optimal hour-by-hour temperature set point sequence as a temperature output value of each room at different preset time points, so as to regulate the temperature and air supply of each room in the region to be optimized, thereby realizing scheduling of the building flexible load.

Description

Technical Field

[0001] This invention relates to the field of building flexible load scheduling technology, and in particular to a building flexible load scheduling method and apparatus based on PPD-guided two-stage method. Background Technology

[0002] Currently, in the context of achieving carbon neutrality, flexible load scheduling in the building sector is crucial, and air conditioning systems typically account for about 50% of building operating energy consumption. Therefore, it is necessary to optimize the operation of air conditioning systems for energy conservation.

[0003] In related technologies, energy conservation and carbon reduction are achieved by starting, stopping, and adjusting air conditioning systems according to a pre-set schedule or fixed temperature setpoint. However, this approach often struggles to adapt to dynamic changes in weather, occupants, and building interior thermal disturbances, resulting in limited energy-saving potential and difficulty in efficient deployment in large-scale, multi-zone buildings.

[0004] Therefore, there is an urgent need for a building flexible load scheduling air conditioning optimization control method and device based on PPD-guided two-stage approach to solve the above-mentioned technical problems. Summary of the Invention

[0005] This invention provides a PPD-guided two-stage air conditioning optimization control method and device, which enables the air conditioning operation scheme to adapt to the dynamic changes of people and thermal disturbances inside the building while achieving the goals of energy saving and carbon reduction, meeting comfort constraints, improving energy-saving potential, and reducing computational complexity.

[0006] In a first aspect, the present invention provides a two-stage building flexible load scheduling method based on PPD guidance, wherein the building flexible load includes a heating, ventilation, and air conditioning system, and the method includes: Determine the candidate domain of control variables for air conditioning terminal equipment in the area to be optimized, as well as the set of time periods during which each room is occupied during working hours; Under preset constraints, with the goal of minimizing the total energy consumption of the HVAC system in the area to be optimized, the temperature setpoint and air supply level of the air conditioning terminal equipment in each room during the entire working day are solved to obtain the optimal fixed temperature setpoint and optimal fixed air supply level for each room. The preset constraints are: the PPD value of all rooms during the occupied period is lower than the preset threshold. Using the optimal fixed air supply level as the air volume output value of each room, under preset constraints, with the goal of minimizing the total energy consumption of the HVAC system in the area to be optimized, and using the optimal fixed temperature setpoint as the reference value, the temperature setpoint corresponding to each room at different preset time points is solved to obtain the optimal hourly temperature setpoint sequence. The optimal hourly temperature setpoint sequence is used as the temperature output value of each room at different preset time points to regulate the temperature and air volume of each room in the area to be optimized, thereby realizing the scheduling of the building's flexible load.

[0007] Secondly, the present invention provides a building flexible load scheduling device based on PPD-guided two-stage method, wherein the building flexible load includes a heating, ventilation, and air conditioning system, and the device includes: Initialize the data acquisition module to determine the candidate domain of control variables for air conditioning terminal equipment in the area to be optimized, as well as the set of time periods during which each room is occupied during working hours; The first-stage optimization module is connected to the initialization data acquisition module. Under preset constraints, with the goal of minimizing the total energy consumption of the HVAC system in the area to be optimized, it solves the temperature setpoint and air supply level of the air conditioning terminal equipment in each room during the whole day's working time, and obtains the optimal fixed temperature setpoint and optimal fixed air supply level for each room. The preset constraint is that the PPD value of all rooms during the occupied period is lower than the preset threshold. The second-stage optimization module is connected to the first-stage optimization module. It takes the optimal fixed air supply level as the air volume output value of each room. Under preset constraints, it takes the minimum total energy consumption of the HVAC system in the area to be optimized as the optimization objective. Using the optimal fixed temperature setpoint as the reference value, it solves the temperature setpoint corresponding to each room at different preset time points to obtain the optimal hourly temperature setpoint sequence. The control command output module is connected to the second-stage optimization module. It uses the optimal hourly temperature setpoint sequence as the temperature output value of each room at different preset time points to regulate the temperature and air volume of each room in the area to be optimized, thereby realizing the scheduling of the building's flexible load.

[0008] Thirdly, the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements the method described in the first aspect of the present invention.

[0009] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to perform the method described in the first aspect of the present invention.

[0010] This invention provides a PPD-guided two-stage method and apparatus for scheduling flexible building loads. Through a two-stage control architecture of coarse-grained optimization followed by fine-grained optimization, it effectively avoids the high-dimensional decision space problem caused by simultaneously optimizing temperature and airflow, significantly reducing computational complexity and improving computational efficiency. It is suitable for large-scale, multi-area buildings. During the optimization process, thermal comfort satisfaction is consistently used as a constraint, enabling the air conditioning operation scheme to adapt to dynamic changes in building occupants and thermal disturbances while achieving energy conservation and carbon reduction goals. This improves energy-saving potential and comfort levels, ultimately achieving the scheduling of flexible building loads. Attached Figure Description

[0011] 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 some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart of a two-stage building flexible load scheduling method based on PPD guidance provided by an embodiment of the present invention; Figure 2 This is a hardware architecture diagram of an electronic device provided in an embodiment of the present invention; Figure 3 This is a structural diagram of a building flexible load scheduling device based on PPD-guided two-stage according to an embodiment of the present invention; Figure 4 This is a schematic diagram of an office building and its floor area division, which is the subject of this study, provided by an embodiment of the present invention; Figure 5 This is a schematic diagram comparing the energy consumption and thermal comfort of three groups of experiments in scenario A, provided by an embodiment of the present invention; Figure 6 This is a schematic diagram comparing the convergence process of three groups of experiments in scenario A at layer 2, provided by an embodiment of the present invention. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0014] Please refer to Figure 1 This invention provides a two-stage building flexible load scheduling method based on PPD guidance, wherein the building flexible load includes a heating, ventilation, and air conditioning system. The method includes: Step 100: Determine the candidate domain of control variables for the air conditioning terminal equipment in the area to be optimized, and the set of time periods during which each room is occupied during working hours; Step 102: Under the preset constraints, with the goal of minimizing the total energy consumption of the HVAC system in the area to be optimized, solve the temperature setpoint and air supply level of the air conditioning terminal equipment in each room during the whole day's working time, and obtain the optimal fixed temperature setpoint and optimal fixed air supply level for each room. The preset constraint is that the PPD value of all rooms is lower than the preset threshold during the period when they are occupied. Step 104: Take the optimal fixed air supply level as the air volume output value of each room. Under the preset constraints, take the minimum total energy consumption of the HVAC system in the area to be optimized as the optimization objective. Take the optimal fixed temperature setpoint as the reference value, solve the temperature setpoint corresponding to each room at different preset time points, and obtain the optimal hourly temperature setpoint sequence. Step 106: Use the optimal hourly temperature setpoint sequence as the temperature output value of each room at different preset time points to regulate the temperature and air volume of each room in the area to be optimized, thereby realizing the scheduling of the building's flexible load.

[0015] In this embodiment of the invention, initial state data and constraints of the heating, ventilation, and air conditioning (HVAC) system within a building (e.g., an office building) in the area to be optimized are obtained. The building's physical information and discrete candidate domains of equipment control variables (including the range of regional temperature setpoints and air supply levels) are acquired. The set of occupancy periods for each room during its working hours is determined, and an upper limit threshold for the predicted percentage of dissatisfaction (PPD) value in thermal comfort assessment is set. Under the dual constraints of thermal comfort assessment and total energy consumption of the HVAC system, the air supply level and temperature setpoint are optimized in two stages to ultimately determine an air conditioning control scheme that satisfies both user thermal comfort and energy consumption. In the first stage of optimization: the total energy consumption of the HVAC system is minimized while ensuring that the maximum PPD value during occupancy periods in all rooms of the area to be optimized is below the threshold. For each independent control zone (typically a single room as an independent control area), the temperature setpoint and air supply level of that zone are set to constants throughout the day, forming a low-dimensional decision space (i.e., the air conditioning temperature and water supply level of that room are constant values ​​throughout the day). An adaptive particle swarm optimization algorithm with PPD-guided directional bias mechanism is used to solve this, outputting the first-stage optimization results: the optimal fixed temperature setpoint and optimal fixed air supply level for each room. In the second-stage optimization process: the optimal fixed air supply level is used as the airflow output value for each room throughout the day, meaning the air supply level is a fixed value that does not change throughout the day. Simultaneously, the temperature at each preset time point throughout the day is adjusted and optimized to further minimize the total energy consumption of the HVAC system. The decision variables are expanded to an hourly temperature setpoint sequence for each zone, forming a high-dimensional decision space. Using the optimal fixed temperature setpoint output from the first stage as the baseline value for the initial population, an adaptive particle swarm optimization algorithm with PPD-guided directional bias mechanism is employed again for solving the problem. The constraints and optimization objective remain unchanged. After iterative solving, the optimization results of the second stage are finally output: the temperature setpoints for each room at different preset time points, i.e., the optimal hourly temperature setpoint sequence. Based on the optimal fixed air supply level output from the first stage and the optimal hourly temperature setpoint sequence for each room output from the second stage, a baseline control strategy is generated and sent to the building automation system for execution. Ultimately, this ensures that the air conditioning terminal equipment in each room outputs the optimal fixed air supply level throughout the day and outputs the optimal hourly temperature setpoint value at different preset time points, minimizing energy consumption while ensuring user thermal comfort, ultimately achieving the scheduling of flexible building loads.

[0016] As can be seen, the technical solution in this embodiment of the invention adopts a two-stage control architecture of "first determining the fixed air supply level and initial temperature throughout the day (low dimension) — then fixing the air supply level and fine-tuning the hourly temperature (high dimension)". This greatly avoids the severe fragmentation of the feasible region caused by the joint optimization of temperature and air volume in large-scale multi-regional buildings, and avoids the problem of premature convergence of local minima during the optimization process. By reducing the dimensionality of the high-dimensional decision space and guiding the model to correct the search direction with the PPD-guided directional bias correction algorithm in this embodiment of the invention, the model can quickly converge to an air conditioning operation scheme that takes into account both high energy efficiency and high comfort, overcoming the shortcomings of existing technologies such as limited energy-saving potential, high computational complexity, and difficulty in effectively handling thermal comfort constraints.

[0017] It should be noted that HVAC systems are a typical example of flexible loads in buildings. Furthermore, due to their thermal inertia and widespread installed capacity, HVAC systems are an important component of demand-side flexible resources. In this invention, by coordinating the optimization of the temperature setpoint and air supply volume, the energy-saving potential of the building side and the effective supply of resources for flexible grid dispatch are realized, supporting the source-load coordination and peak shaving / valley filling goals under the new power system.

[0018] In one embodiment of the present invention, under preset constraints, with the goal of minimizing the total energy consumption of the HVAC system in the area to be optimized, the temperature setpoint and air supply level of the air conditioning terminal equipment in each room are solved throughout the day to obtain the optimal fixed temperature setpoint and optimal fixed air supply level for each room, including: Within the candidate domain of control variables, select candidate values ​​for air supply level and temperature setpoint, and calculate the PPD value of each room during the occupied period for each set of candidate values. For candidate values ​​where the PPD value is higher than the preset threshold, the current temperature setpoint of the room is directionally offset and corrected according to the preset neutral comfort temperature of the corresponding room, so as to obtain a candidate temperature setpoint value that moves closer to the direction of satisfying the constraint conditions. The optimization objective is to minimize the total energy consumption of the HVAC system in the area to be optimized. The process is iterated repeatedly until the optimal fixed temperature setpoint and the optimal fixed air supply level for each room are selected.

[0019] In this embodiment, for example, the candidate set of temperature setpoints within the candidate domain of the control variables is set to [19.0℃, 19.5℃, 20.0℃, ..., 26.0℃] and the candidate set of air supply levels is set to [0, 0.2, 0.4, 0.6, 0.8, 1.0]. Multiple candidate solutions are generated based on the candidate values ​​of the air supply levels and temperature setpoints. Each candidate solution represents the temperature setpoint and fixed air supply level output throughout the day for that room. For each candidate solution, the operation of the air conditioning system under the corresponding operating scheme is simulated at various times throughout a typical weather day, and the PPD value of each room during the occupied period is calculated. If the PPD value of a room exceeds a preset threshold τ (e.g., τ=0.2), a directional bias correction is applied to the temperature setpoint based on the room's neutral comfort temperature (e.g., 23°C): when the temperature setpoint is higher than the neutral comfort temperature, a cooling correction is applied; when it is lower than the neutral comfort temperature, a heating correction is applied. After multiple iterations and optimizations, with the goal of minimizing the total energy consumption of the HVAC system, candidate solutions that satisfy the condition that the PPD of all rooms is lower than τ were selected as the optimal fixed temperature setpoint and the optimal fixed air supply level output.

[0020] In one embodiment of the present invention, the optimal fixed air supply level is used as the air volume output value for each room. Under preset constraints, the optimization objective is to minimize the total energy consumption of the HVAC system in the area to be optimized. Using the optimal fixed temperature setpoint as the reference value, the temperature setpoints corresponding to each room at different preset time points are solved to obtain the optimal hourly temperature setpoint sequence, including: Using the optimal fixed temperature setpoint as the baseline value, generate hourly temperature setpoint candidate values ​​for each room at different preset time points, and calculate the hourly PPD value of each room during the occupied period corresponding to each set of candidate values. For candidate values ​​corresponding to time points where the PPD value is higher than the preset threshold, the temperature setpoint at the time point is directionally offset and corrected according to the preset neutral comfort temperature of the corresponding room, so as to obtain candidate temperature setpoint values ​​that move closer to the direction of satisfying the constraints. The optimization objective is to minimize the total energy consumption of the HVAC system in the area to be optimized. The process is iterated repeatedly until the optimal hourly temperature setpoint for each room at different preset time points is selected.

[0021] In this embodiment, the optimal fixed temperature setpoint (e.g., 22.5°C) output in the first stage is used as the baseline value, and initial temperature setpoint candidate values ​​for each time step are generated according to a preset time step (e.g., one hour). The temperature setpoint candidate values ​​can be generated by perturbation within a small range around the baseline value, such as discrete values ​​within the ±1°C range of the baseline value. For each set of candidate solutions, its operation is simulated throughout a typical weather day, and the hourly PPD value for each room during the occupied period is calculated. If the PPD value at a certain time point exceeds a preset threshold τ (e.g., τ=0.2), a directional bias correction is applied to the temperature setpoint at that time point based on the room's neutral comfort temperature (e.g., 23°C). When the temperature setpoint is higher than the neutral comfort temperature, a cooling correction is applied; when it is lower than the neutral comfort temperature, a heating correction is applied. This bias correction brings the temperature setpoint closer to satisfying comfort constraints. After multiple iterations and optimizations, with the goal of minimizing the total energy consumption of the HVAC system, candidate solutions were selected that ensured that the hourly PPD of all rooms was below τ during the occupied period. These solutions were then output as the optimal hourly temperature setpoint sequence for each time step.

[0022] In one embodiment of the present invention, the optimization objective is expressed by the following formula: The total energy consumption of the HVAC system. The penalty coefficient is... This indicates that when the PPD value during any occupied time period exceeds a preset threshold, the function is activated. At that time, penalty items should be included in the optimization objectives. Indicates the first The room at the preset time point PPD value at the location, Indicates the temperature output value. Candidate domain for control variables representing temperature. This indicates the air volume output value. Candidate domain for control variables representing air volume. This represents the set of time periods that are occupied by someone.

[0023] In this embodiment, a penalty coefficient is added to candidate solutions that do not satisfy the PPD constraint in each iteration, ensuring that the priority of infeasible solutions is always lower than that of feasible solutions, thus reducing their priority during the selection process. Specifically, the magnitude of the penalty coefficient can be positively correlated with the degree to which the PPD exceeds a preset threshold; the greater the exceedance of the preset threshold, the lower the priority. This ensures that the algorithm prioritizes feasible solutions and guides the search closer to the feasible region boundary, while retaining valuable information from infeasible solutions (for example, in some candidate solutions, although the PPD value is greater than the preset threshold, it is only slightly exceeded, and the energy consumption corresponding to the candidate solution is extremely low; directly eliminating the candidate solution may result in the loss of valuable search direction information), accelerating convergence to the globally optimal solution that satisfies the comfort constraint and minimizes energy consumption.

[0024] In one embodiment of the present invention, the directional bias correction is as follows: if the current temperature setpoint is greater than the preset neutral comfort temperature, a cooling direction bias correction is applied to the current temperature setpoint to obtain a candidate temperature setpoint value that moves toward the direction that satisfies the constraint conditions. If the current temperature setpoint is lower than the preset neutral comfort temperature, apply a bias correction to the current temperature setpoint in the direction of temperature rise to obtain a candidate temperature setpoint value that moves closer to the direction that satisfies the constraint conditions.

[0025] In this embodiment, during the bias correction process: when the PPD value of a room is higher than a preset threshold and the current temperature setpoint is higher than a preset neutral comfort temperature, a cooling bias correction is applied to the temperature setpoint in that room to bring it closer to the neutral comfort temperature. If the PPD value of a room is higher than the preset threshold and the current temperature setpoint is lower than the neutral comfort temperature, a heating bias correction is applied to the temperature setpoint in that room to bring it closer to the neutral comfort temperature. Through this directional bias correction, the temperature setpoint is guided to gradually adjust to a value range that satisfies the PPD constraint condition.

[0026] Specifically, both the first and second stage optimizations mentioned above involve optimization problems with comfort constraints. In this embodiment of the invention, an Adaptive Particle Swarm Optimization (APSO) algorithm is used as the solver for the optimization problem, and a PPD guidance mechanism is introduced to improve the efficiency of generating feasible solutions. Particle Swarm Optimization (PSO) searches for the optimal solution in the solution space through group cooperation, without relying on gradient information, making it suitable for simulation-driven energy consumption optimization. The adaptive mechanism dynamically adjusts the inertia weight and learning factor during the iteration process, giving the algorithm stronger global exploration capabilities in the early stages and stronger local exploration capabilities in the later stages, thereby improving convergence speed and solution stability. Specific steps include: initializing multiple particles, where the position vector of each particle contains the temperature setpoint and air supply level of each room. In each iteration, the velocity vector of each particle is first updated based on its individual historical best position and the group historical best position, and then its position vector is updated (the individual historical best position refers to the position vector found by the candidate solution in previous iterations that satisfies the PPD constraint and minimizes the total energy consumption of the HVAC system; the group historical best position refers to the position vector found by all candidate solutions in previous iterations that satisfies the PPD constraint and minimizes the total energy consumption of the HVAC system; the velocity vector refers to the variable used to control the search direction and step size of the particle in the solution space, which determines the adjustment magnitude and direction of the position vector in the next iteration).

[0027] The speed update rule is expressed by the following formula: The location update rule is expressed by the following formula: Indicates the first iteration For particle position, For particle velocity, For the individual's historical best, For the best historical performance of the group, and For use as random vectors to enhance search diversity, For inertial weights, As a cognitive learning factor, This is a social learning factor. After each new location is generated, energy consumption and PPD are calculated using a simulation model to update the data. and .

[0028] After each new position is generated, the control scheme for the current particle is simulated, and the PPD value of each room during the occupied period is calculated. If the PPD of a room exceeds a preset threshold τ, the current candidate solution is determined to have violated the thermal comfort hard constraint. At this time, a directional bias correction is applied to the velocity component corresponding to the temperature setpoint of that room. If the current temperature setpoint is higher than the theoretical neutral comfort temperature, a negative correction (guided cooling) is applied; if it is lower than the neutral comfort temperature, a positive correction (guided heating) is applied. This bias correction algorithm enables the particle's search direction to actively move towards the region that meets the comfort constraint. After the position is updated, the nearest neighbor projection operator is used to project the continuous temperature and air supply level values ​​to the nearest discrete value in the candidate domain of discrete control variables, ensuring that the generated solution can be directly executed by the actual air conditioning terminal equipment (each air conditioning terminal equipment has a different candidate domain). The total energy consumption of the HVAC system corresponding to the current candidate solution is calculated, and a continuous penalty function is added to the solutions that do not meet the constraints, so that the priority of feasible solutions is always higher than that of infeasible solutions (infeasible solutions are candidate solutions that do not meet the constraints). The individual historical best and the group historical best are updated based on the calculation results. Repeat the above iterative process until convergence, and finally output the group's historical best position as the optimal fixed temperature setpoint and optimal fixed air supply level for each room.

[0029] In one embodiment of the present invention, the directional bias correction is expressed by the following formula: Candidate temperature setpoint values ​​that are closer to satisfying the constraints. For the current temperature setpoint, For the preset threshold, To adaptively adjust the step size, , This is the preset maximum correction step size. The temperature is set at a neutral and comfortable level.

[0030] In this embodiment, a PPD-guided correction mechanism is introduced to reduce infeasible sampling and accelerate entry into the feasible region. If a set of candidate solutions corresponds to... After the standard APSO velocity update and before the position update, a directional bias is applied to the velocity component of the corresponding temperature dimension, making particles more inclined to move in directions that improve comfort. It should be noted that this correction is only activated when an individual's historical best violates the PPD threshold, utilizing historical evaluation results without incurring additional computational overhead. If the threshold is not exceeded, the standard APSO update is retained to balance feasibility guidance and global exploration capabilities.

[0031] In the aforementioned directional bias correction, an adaptive correction step size is also introduced. This adaptive correction step size can be dynamically adjusted based on the degree to which the current PPD value exceeds a preset threshold. The more severe the PPD exceedance, the larger the correction step size, resulting in a greater shift of the temperature setpoint towards a neutral and comfortable temperature. When the PPD exceedance is relatively minor, the correction step size is correspondingly smaller to avoid overcorrection. Large step size corrections can be performed when PPD is severely exceeded to quickly pull infeasible solutions into the feasible region, improving search efficiency. When PPD is only slightly exceeded, small step size corrections are performed to prevent "overcorrection," avoid missing the optimal solution region, and ensure the quality of the final solution.

[0032] In one embodiment of the present invention, the above method further includes: During the directional bias correction process, the nearest neighbor projection operator is used to project the temperature setpoint value obtained after adaptive correction step size correction to the nearest discrete value in the candidate domain of the control variable, which is used as a candidate temperature setpoint value to move closer to the direction that satisfies the constraint conditions. The nearest neighbor projection operator is expressed by the following formula: Candidate temperature setpoint values ​​that are closer to satisfying the constraints. For the candidate domain of control variables, The closest discrete value within the candidate domain of the control variable. This is the temperature setpoint value obtained after adaptive correction step size adjustment.

[0033] In this embodiment, the nearest neighbor projection operator is used to project continuous values ​​onto a discrete candidate domain. This allows the temperature setpoint and air supply level output by the algorithm to be applicable to different air conditioning devices, ultimately enabling them to be identified and executed. This avoids control deviations caused by mismatches between control commands and device execution precision. Performing this nearest neighbor projection operation in each iteration ensures that the particle swarm search process always occurs within the discrete solution space, eliminating the need for subsequent additional discretization steps.

[0034] In a specific application example, an office building in a certain area is selected as the research object, and floors 2 to 4 of the office building are defined as the research scope. Controllable quantities that can be directly deployed through the Building Management System (BMS) include air volume levels and zone temperature setpoints. The building is functionally divided into 3 floors with a total of 29 zones (rooms), such as... Figure 4 As shown. Since there are no interconnected spaces between the floors, the optimization process is carried out floor by floor.

[0035] The mapping relationship between air volume level and wind speed is shown in Table 1. Optimization and evaluation were performed on a discrete-time grid with a simulation time step of 1 hour. Meteorological boundary conditions were based on typical local meteorological parameters. The working hours in the model were set from 8:00 to 19:00. Personnel occupancy, lighting, and equipment heat gain in the simulation were configured according to the design drawings and actual site conditions, and remained consistent across all scenarios. Apply hard constraints, threshold Set it to 0.2.

[0036] Table 1: Mapping Relationship between Air Supply Volume Level and Model Wind Speed Experiments were conducted in three groups (A1, A2, and A3) under scenario A, as shown in Table 2. All scenarios were conducted under the same building objects, evaluation periods, and comfort constraints, with consistent model parameters and simulation settings to ensure comparability between different scenarios.

[0037] Scenario A aims to verify the performance advantages of the two-stage solution framework over the traditional one-time joint optimization method, as well as its efficiency advantages over methods without a PPD guidance mechanism. It includes three comparative experiments: A1 employs a one-time strategy to jointly optimize temperature and airflow in the full variable space; A2 introduces a two-stage decomposition framework; and A3, based on A2, adds an adaptive PPD guidance mechanism to the particle velocity update.

[0038] Table 2: Comparison of Optimization Strategies for Three Groups of Experiments in Scenario A Energy consumption and PPD values ​​under scenario A are as follows Figure 5 As shown: In experimental group A1, the energy consumption was 843.61 kWh ( After introducing the two-stage decomposition framework, the energy consumption of experimental group A2 was reduced to 814.91 kWh. A3, which further incorporates the PPD guidance mechanism based on A2, achieves an energy consumption of 813.95 kWh. The results are almost identical to A2, but both are significantly better than A1. This indicates that, under the same constraints, the two-stage framework can achieve higher-quality energy-saving solutions. Meanwhile, as... Figure 4 As shown, the three scenarios All values ​​remained below the threshold value 1, with no comfort violations, indicating that the energy consumption differences mainly stemmed from the quality differences in the control combinations identified during the search process.

[0039] Taking 2F as an example, we further compared the hourly temperature setpoints and supply air volume levels obtained from three scenarios (A1, A2, and A3). The comparison results show that A2 and A3 produced highly consistent hourly temperature settings and distribution patterns, and also exhibited similarities in the regional distribution characteristics of supply air volume levels. In contrast, the temperature-supply air volume combination obtained in A1 failed to fully exploit the energy-saving potential within the feasible region. The comparison results are consistent with... Figure 5 The energy consumption results are consistent with those in the two-stage strategy, indicating that determining the low-dimensional parameters first and then conducting a refined search on the high-dimensional temperature sequence is more likely to yield better global optimization results.

[0040] also, Figure 6 The convergence process of the 2F strategy is illustrated. A1 experiences a rapid decrease in energy consumption in the early iterations before quickly plateauing, ultimately converging at a relatively high energy consumption level. This indicates that a one-time search in a high-dimensional joint decision space easily leads to premature convergence. For A2 and A3, the convergence behavior in the first stage is improved due to the reduced dimensionality, and the second stage maintains a stable downward trend, ultimately reaching a lower energy consumption level. Further comparison reveals that A3 exhibits a more compact descent trajectory in the second stage, reaching a final optimum close to that of A2 with fewer iterations. This demonstrates that the PPD-guided mechanism can improve search efficiency and accelerate the convergence of the two-stage strategy without compromising the quality of the final solution. Overall, A2 achieves a better energy-efficient solution than A1, but with a slower convergence speed; A3 achieves a similar energy efficiency to A2 while maintaining higher convergence efficiency, demonstrating better overall performance than A1.

[0041] In summary, the embodiments of the present invention provide a method and apparatus for optimizing air conditioning control based on PPD-guided two-stage operation, which has the following beneficial effects: The adopted two-stage control architecture, through the first stage reducing the air supply level to a fixed constant and the second stage optimizing the hourly temperature setpoint under a fixed air volume, greatly avoids the severe fragmentation of the feasible region caused by the joint optimization of temperature and air volume in large-scale multi-regional buildings, significantly reducing the computational dimensionality and effectively overcoming premature convergence due to local minima during the optimization process. A PPD-guided directional bias mechanism is proposed, applying a specific directional bias to the velocity component of particles without increasing additional computational costs. This forces candidate solutions with excessive PPD to move towards a neutral, comfortable temperature direction, enabling the algorithm to quickly escape infeasible solution regions without sacrificing the rigid requirements of human thermal comfort, significantly improving the efficiency of feasible solution generation. The bias correction process also introduces an adaptive correction step size, which can be dynamically adjusted according to the degree to which the current PPD value exceeds a preset threshold. When the PPD is severely excessive, a large step size correction is performed to quickly pull infeasible solutions into the feasible region, improving search efficiency. When the PPD is only slightly excessive, a small step size correction is performed to prevent "overcorrection," avoid missing the optimal solution region, and ensure the quality of the final solution. The nearest neighbor projection operator is used to discretize the continuous optimization results into adjustment step size and specific fan speed that can be identified by the building automation system. The output control commands can be directly issued and executed, which has real engineering deployability. In real building tests, a stable energy saving rate of about 9% was achieved while maintaining the thermal comfort of all personnel. Moreover, this method can quantify the flexible load boundary under the building comfort constraint, which can serve as a standard basis for building response to grid peak demand.

[0042] like Figure 2 , Figure 3 As shown, this embodiment of the invention provides a building flexible load scheduling device based on PPD-guided two-stage approach. The device embodiment can be implemented through software, hardware, or a combination of both. From a hardware perspective, as... Figure 2 The diagram shown is a hardware architecture diagram of an electronic device for a PPD-guided two-stage flexible building load scheduling device provided in an embodiment of the present invention. Besides... Figure 2 In addition to the processor, memory, network interface, and non-volatile memory shown, the electronic device in the embodiment may also include other hardware, such as a forwarding chip responsible for processing packets. Taking software implementation as an example, such as... Figure 3 As shown, a device in a logical sense is formed by the CPU of the electronic device in which it is located reading the corresponding computer program from the non-volatile memory into the memory for execution.

[0043] like Figure 3 As shown in the figure, this embodiment provides a PPD-guided two-stage building flexible load scheduling device, wherein the building flexible load includes a heating, ventilation, and air conditioning system, and the device includes: Initialize the data acquisition module 300 to determine the candidate domain of control variables for the air conditioning terminal equipment in the area to be optimized, as well as the set of time periods during which each room is occupied during working hours; The first-stage optimization module 302 is connected to the initialization data acquisition module. Under preset constraints, with the goal of minimizing the total energy consumption of the HVAC system in the area to be optimized, it solves the temperature setpoint and air supply level of the air conditioning terminal equipment in each room during the whole day's working time, and obtains the optimal fixed temperature setpoint and optimal fixed air supply level for each room. The preset constraint is that the PPD value of all rooms during the occupied period is lower than the preset threshold. The second-stage optimization module 304 is connected to the first-stage optimization module. It takes the optimal fixed air supply level as the air volume output value of each room. Under preset constraints, it takes the minimum total energy consumption of the HVAC system in the area to be optimized as the optimization objective. Using the optimal fixed temperature setpoint as the reference value, it solves the temperature setpoint corresponding to each room at different preset time points to obtain the optimal hourly temperature setpoint sequence. The control command output module 306 is connected to the second-stage optimization module. It uses the optimal hourly temperature setpoint sequence as the temperature output value of each room at different preset time points to regulate the temperature and air volume of each room in the area to be optimized, thereby realizing the scheduling of the building's flexible load.

[0044] In this embodiment of the invention, the initialization data acquisition module 300 can be used to execute step 100 in the above method embodiment, the first stage optimization module 302 can be used to execute step 102 in the above method embodiment, the second stage optimization module 304 can be used to execute step 104 in the above method embodiment, and the control instruction output module 306 can be used to execute step 106 in the above method embodiment.

[0045] In one embodiment of the present invention, the step of minimizing the total energy consumption of the HVAC system in the area to be optimized under preset constraints, and solving for the temperature setpoint and air supply level of the air conditioning terminal equipment in each room throughout the day to obtain the optimal fixed temperature setpoint and optimal fixed air supply level for each room, includes: Within the candidate domain of the control variables, select candidate values ​​for air supply level and temperature setpoint, and calculate the PPD value of each room during the occupied period for each set of candidate values. For candidate values ​​where the PPD value is higher than the preset threshold, the current temperature setpoint of the room is directionally offset and corrected according to the preset neutral comfort temperature of the corresponding room, so as to obtain a candidate temperature setpoint value that moves closer to the direction of satisfying the constraint conditions. The optimization objective is to minimize the total energy consumption of the HVAC system in the area to be optimized. The process is iterated repeatedly until the optimal fixed temperature setpoint and the optimal fixed air supply level for each room are selected.

[0046] In one embodiment of the present invention, the optimal fixed air supply level is used as the air volume output value for each room. Under preset constraints, with the goal of minimizing the total energy consumption of the HVAC system in the area to be optimized, and using the optimal fixed temperature setpoint as the reference value, the temperature setpoints corresponding to each room at different preset time points are solved to obtain the optimal hourly temperature setpoint sequence, including: Using the optimal fixed temperature setpoint as the baseline value, generate hourly temperature setpoint candidate values ​​for each room at different preset time points, and calculate the hourly PPD value of each room corresponding to each set of candidate values ​​during the occupied period. For the candidate values ​​corresponding to the time points where the PPD value is higher than the preset threshold, the temperature setpoint at the time point is directionally offset and corrected according to the preset neutral comfort temperature of the corresponding room, so as to obtain the candidate values ​​of the temperature setpoint that are closer to the direction of satisfying the constraint conditions. The optimization objective is to minimize the total energy consumption of the HVAC system in the area to be optimized. The process is iterated repeatedly until the optimal hourly temperature setpoint for each room at different preset time points is selected.

[0047] In one embodiment of the present invention, the optimization objective is expressed by the following formula: The total energy consumption of the HVAC system. The penalty coefficient is... This indicates that when the PPD value during any occupied time period exceeds a preset threshold, the function is activated. At that time, a penalty term is included in the optimization objective. Indicates the first The room at the preset time point PPD value at the location, Indicates the temperature output value. Candidate domain for control variables representing temperature. This indicates the air volume output value. Candidate domain for control variables representing air volume. This represents the set of time periods that are occupied by someone.

[0048] In one embodiment of the present invention, the directional bias correction is as follows: If the current temperature setpoint is greater than the preset neutral comfort temperature, apply a cooling direction bias correction to the current temperature setpoint to obtain a candidate temperature setpoint value that moves closer to the direction that satisfies the constraint conditions. If the current temperature setpoint is lower than the preset neutral comfort temperature, apply a bias correction to the current temperature setpoint in the direction of temperature rise to obtain a candidate temperature setpoint value that moves closer to the direction that satisfies the constraint conditions.

[0049] In one embodiment of the present invention, the directional bias correction is expressed by the following formula: Candidate temperature setpoint values ​​that are closer to satisfying the constraints. For the current temperature setpoint, For the preset threshold, To adaptively adjust the step size, , This is the preset maximum correction step size. The temperature is set at a neutral and comfortable level.

[0050] In one embodiment of the present invention, the method further includes: During the directional bias correction process, the nearest neighbor projection operator is used to project the temperature setpoint value obtained after adaptive correction step size correction to the nearest discrete value in the candidate domain of the control variable, which is used as a candidate temperature setpoint value to move closer to the direction that satisfies the constraint conditions. The nearest neighbor projection operator is expressed by the following formula: Candidate temperature setpoint values ​​that are closer to satisfying the constraints. For the candidate domain of control variables, The nearest discrete value within the candidate domain of the control variable. This is the temperature setpoint value obtained after adaptive correction step size adjustment.

[0051] It is understood that the structures illustrated in the embodiments of the present invention do not constitute a specific limitation on an air conditioning optimization control device based on PPD-guided two-stage control. In other embodiments of the present invention, an air conditioning optimization control device based on PPD-guided two-stage control may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0052] The information interaction and execution process between the modules in the above-mentioned device are based on the same concept as the method embodiment of the present invention, and the specific details can be found in the description of the method embodiment of the present invention, and will not be repeated here.

[0053] This invention also provides an electronic device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, it implements a PPD-guided two-stage air conditioning optimization control method according to any embodiment of this invention.

[0054] This invention also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program causes the processor to perform a PPD-guided two-stage air conditioning optimization control method according to any embodiment of this invention.

[0055] Specifically, a system or apparatus equipped with a storage medium may be provided, on which software program code implementing the functions of any of the embodiments described above is stored, and the computer (or CPU or MPU) of the system or apparatus may read and execute the program code stored in the storage medium.

[0056] In this case, the program code read from the storage medium can itself implement the function of any of the above embodiments, and therefore the program code and the storage medium storing the program code constitute part of the present invention.

[0057] Storage media embodiments for providing program code include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD+RW), magnetic tapes, non-volatile memory cards, and ROMs. Alternatively, program code can be downloaded from a server computer via a communication network.

[0058] Furthermore, it should be clear that not only can the program code read by the computer be executed, but also the operating system or other components operating on the computer can be instructed based on the program code to perform some or all of the actual operations, thereby achieving the function of any of the embodiments described above.

[0059] Furthermore, it is understood that the program code read from the storage medium is written to the memory set in the expansion board inserted into the computer or to the memory set in the expansion module connected to the computer. Then, based on the instructions of the program code, the CPU or other components installed on the expansion board or expansion module execute some and all of the actual operations, thereby realizing the function of any of the above embodiments.

[0060] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0061] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as ROM, RAM, magnetic disk, or optical disk.

[0062] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A building flexible load scheduling method based on PPD-guided two-stage process, characterized in that, The building's flexible load includes a heating, ventilation, and air conditioning system, and the method includes: Determine the candidate domain of control variables for air conditioning terminal equipment in the area to be optimized, as well as the set of time periods during which each room is occupied during working hours; Under preset constraints, with the goal of minimizing the total energy consumption of the HVAC system in the area to be optimized, the temperature setpoint and air supply level of the air conditioning terminal equipment in each room during the entire working day are solved to obtain the optimal fixed temperature setpoint and optimal fixed air supply level for each room. The preset constraints are: the PPD value of all rooms during the occupied period is lower than the preset threshold. Using the optimal fixed air supply level as the air volume output value of each room, under preset constraints, with the goal of minimizing the total energy consumption of the HVAC system in the area to be optimized, and using the optimal fixed temperature setpoint as the reference value, the temperature setpoint corresponding to each room at different preset time points is solved to obtain the optimal hourly temperature setpoint sequence. The optimal hourly temperature setpoint sequence is used as the temperature output value of each room at different preset time points to regulate the temperature and air volume of each room in the area to be optimized, thereby realizing the scheduling of the building's flexible load.

2. The method according to claim 1, characterized in that, Under preset constraints, with the goal of minimizing the total energy consumption of the HVAC system in the area to be optimized, the process involves solving for the temperature setpoint and air supply level of the air conditioning terminal equipment in each room throughout the day, obtaining the optimal fixed temperature setpoint and optimal fixed air supply level for each room, including: Within the candidate domain of the control variables, select candidate values ​​for air supply level and temperature setpoint, and calculate the PPD value of each room during the occupied period for each set of candidate values. For candidate values ​​where the PPD value is higher than the preset threshold, the current temperature setpoint of the room is directionally offset and corrected according to the preset neutral comfort temperature of the corresponding room, so as to obtain a candidate temperature setpoint value that moves closer to the direction of satisfying the constraint conditions. The optimization objective is to minimize the total energy consumption of the HVAC system in the area to be optimized. The process is iterated repeatedly until the optimal fixed temperature setpoint and the optimal fixed air supply level for each room are selected.

3. The method according to claim 1, characterized in that, The optimal fixed air supply level is used as the air volume output value for each room. Under preset constraints, the optimization objective is to minimize the total energy consumption of the HVAC system in the area to be optimized. Using the optimal fixed temperature setpoint as the reference value, the temperature setpoints corresponding to each room at different preset time points are solved to obtain the optimal hourly temperature setpoint sequence, including: Using the optimal fixed temperature setpoint as the baseline value, generate hourly temperature setpoint candidate values ​​for each room at different preset time points, and calculate the hourly PPD value of each room corresponding to each set of candidate values ​​during the occupied period. For the candidate values ​​corresponding to the time points where the PPD value is higher than the preset threshold, the temperature setpoint at the time point is directionally offset and corrected according to the preset neutral comfort temperature of the corresponding room, so as to obtain the candidate values ​​of the temperature setpoint that are closer to the direction of satisfying the constraint conditions. The optimization objective is to minimize the total energy consumption of the HVAC system in the area to be optimized. The process is iterated repeatedly until the optimal hourly temperature setpoint for each room at different preset time points is selected.

4. The method according to any one of claims 1-3, characterized in that, The optimization objective is expressed by the following formula: The total energy consumption of the HVAC system. The penalty coefficient is... This indicates that when the PPD value during any occupied time period exceeds a preset threshold, the function is activated. At that time, a penalty term is included in the optimization objective. Indicates the first The room at the preset time point PPD value at the location, Indicates the temperature output value. Candidate domain for control variables representing temperature. This indicates the air volume output value. Candidate domain for control variables representing air volume. This represents the set of time periods that are occupied by someone.

5. The method according to claim 2 or 3, characterized in that, The directional bias correction is as follows: If the current temperature setpoint is greater than the preset neutral comfort temperature, apply a cooling direction bias correction to the current temperature setpoint to obtain a candidate temperature setpoint value that moves closer to the direction that satisfies the constraint conditions. If the current temperature setpoint is lower than the preset neutral comfort temperature, apply a bias correction to the current temperature setpoint in the direction of temperature rise to obtain a candidate temperature setpoint value that moves closer to the direction that satisfies the constraint conditions.

6. The method according to claim 5, characterized in that, The directional bias correction is expressed by the following formula: Candidate temperature setpoint values ​​that are closer to satisfying the constraints. For the current temperature setpoint, For the preset threshold, To adaptively adjust the step size, , This is the preset maximum correction step size. The temperature is set at a neutral and comfortable level.

7. The method according to claim 6, characterized in that, The method further includes: During the directional bias correction process, the nearest neighbor projection operator is used to project the temperature setpoint value obtained after adaptive correction step size correction to the nearest discrete value in the candidate domain of the control variable, which is used as a candidate temperature setpoint value to move closer to the direction that satisfies the constraint conditions. The nearest neighbor projection operator is expressed by the following formula: Candidate temperature setpoint values ​​that are closer to satisfying the constraints. For the candidate domain of control variables, The nearest discrete value within the candidate domain of the control variable. This is the temperature setpoint value obtained after adaptive correction step size adjustment.

8. A building flexible load dispatching device based on PPD-guided two-stage process, characterized in that, The building's flexible load includes a heating, ventilation, and air conditioning system, and the device includes: Initialize the data acquisition module to determine the candidate domain of control variables for air conditioning terminal equipment in the area to be optimized, as well as the set of time periods during which each room is occupied during working hours; The first-stage optimization module is connected to the initialization data acquisition module. Under preset constraints, with the goal of minimizing the total energy consumption of the HVAC system in the area to be optimized, it solves the temperature setpoint and air supply level of the air conditioning terminal equipment in each room during the whole day's working time, and obtains the optimal fixed temperature setpoint and optimal fixed air supply level for each room. The preset constraint is that the PPD value of all rooms during the occupied period is lower than the preset threshold. The second-stage optimization module is connected to the first-stage optimization module. It takes the optimal fixed air supply level as the air volume output value of each room. Under preset constraints, it takes the minimum total energy consumption of the HVAC system in the area to be optimized as the optimization objective. Using the optimal fixed temperature setpoint as the reference value, it solves the temperature setpoint corresponding to each room at different preset time points to obtain the optimal hourly temperature setpoint sequence. The control command output module is connected to the second-stage optimization module. It uses the optimal hourly temperature setpoint sequence as the temperature output value of each room at different preset time points to regulate the temperature and air volume of each room in the area to be optimized, thereby realizing the scheduling of the building's flexible load.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor, when executing the computer program, implements the method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, which, when executed in a computer, causes the computer to perform the method of any one of claims 1-7.

Images