Feasible domain aggregation methods, apparatus, equipment, storage media and products

By acquiring the operating parameters of the temperature control equipment and the uncertainty information of the external environment, an internal approximate polyhedron parallel to the constraint hyperplane is constructed, which solves the problems of insufficient constraint capture and uncertainty handling in temperature control load aggregation, and realizes efficient and robust feasible domain aggregation.

CN123069684APending Publication Date: 2026-11-10GUO WANG ZHE JIANG SHENG DIAN LI YOU XIAN GONG SI CI XI SHI GONG DIAN GONG SI +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611278413.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-21
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing technologies cannot accurately capture constraints with continuous thermal dynamic characteristics when polymerizing temperature-controlled loads, resulting in poor internal approximation accuracy and failing to effectively handle the uncertainty of outdoor temperature prediction, leading to poor robustness of polymerization results in real-world scenarios.

Method used

By acquiring the operating parameters of controllable equipment and information on external environmental uncertainties, a set of candidate operating intervals is constructed, and an internal approximate polyhedron parallel to the constraint hyperplane is generated. Summation and aggregation are then performed to ensure that the aggregation result is strictly feasible within each individual device.

Benefits of technology

It achieves robustness in aggregating feasible domains under uncertain external environments, avoids the risk of scheduling instructions exceeding limits, and ensures the flexibility and schedulability of temperature-controlled load clusters.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123069684A_ABST
    Figure CN123069684A_ABST
Patent Text Reader

Abstract

This application discloses a feasible domain aggregation method, apparatus, device, storage medium, and product, relating to the field of demand-side resource aggregation technology in power systems. The feasible domain aggregation method includes: acquiring operating parameter information of each controllable device to be aggregated and uncertainty information of the external environment in which the controllable devices are located; constructing candidate operation interval sets for each controllable device based on the operating parameter information and the uncertainty information, where each element in the candidate operation interval set corresponds to an operational constraint under a deviation state of the external environment; performing internal approximation processing on each candidate operation interval set to generate an internal approximation polyhedron parallel to the boundary hyperplane of the candidate operation interval set; and summing and aggregating the internal approximation polyhedra of each controllable device to obtain the aggregated feasible domain of the entire controllable device cluster. This application solves the technical problem of poor robustness in feasible domain aggregation of temperature-controlled loads in real-world scenarios.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of demand-side resource aggregation technology in power systems, and in particular to feasible domain aggregation methods, apparatus, equipment, storage media, and products. Background Technology

[0002] With a high proportion of renewable energy being integrated into the power system, massive flexible loads on the demand side have become a crucial resource for enhancing grid flexibility. Temperature-controlled loads are numerous and exhibit diverse thermal dynamic characteristics; aggregating them into a unified, dispatchable model is key to unlocking their regulatory potential. In conventional techniques, aggregation methods based on chino polyhedra leverage their low computational complexity (Minkowski sum) to achieve rapid aggregation of large-scale resources. However, the chino polyhedron generator construction in these methods can only handle standardized constraints with simple numerical values. In contrast, the constraint coefficients of temperature-controlled loads, after transformation by the thermal dynamic model, contain continuously changing thermodynamic parameters. Conventional generator construction methods cannot accurately capture the geometric characteristics of such constraints, resulting in poor internal approximation accuracy.

[0003] To address the aforementioned issues, some technologies employ virtual battery models to simplify the description of temperature-controlled loads and combine them with external approximation methods for aggregation. However, virtual battery models cannot accurately depict real thermal dynamic processes, and the aggregated feasible region obtained through external approximation may include power points that are practically infeasible. This poses a risk of violating temperature comfort constraints when decomposing scheduling instructions to individual devices. Furthermore, conventional technologies are all based on deterministic environmental parameter modeling and do not consider the impact of uncertainties in outdoor temperature prediction on the feasible region boundary, resulting in poor robustness of the aggregation results in real-world scenarios. Summary of the Invention

[0004] The main objective of this application is to provide a feasible domain aggregation method, apparatus, device, storage medium, and product, which aims to solve the technical problem of poor robustness in feasible domain aggregation of temperature-controlled loads in real-world scenarios.

[0005] To achieve the above objectives, this application proposes a feasible domain aggregation method, which includes: Obtain the operating parameter information of each controllable device to be aggregated, as well as the uncertainty information of the external environment in which the controllable devices are located; Based on the operating parameter information and the uncertainty information, a candidate operation interval set is constructed for each controllable device. Each element in the candidate operation interval set corresponds to an operation constraint under a deviation state of the external environment. The candidate operation interval sets are subjected to internal approximation processing to generate an internal approximate polyhedron that is parallel to the boundary hyperplane of the candidate operation interval sets. The approximate polyhedra of each controllable device are summed and aggregated to obtain the aggregated feasible region of the entire controllable device cluster.

[0006] In one embodiment, the step of constructing a set of candidate operating ranges for each controllable device based on the operating parameter information and the uncertainty information includes: Based on the operating parameter information and the uncertainty information, calculate the deviation state values ​​of the external environment; Calculate the constraint boundaries corresponding to each of the aforementioned deviation state values; The intersection of the constraint boundaries corresponding to each deviation state value is used as the set of candidate operation intervals.

[0007] In one embodiment, the step of calculating the deviation state values ​​of the external environment based on the operating parameter information and the uncertainty information includes: Read the pre-configured baseline state value and the pre-configured deviation distribution information, wherein the deviation distribution information includes the magnitude of each deviation and its corresponding probability of occurrence; The deviation state values ​​of the external environment are generated based on the baseline state value and the deviation distribution information.

[0008] In one embodiment, the operating parameter information of the controllable device includes thermal resistance parameter, heat capacity parameter and energy efficiency ratio parameter, and the operating parameter information is described by a second-order equivalent thermal parameter model; the operating constraints include device output power limit constraints and temperature state constraints, and the temperature state constraints are obtained by iterative expansion of the state equations from the operating parameter information, and the coefficients of the temperature state constraints include the thermal resistance parameter and the heat capacity parameter.

[0009] In one embodiment, the step of calculating the constraint boundary corresponding to each of the deviation state values ​​includes: For each of the aforementioned deviation state values, corresponding constraint boundaries are generated. Each of the aforementioned constraint boundaries is represented in the form of a system of linear inequalities. The coefficient matrix of each of the aforementioned linear inequalities is calculated by combining the aforementioned deviation state values ​​with the operating parameter information. The constant vector of each of the aforementioned linear inequalities is calculated by combining the aforementioned deviation state values ​​with the initial state of the controllable device.

[0010] In one embodiment, the step of performing inner approximation processing on each of the candidate operation interval sets to generate an inner approximate polyhedron that remains parallel to the boundary hyperplane of the candidate operation interval sets includes: Obtain the normal vector of each constraint hyperplane in the candidate operation interval set; Based on each of the normal vectors, a generator set parallel to the corresponding constraint hyperplane is constructed, wherein each generator in the generator set is in a parallel direction to the constraint hyperplane; Construct the inner approximate polyhedron based on the generator set.

[0011] In one embodiment, the step of constructing a generator set parallel to the corresponding constraint hyperplane based on each of the said normal vectors includes: Determine the number of dimensions in the decision variable space of the controllable device; For any of the constrained hyperplanes, the total number of generators is calculated by subtracting one from the dimension number. Each generator corresponding to the constrained hyperplane is constructed, and the position of the non-zero component in each generator is determined alternately based on the adjacent components of the normal vector. Each of the generators is used as the generator set corresponding to the constrained hyperplane.

[0012] In one embodiment, the step of summing and aggregating the inner approximate polyhedra of each controllable device includes: Obtain the center point, generator matrix, and upper limit of the scaling factor for each of the aforementioned internal approximate polyhedra; The center points are merged into an aggregate center point, the generator matrices are merged into an aggregate generator matrix, and the upper limits of the scaling coefficients are merged into an aggregate scaling coefficient upper limit. The feasible region for aggregation is generated based on the aggregation center point, the aggregation generator matrix, and the upper limit of the aggregation scaling factor.

[0013] In one embodiment, the steps of merging the central points into an aggregate central point, merging the generator matrices into an aggregate generator matrix, and merging the upper limits of the scaling factors into an aggregate scaling factor upper limit include: The corresponding dimensional coordinates of each of the aforementioned center points are added together, and the resulting coordinates are used as the coordinates of the aggregate center point. The generator matrices are concatenated sequentially along the column direction, and the resulting matrix is ​​used as the aggregate generator matrix. The upper limits of each of the above-mentioned scaling coefficients are concatenated sequentially in the row direction, and the resulting vector is used as the upper limit of the aggregated scaling coefficient.

[0014] In one embodiment, after the step of performing inner approximation processing on each of the candidate operation interval sets, the method further includes: Obtain the coordinates of all vertices of each of the aforementioned internal approximate polyhedra; Determine whether the coordinates of each vertex satisfy the operation constraints in the corresponding candidate operation interval set; When all vertex coordinates satisfy the aforementioned operational constraints, it is determined that all points of the inner approximate polyhedron are within the corresponding candidate operation interval set.

[0015] In one embodiment, before the step of obtaining the operating parameter information of each controllable device to be aggregated and the uncertainty information of the external environment in which the controllable devices are located, the method further includes: Obtain a pre-configured list of device clusters, which contains the device identifiers of each controllable device to be aggregated; Based on the device identifier, the inherent parameters of each controllable device are read as the operating parameter information, and the statistical characteristic values ​​of the historical measurement data of the environmental sensors at the location of each controllable device are read as the uncertainty information.

[0016] Furthermore, to achieve the above objectives, this application also proposes a feasible domain aggregation apparatus, which includes: The acquisition module is used to acquire the operating parameter information of each controllable device to be aggregated, as well as the uncertainty information of the external environment in which the controllable devices are located; The construction module is used to construct a candidate operation interval set for each controllable device based on the operating parameter information and the uncertainty information. Each element in the candidate operation interval set corresponds to the operation constraint under a deviation state of the external environment. The generation module is used to perform inner approximation processing on each of the candidate operation interval sets to generate an inner approximate polyhedron that is parallel to the boundary hyperplane of the candidate operation interval set. The aggregation module is used to sum and aggregate the internal approximate polyhedra of each controllable device to obtain the aggregated feasible region of the entire controllable device cluster.

[0017] In addition, to achieve the above objectives, this application also proposes a feasible domain aggregation device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the feasible domain aggregation method as described above.

[0018] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and which, when executed by a processor, implements the steps of the feasible domain aggregation method as described above.

[0019] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the feasible domain aggregation method as described above.

[0020] One or more technical solutions proposed in this application have at least the following technical effects: Compared to related technologies that construct feasible regions based solely on deterministic environmental parameters and cannot handle constraints with continuous thermodynamic parameters, this application obtains the operating parameter information of each controllable device to be aggregated, as well as the uncertainty information of the external environment in which the controllable devices are located; based on the operating parameter information and the uncertainty information, it constructs candidate operating interval sets for each controllable device, where each element in the candidate operating interval set corresponds to an operating constraint under a deviation state of the external environment; it performs internal approximation processing on each candidate operating interval set to generate an internal approximation polyhedron parallel to the boundary hyperplane of the candidate operating interval set; and it sums and aggregates the internal approximation polyhedra of each controllable device to obtain the aggregated feasible region of the controllable device cluster as a whole. Understandably, this application introduces uncertainty information about the external environment during the feasible region construction stage. It generates various deviation state values ​​of the external environment through pre-configured baseline state values ​​and deviation distribution information, and generates a corresponding set of operational constraints for each deviation state. The intersection of operational constraints under each deviation state is taken as the candidate operational interval set, ensuring that the aggregation result naturally covers the physical feasible boundary under various possible deviations in the external environment. Simultaneously, by performing internal approximation processing on the candidate operational interval set and generating an internal approximation polyhedron parallel to the constraint hyperplane, it ensures that the aggregation result is strictly within the true feasible region of each individual unit, preventing the introduction of practically infeasible power points during the aggregation stage. This ensures that even when there are deviations in the external environment prediction, the aggregated feasible region can still guarantee that hard constraints such as the temperature comfort of each individual device are not violated. Therefore, it avoids the risk of scheduling instructions exceeding limits in real-world scenarios in deterministic feasible regions, as seen in related technologies, thus ensuring the effectiveness of the aggregated feasible region under uncertain environments. Ultimately, it achieves robust aggregation of temperature-controlled load clusters, solving the technical problem of poor robustness in the aggregation of temperature-controlled load feasible regions in real-world scenarios. Attached Figure Description

[0021] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0022] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 A flowchart illustrating an embodiment of the feasible domain aggregation method of this application; Figure 2 A method flowchart is provided for an embodiment of the feasible domain aggregation method of this application; Figure 3A schematic diagram of the improved Zonotope generator construction and inner approximation principle provided for embodiments of the feasible domain aggregation method of this application; Figure 4 A comparison diagram of the robust feasible domain of a single entity and the approximate size of the Chino polyhedron provided in the embodiments of the feasible domain aggregation method of this application; Figure 5 A comparison diagram of Minkowski and aggregation similarity of cluster aggregation feasible regions provided for embodiments of the feasible region aggregation method of this application; Figure 6 This is a schematic diagram of the modular structure of a feasible domain aggregation device according to an embodiment of this application; Figure 7 This is a schematic diagram of the device structure of the hardware operating environment involved in the feasible domain aggregation method in this application embodiment.

[0024] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0025] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0026] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0027] The main solution in this application embodiment is: Obtain the operating parameter information of each controllable device to be aggregated, as well as the uncertainty information of the external environment in which the controllable devices are located; Based on the operating parameter information and the uncertainty information, a candidate operation interval set is constructed for each controllable device. Each element in the candidate operation interval set corresponds to an operation constraint under a deviation state of the external environment. The candidate operation interval sets are subjected to internal approximation processing to generate an internal approximate polyhedron that is parallel to the boundary hyperplane of the candidate operation interval sets. The approximate polyhedra of each controllable device are summed and aggregated to obtain the aggregated feasible region of the entire controllable device cluster.

[0028] In this embodiment, the feasible domain aggregation device is used as the execution subject. For ease of description, it will be referred to as "device" in the following detailed description.

[0029] Because related technologies construct feasible regions based solely on deterministic environmental parameters, and the Kino polyhedron generator construction method cannot handle complex constraints containing continuous thermodynamic parameters, the robustness of the aggregation feasible region in real-world scenarios is poor. This application provides a solution by introducing external environmental uncertainty information into the feasible region construction stage and employing a generator construction method parallel to the constraint hyperplane to achieve high-precision internal approximation, ensuring that the physical constraints of each individual device are not violated even when environmental parameters deviate.

[0030] This embodiment aims to address the robust aggregation problem of temperature-controlled load clusters when outdoor temperature prediction is uncertain. The research background addressed in this embodiment is as follows: conventional feasible region aggregation methods for temperature-controlled loads either model based on deterministic outdoor temperatures, ignoring the impact of prediction errors on the feasible region boundary; or they employ simplified virtual battery models, failing to accurately depict real thermal dynamics. Both methods result in poor robustness of the aggregated feasible region in real-world scenarios. The application scenario addressed in this embodiment is: when large-scale heterogeneous temperature-controlled loads participate in day-ahead dispatch or intraday rolling control of the power grid, it is necessary to aggregate a massive number of individual devices into a flexible overall model that can be directly invoked by the dispatch center, and this model must remain effective even when there are deviations between the actual and predicted outdoor temperatures.

[0031] This application proposes, firstly, obtaining the operating parameter information and external environmental uncertainty information of each controllable device to be aggregated. The operating parameter information describes the thermal dynamic characteristics of the device, and the uncertainty information characterizes the possible deviation range of outdoor temperature prediction. Based on this information, a candidate operating interval set for each controllable device is constructed—each element in this set corresponds to an operating constraint under a deviation state of the external environment. Specifically, for each possible outdoor temperature deviation scenario, a set of constraints that the device power and indoor temperature must satisfy is generated. The intersection of the constraints under each scenario is taken as the candidate operating interval set, ensuring that any power sequence within the set does not violate the device's physical constraints and user temperature comfort requirements under any deviation scenario. Furthermore, an internal approximation process is performed on each candidate operating interval set: the normal vectors of each constraint hyperplane in the candidate operating interval set are obtained; a generator set parallel to the corresponding constraint hyperplane is constructed based on each normal vector; and then an internal approximation polyhedron is constructed based on the generator set. Because the generator remains parallel to the constraint hyperplane, the faces of the inner approximate polyhedra fit tightly with the constraint boundaries of the original feasible region. This minimizes the approximate volume loss while ensuring that the inner approximate polyhedra are completely contained within the original feasible region. Finally, the inner approximate polyhedra of each controllable device are summed and aggregated—the coordinates of the center points of each inner approximate polyhedron are merged into an aggregate center point, the generator matrices are merged into an aggregate generator matrix, and the upper limits of each scaling factor are merged into an aggregate scaling factor upper limit. The aggregated feasible region is then generated based on the merged parameters. Since the aggregation process only involves vector addition and matrix concatenation, the computational complexity is linearly related to the number of devices, allowing large-scale cluster aggregation to be completed in milliseconds.

[0032] Therefore, this application incorporates external environmental uncertainties during the feasible region construction stage, enabling the aggregated feasible region to naturally withstand outdoor temperature prediction deviations. Simultaneously, by employing an improved generator construction method to achieve high-precision internal approximation, it ensures that the aggregated feasible region strictly resides within the true feasible regions of each individual unit, fundamentally eliminating the risk of scheduling overruns caused by infeasible power points introduced through external approximation in related technologies. Thus, this application effectively improves the robustness of temperature-controlled load feasible region aggregation in real-world scenarios.

[0033] Based on this, embodiments of this application provide a feasible domain aggregation method, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the feasible domain aggregation method of this application.

[0034] In this embodiment, the feasible domain aggregation method includes steps S10 to S40: Step S10: Obtain the operating parameter information of each controllable device to be aggregated and the uncertainty information of the external environment in which the controllable devices are located; It should be noted that controllable equipment refers to electrical equipment with power regulation capabilities whose operating state is constrained by a temperature range, including but not limited to heat pump equipment, air conditioning equipment, electric heating equipment, and electric refrigeration equipment. Operating parameter information refers to the set of parameters used to describe the physical characteristics and operating conditions of the controllable equipment, including but not limited to the equipment's thermal resistance parameters, heat capacity parameters, energy efficiency ratio parameters, rated power limits, and the user-set temperature comfort range. External environment refers to the external meteorological conditions of the space where the controllable equipment is located. Uncertainty information refers to the probability distribution characteristics of the possible deviation range and magnitude between the predicted and actual values ​​of the external environment in the future, used to quantify the degree of uncertainty in outdoor temperature prediction.

[0035] Understandably, this step, by acquiring operational parameter information of the controllable equipment and uncertainty information of the external environment, provides a data foundation for subsequently constructing a robust feasible region capable of withstanding environmental prediction biases. Since the operational parameter information encompasses the thermodynamic characteristics and operational constraint boundaries of the equipment, and the uncertainty information covers the possible deviation range of environmental parameters, this step ensures that the subsequently constructed feasible region possesses a complete physical description and a quantitative characterization of uncertainty at the input information level, thus supporting the inclusion of robustness considerations during the feasible region construction phase.

[0036] Step S20: Based on the operating parameter information and the uncertainty information, construct a candidate operation interval set for each controllable device. Each element in the candidate operation interval set corresponds to an operation constraint under a deviation state of the external environment. It should be noted that the candidate operating interval set refers to the power feasible region jointly enclosed by multiple operating constraint boundaries corresponding to various deviation states of the external environment. Each element in this set corresponds to all the operating constraints that the device power sequence must satisfy under a certain deviation state of the external environment. A deviation state refers to the deviation of the actual value of the external environment from the baseline predicted value, and is used to characterize a specific realization of environmental uncertainty. Operating constraints refer to the physical limitations that the device must comply with during operation, including but not limited to the rated limit constraint of the device's output power and the temperature state constraint of maintaining the indoor temperature within the user's comfort range. The temperature state constraint is obtained by iteratively expanding the state equation described by the second-order equivalent thermal parameter model based on the operating parameter information.

[0037] Understandably, this step, by constructing a set of candidate operation intervals based on operational parameters and uncertainty information, essentially extends the deterministic feasible region construction method to uncertain scenarios. Since the candidate operation interval set requires the power sequence to satisfy all operational constraints under every deviation state of the external environment, any power sequence within this set possesses a natural resilience against outdoor temperature prediction deviations. Compared to conventional techniques that construct feasible regions based on only a single prediction scenario, this step introduces a constraint intersection mechanism under multiple deviation states, ensuring that the aggregated feasible region remains effective even when real-world environmental parameters deviate from predicted values. This fundamentally solves the scheduling limit exceedance risk caused by ignoring uncertainty in related technologies.

[0038] Step S30: Perform inner approximation processing on each candidate operation interval set to generate an inner approximation polyhedron that is parallel to the boundary hyperplane of the candidate operation interval set; It should be noted that internal approximation refers to the operation of approximating the candidate operation interval set internally using a pre-defined geometric shape. The internal approximation polyhedron is a polyhedron completely contained within the candidate operation interval set, with each face parallel to the boundary hyperplane of the candidate operation interval set. It is determined by three parameters: the center point, the generator matrix, and the upper limit of the scaling coefficient. Each column vector in the generator matrix is ​​a generator, and each generator is parallel to the boundary hyperplane of the candidate operation interval set. The boundary hyperplane refers to the planar boundary in the power variable space corresponding to each linear constraint enclosing the candidate operation interval set.

[0039] Understandably, this step, by performing an internal approximation on each candidate operation interval set and generating an internal approximation polyhedron parallel to the boundary hyperplane, essentially approximates the original feasible region internally using a polyhedron aligned with the constraint direction. Since the faces of the internal approximation polyhedron remain parallel to the boundary hyperplane, they form surface contact rather than degenerate point contact, allowing the internal approximation polyhedron to approximate the original boundary to the maximum extent possible while remaining strictly within the candidate operation interval set. Compared to conventional techniques where the generator can only be aligned with simple coordinate axes, this step, by making the generator parallel to the constraint hyperplane in any direction, can adapt to complex constraint geometries containing continuous thermodynamic parameters, significantly reducing the internal approximation volume loss and improving flexibility.

[0040] Step S40: Summate and aggregate the internal approximate polyhedra of each controllable device to obtain the aggregated feasible region of the entire controllable device cluster.

[0041] It should be noted that summation aggregation refers to the operation of merging the parameters of multiple internal approximate polyhedra according to vector addition and matrix concatenation rules. The aggregation feasible region refers to the flexible and adjustable range of the entire controllable device cluster, which is determined by three parameters: the aggregation center point, the aggregation generator matrix, and the upper limit of the aggregation scaling coefficient.

[0042] Understandably, this step, by summing and aggregating the internal approximate polyhedra of each controllable device, essentially leverages the closure property of these internal approximate polyhedra in the summation and aggregation operation to transform the aggregation of individual feasible regions into a simple merging at the parameter level. Since the aggregation process only involves vector addition of the center point coordinates, column concatenation of the generator matrix, and row concatenation of the upper limit of the scaling factor, the computational cost is linearly related to the number of devices, enabling large-scale cluster aggregation to be completed in milliseconds. Compared to the computational explosion problem faced by conventional techniques when accurately aggregating high-dimensional complex polyhedra, this step significantly improves computational efficiency while maintaining aggregation accuracy, making online aggregation and re-aggregation possible and possessing engineering practicality.

[0043] This embodiment provides a feasible region aggregation method. This application introduces uncertainty information of the external environment during the feasible region construction stage. Various deviation state values ​​of the external environment are generated using pre-configured baseline state values ​​and deviation distribution information. A corresponding set of operational constraints is generated for each deviation state. The intersection of operational constraints under each deviation state is taken as a set of candidate operational intervals, ensuring that the aggregation result naturally covers the physical feasible boundary under various possible deviations in the external environment. Simultaneously, by performing an internal approximation on the candidate operational interval set and generating an internal approximation polyhedron parallel to the constraint hyperplane, it is ensured that the aggregation result is strictly within the true feasible region of each individual unit, preventing the introduction of practically infeasible power points during the aggregation stage. This ensures that even when there are deviations in the predicted external environment, the aggregated feasible region can still guarantee that hard constraints such as temperature comfort of each individual device are not violated. Therefore, it avoids the risk of scheduling instructions exceeding limits in real-world scenarios in deterministic feasible regions, as seen in related technologies, thus ensuring the effectiveness of the aggregated feasible region under uncertain environments. Ultimately, robust aggregation of temperature-controlled load clusters is achieved, solving the technical problem of poor robustness in feasible region aggregation of temperature-controlled loads in real-world scenarios.

[0044] In one feasible implementation, the step of constructing a set of candidate operating ranges for each controllable device based on the operating parameter information and the uncertainty information includes: Based on the operating parameter information and the uncertainty information, calculate the deviation state values ​​of the external environment; Calculate the constraint boundaries corresponding to each of the aforementioned deviation state values; The intersection of the constraint boundaries corresponding to each deviation state value is used as the set of candidate operation intervals.

[0045] It should be noted that the deviation state value refers to the state value determined by the deviation of the actual parameters of the external environment from the baseline predicted parameters, and is a specific implementation used to quantify the uncertainty of the external environment. The constraint boundary refers to the linear inequality boundary formed by the physical constraints of equipment operation in the power variable space, including the rated limit boundary of the equipment's output power and the temperature state boundary for maintaining the indoor temperature within the user's comfort range. These constraint boundaries together enclose the power feasible region of the equipment under a single deviation state. The intersection refers to the common part of the regions enclosed by the corresponding constraint boundaries under different deviation states. Any power sequence within the intersection satisfies all operational constraints under every deviation state of the external environment.

[0046] It is understandable that this implementation calculates the deviation state values ​​of the external environment based on operating parameter information and uncertainty information, and then calculates the constraint boundaries corresponding to each deviation state value and takes the intersection. Essentially, it integrates external environmental uncertainty into the feasible region construction process in the form of discrete deviation scenarios. Since the candidate operation interval set requires the power sequence to satisfy all physical constraints in every deviation state, any power sequence within this set has the ability to withstand outdoor temperature prediction deviations. Compared to related technologies that construct feasible regions based only on a single prediction scenario, this step, through a constraint intersection mechanism under multiple deviation states, ensures that the aggregated feasible region remains effective when real environmental parameters deviate from the predicted values, incorporating robustness considerations from the very beginning of feasible region construction.

[0047] In one feasible implementation, the step of calculating the deviation state values ​​of the external environment based on the operating parameter information and the uncertainty information includes: Read the pre-configured baseline state value and the pre-configured deviation distribution information, wherein the deviation distribution information includes the magnitude of each deviation and its corresponding probability of occurrence; The deviation state values ​​of the external environment are generated based on the baseline state value and the deviation distribution information.

[0048] It should be noted that the baseline state value refers to the predicted parameter value of the external environment under unbiased conditions, such as the predicted average outdoor temperature, which serves as the benchmark for generating each deviation state value. Deviation distribution information refers to the statistical characteristic description of the predicted deviations of the external environment, including the magnitude of each deviation and its corresponding probability of occurrence. The deviation magnitude refers to the amount of deviation of the actual external environment value from the baseline state value, and the probability of occurrence refers to the probability that each deviation magnitude may occur during actual operation. Reading the pre-configuration refers to retrieving the pre-set baseline state value and deviation distribution information from pre-stored configuration data. This configuration behavior is completed before the method execution and is not considered an online iterative step in this method.

[0049] It is understandable that this implementation method, by reading pre-configured baseline state values ​​and deviation distribution information, and generating various deviation state values ​​of the external environment based on these two, essentially transforms the uncertainty of the external environment from an abstract description of statistical characteristics into a set of specific scenarios that can be used for constraint calculation. Since the deviation distribution information includes the magnitude of each deviation and its corresponding probability of occurrence, this step can flexibly configure the generation strategy of deviation state values ​​according to the actual application scenario. For example, it can select deviation magnitudes with a higher probability of occurrence as typical scenarios, or select a set of scenarios covering a wider range of deviations based on the robustness requirements of the scheduling. This implementation method also considers the specific application scenario of subsequently constructing a set of candidate operation intervals when generating each deviation state value. Simultaneously, by reading the pre-configured information, the configuration of baseline state values ​​and deviation distribution information is excluded from the execution loop, eliminating the need to repeatedly configure baseline and deviation-related data during multiple aggregation calculations, thus reducing the online computational burden.

[0050] In one feasible implementation, the operating parameter information of the controllable device includes thermal resistance parameter, heat capacity parameter and energy efficiency ratio parameter, and the operating parameter information is described by a second-order equivalent thermal parameter model; the operating constraints include device output power limit constraints and temperature state constraints, and the temperature state constraints are obtained by transforming the operating parameter information through iterative expansion of the state equation, and the coefficients of the temperature state constraints include the thermal resistance parameter and the heat capacity parameter.

[0051] It should be noted that thermal resistance parameters refer to physical quantities in a building envelope that impede heat transfer, used to characterize the ability of walls, doors, windows, and other structures to hinder heat conduction. Heat capacity parameters refer to physical quantities in a building envelope and indoor air that store thermal energy, used to characterize the thermal inertia of the building itself and its interior air. Energy efficiency ratio parameters refer to the ratio of heat transferred per unit of electrical energy consumed by equipment, used to establish the conversion relationship between equipment power and heating or cooling power. A second-order equivalent thermal parameter model is a physical model that uses two heat capacity nodes to describe the dynamic temperature changes of indoor air and the building's internal thermal mass. Compared to a single-order model, it can more accurately characterize the heat storage effect of the building envelope and the dynamic coupling relationship between indoor temperature and equipment power. Equipment output power limit constraints mean that the actual output power of the equipment during operation must not exceed the upper and lower limits of the equipment's rated capacity. Temperature state constraints are the mathematical expression of the physical constraint that the indoor temperature must be maintained within the user-set comfort temperature range during equipment operation. The coefficient matrix of this constraint includes thermal resistance and heat capacity parameters, reflecting the dynamic characteristics of the impact of equipment power changes on indoor temperature. Iterative expansion of the state equation refers to the process of successively substituting the difference equation describing the thermal dynamics of the equipment into the equation in chronological order, thereby expressing the indoor temperature at future moments as a linear combination of historical power sequences and constant terms contributing to the initial state and environmental parameters.

[0052] Understandably, this implementation uses a second-order equivalent thermal parameter model to describe the operating parameters of controllable equipment. Compared to simplified descriptions such as virtual battery models used in related technologies, the second-order equivalent thermal parameter model can more accurately characterize the thermal storage effect of the building envelope and the dynamic coupling relationship between indoor temperature and equipment power. Based on this, the temperature state constraint is transformed into a linear inequality concerning the power sequence through iterative expansion of the state equation, allowing complex physical constraints containing continuous thermodynamic parameters to be incorporated into a unified linear constraint framework. Furthermore, since the coefficient matrix of the temperature state constraint explicitly includes thermal resistance and heat capacity parameters, the individual thermal dynamic differences of different devices can be accurately reflected during the feasible region construction stage, providing an accurate constraint boundary description for subsequent high-precision internal approximation processing and avoiding feasible region deviations caused by model simplification.

[0053] In one feasible implementation, the step of calculating the constraint boundary corresponding to each of the deviation state values ​​includes: For each of the aforementioned deviation state values, corresponding constraint boundaries are generated. Each of the aforementioned constraint boundaries is represented in the form of a system of linear inequalities. The coefficient matrix of each of the aforementioned linear inequalities is calculated by combining the aforementioned deviation state values ​​with the operating parameter information. The constant vector of each of the aforementioned linear inequalities is calculated by combining the aforementioned deviation state values ​​with the initial state of the controllable device.

[0054] It should be noted that the system of linear inequalities refers to a set of multiple linear inequalities, which together enclose the feasible region of the equipment in the power variable space; the coefficient matrix is ​​the set of coefficients corresponding to each decision variable in the system of linear inequalities. Each row of this matrix corresponds to a linear inequality, and each column corresponds to a decision variable. The values ​​of each element in the matrix are calculated by combining the deviation state values ​​with the operating parameter information; the constant vector is the set of constants on the right side of each inequality in the system of linear inequalities. Each element of this vector corresponds to the same row of the coefficient matrix. The constant values ​​are calculated by combining the deviation state values ​​with the initial state of the controllable equipment; the initial state refers to the value of the state variables of the controllable equipment at the beginning of the current scheduling cycle, including but not limited to the initial value of the indoor air temperature and the initial value of the internal thermal mass temperature.

[0055] Understandably, this implementation method unifies the physical constraints under various deviation state values ​​into a mathematical expression of a system of linear inequalities, enabling complex constraints containing continuous thermodynamic parameters to be processed within a unified linear framework. Since the coefficient matrix and constant vector are calculated based on operating parameter information and initial states, respectively, the individual physical differences of different devices and the environmental differences under different deviation states are accurately reflected in the constraint expression. Simultaneously, the system of linear inequalities serves as the standard input format for subsequent intersection and inner approximation processing, allowing direct use by the constraint boundary intersection operation. This ensures that data transfer between steps in the entire aggregation process does not require format conversion, improving computational efficiency.

[0056] In one feasible implementation, the step of performing inner approximation processing on each of the candidate operation interval sets to generate an inner approximate polyhedron that remains parallel to the boundary hyperplane of the candidate operation interval sets includes: Obtain the normal vector of each constraint hyperplane in the candidate operation interval set; Based on each of the normal vectors, a generator set parallel to the corresponding constraint hyperplane is constructed, wherein each generator in the generator set is in a parallel direction to the constraint hyperplane; Construct the inner approximate polyhedron based on the generator set.

[0057] It should be noted that a constraint hyperplane refers to the planar boundary in the power variable space corresponding to each linear constraint that encloses the set of candidate operation intervals. Each linear inequality corresponds to a constraint hyperplane. A normal vector is a vector perpendicular to the constraint hyperplane and pointing into the region bounded by the constraint inequality. It represents the direction of the constraint hyperplane in the power variable space, and its components are composed of the coefficients of the decision variables in the corresponding linear inequality. A generator set is a collection of multiple generators. A generator is the edge direction vector of an inner approximate polyhedron. Each generator in the generator set is parallel to its corresponding constraint hyperplane. Parallel direction means that the angle between the generator's direction vector and the constraint hyperplane's direction vector is zero; that is, the generator's direction completely conforms to the direction of the constraint hyperplane, thus ensuring that the faces of the inner approximate polyhedron constructed based on this generator set remain parallel to the original constraint hyperplane.

[0058] It is understandable that this implementation obtains the normal vectors of each constraint hyperplane in the candidate operation interval set and constructs a generator set parallel to the corresponding constraint hyperplane based on each normal vector. Essentially, this aligns the edge directions of the inner approximate polyhedron perfectly with the constraint boundary directions of the original feasible region. Since each generator in the generator set is parallel to the constraint hyperplane, the faces of the inner approximate polyhedron spanned by these generators can form complete surface contact with the corresponding constraint hyperplane, unlike the degenerate point contact formed in conventional techniques due to the inconsistency between the generator direction and the constraint boundary direction. This surface contact method allows the inner approximate polyhedron to approximate the boundary of the original candidate operation interval set to the maximum extent while strictly satisfying the inner approximation constraints, significantly reducing the volume loss caused by mismatched approximate directions and improving the utilization of flexibility resources.

[0059] In one feasible implementation, the step of constructing a generator set parallel to the corresponding constraint hyperplane based on each of the said normal vectors includes: Determine the number of dimensions in the decision variable space of the controllable device; For any of the constrained hyperplanes, the total number of generators is calculated by subtracting one from the dimension number. Each generator corresponding to the constrained hyperplane is constructed, and the position of the non-zero component in each generator is determined alternately based on the adjacent components of the normal vector. Each of the generators is used as the generator set corresponding to the constrained hyperplane.

[0060] It should be noted that the decision variable space refers to the mathematical space spanned by all the power variables to be decided by the controllable equipment. The dimension of this space is equal to the total number of power variables to be decided within the scheduling period. For example, when the scheduling period contains multiple time periods, the power values ​​of each time period constitute one dimension of the decision variable space. The total number of generators refers to the number of generators that need to be constructed for a constrained hyperplane. Its value is equal to the dimension of the decision variable space minus one, so that the constructed generators can span a complete generator subspace parallel to the constrained hyperplane. Non-zero components refer to the components in the generator vector whose values ​​are not zero. The position of the non-zero components in each generator is determined alternately by the adjacent components of the normal vector. Specifically, for the i-th component and the (i+1)-th component of the normal vector, the value of the i-th component is assigned to the (i+1)-th position of the generator, and the value of the negative (i+1)-th component is assigned to the i-th position of the generator. By traversing the combinations of each adjacent component of the normal vector, a generator with a dimension minus one is constructed. The methods for determining the values ​​of adjacent components alternately include, but are not limited to, combining them sequentially according to the index order of two adjacent components in the normal vector, and in each combination, the positions of the two non-zero components in the generator are interchanged relative to their positions in the normal vector.

[0061] Understandably, this implementation constructs each generator by subtracting one from the dimension of the decision variable space, ensuring that the generator set completely covers all degrees of freedom in the direction of the constraint hyperplane. Since the positions of non-zero components in each generator are determined alternately by adjacent components of the normal vector, each constructed generator is strictly parallel to the direction of the constraint hyperplane, and the generators are linearly independent, collectively spanning a directional subspace of the same dimension as the constraint hyperplane. The inner approximate polyhedron constructed based on this generator set has its face orientation perfectly aligned with the constraint hyperplane, achieving the maximum approximate volume while remaining strictly within the candidate operation interval set. Through this method, this implementation, while ensuring strict inner approximation, can construct a complete generator set parallel to any constraint hyperplane, applicable to feasible region constraints with arbitrary continuous coefficients and arbitrary dimensions, thus overcoming the limitation of related technologies where generator construction methods are only applicable to simple standardized constraints.

[0062] In one feasible implementation, the step of summing and aggregating the internal approximate polyhedra of each controllable device includes: Obtain the center point, generator matrix, and upper limit of the scaling factor for each of the aforementioned internal approximate polyhedra; The center points are merged into an aggregate center point, the generator matrices are merged into an aggregate generator matrix, and the upper limits of the scaling coefficients are merged into an aggregate scaling coefficient upper limit. The feasible region for aggregation is generated based on the aggregation center point, the aggregation generator matrix, and the upper limit of the aggregation scaling factor.

[0063] It should be noted that the inner approximation polyhedron refers to the polyhedron used for internal approximation of the candidate operation interval set, and is determined by three parameters: the center point, the generator matrix, and the upper limit of the scaling factor. The center point refers to the geometric center coordinates of the inner approximation polyhedron, used to characterize its position reference in the decision variable space. The generator matrix is ​​a matrix composed of multiple generators arranged as column vectors, where each column vector represents the edge direction of the inner approximation polyhedron in various directions. The number of columns in the generator matrix equals the total number of generators, and the number of rows equals the dimension of the decision variable space. The upper limit of the scaling factor is a vector used to limit the scaling amplitude of each generator. Each element of this vector corresponds to each column of the generator matrix, controlling the maximum scaling amount in each generator direction. By adjusting the value of the upper limit of the scaling factor, the inner approximation polyhedron achieves the maximum approximate volume while ensuring it is completely contained within the candidate operation interval set. Merging refers to the operation of combining the corresponding parameters of the inner approximation polyhedra of each controllable device according to specific rules to form the overall parameters of the cluster. The aggregation center point refers to the merged result obtained by adding the center points of each individual unit according to their corresponding dimensional coordinates, and serves as the center position of the cluster's aggregation feasible region. The aggregation generator matrix is ​​the merged result obtained by sequentially concatenating the individual unit generator matrices along their column directions; its number of columns is the sum of the number of columns in each individual unit generator matrix, while the number of rows remains unchanged. The aggregation scaling factor upper limit is the merged result obtained by sequentially concatenating the scaling factor upper limits of each individual unit along their row directions; its dimension is equal to the sum of the dimensions of the scaling factor upper limits of each individual unit. The aggregation feasible region refers to the flexible and adjustable range of the cluster as a whole, jointly determined by the aggregation center point, the aggregation generator matrix, and the aggregation scaling factor upper limit, used as an overall flexibility model that can be directly invoked by the power grid dispatch center.

[0064] It is understandable that this implementation method obtains the center point, generator matrix, and upper limit of scaling factor of each internal approximate polyhedron, and merges the three types of parameters corresponding to each individual unit to generate an aggregated feasible region. Essentially, it utilizes the closure property of the internal approximate polyhedron under summation aggregation operations to transform the aggregation problem of each individual unit's feasible region into three independent merging operations at the parameter level. Since merging the center points only involves vector addition of corresponding dimension coordinates, merging the generator matrices only involves matrix concatenation in the column direction, and merging the upper limit of scaling factor only involves vector concatenation in the row direction, the computational cost of the aggregation process is linearly related to the number of devices, overcoming the curse of dimensionality problem in the accurate aggregation of high-dimensional complex polyhedra. Simultaneously, since the merging operations of the three types of parameters are independent and do not interfere with each other, the flexibility information of each device in its respective degree of freedom direction is preserved during the aggregation process without loss or distortion, ensuring the accuracy of the aggregation result. Through the above method, this implementation method can quickly complete the aggregation of large-scale heterogeneous temperature-controlled load clusters and has the engineering applicability of online rapid re-aggregation.

[0065] In one feasible implementation, the steps of merging the central points into an aggregate central point, merging the generator matrices into an aggregate generator matrix, and merging the upper limits of the scaling factors into an aggregate scaling factor upper limit include: The corresponding dimensional coordinates of each of the aforementioned center points are added together, and the resulting coordinates are used as the coordinates of the aggregate center point. The generator matrices are concatenated sequentially along the column direction, and the resulting matrix is ​​used as the aggregate generator matrix. The upper limits of each of the above-mentioned scaling coefficients are concatenated sequentially in the row direction, and the resulting vector is used as the upper limit of the aggregated scaling coefficient.

[0066] It should be noted that the center point refers to the geometric center coordinates of the inner approximate polyhedron. The dimension of this coordinate is the same as the dimension of the decision variable space. The coordinate values ​​of each dimension correspond to the center position of the power variable in each time period within the scheduling cycle. The corresponding dimension coordinates of the center point refer to the coordinate values ​​of the center points of the inner approximate polyhedrons of each controllable device in the same decision variable dimension. When these coordinate values ​​are added, the coordinates of each device in the same dimension are scalarly added. The generator matrix is ​​a matrix composed of multiple generators arranged as column vectors. The number of rows in the matrix is ​​equal to the dimension of the decision variable space, and the number of columns is equal to the total number of generators. Sequential concatenation in the column direction means expanding multiple generator matrices horizontally. That is, all columns of the first matrix are arranged in their original order, and all columns of the second matrix are concatenated to its rightmost position, and so on. The number of rows in the resulting matrix remains unchanged, and the number of columns is the sum of the number of columns of each matrix. The upper limit of the scaling factor is a vector used to limit the scaling range of each column vector in the generator matrix. Its dimension is equal to the number of columns of the generator matrix. Each element and the corresponding column of the generator matrix jointly determine the maximum scaling amount in that direction. Sequential concatenation along the row direction refers to expanding multiple vectors with upper limits of scaling coefficients in the vertical direction. That is, the elements of the first vector are kept in their original order, and the elements of the second vector are concatenated to the end of the first vector, and so on. The dimension of the concatenated vector is the sum of the dimensions of the individual vectors.

[0067] Understandably, this implementation method achieves the aggregated feasible region by summing the corresponding dimensional coordinates of each center point, concatenating the generator matrices along the column direction, and concatenating the upper limits of each scaling factor along the row direction. This results in the aggregated feasible region being an exact summation of the feasible regions of each individual unit. Since the summation of center point coordinates geometrically corresponds to the translation and superposition of the center points of each individual unit, the column concatenation of the generator matrices corresponds to the summation of all degrees of freedom of each individual unit, and the row concatenation of the upper limits of the scaling factors corresponds to the summation of the maximum scaling amplitude in each direction, these three concatenation operations completely preserve the independent degrees of freedom information of each individual unit's feasible region at the parameter level. This prevents information from being mixed or lost due to the aggregation operation, thus ensuring that the aggregated feasible region is as accurate as the exact summation of the feasible regions of each individual unit.

[0068] In one feasible implementation, after the step of performing inner approximation processing on each of the candidate operation interval sets, the method further includes: Obtain the coordinates of all vertices of each of the aforementioned internal approximate polyhedra; Determine whether the coordinates of each vertex satisfy the operation constraints in the corresponding candidate operation interval set; When all vertex coordinates satisfy the aforementioned operational constraints, it is determined that all points of the inner approximate polyhedron are within the corresponding candidate operation interval set.

[0069] It should be noted that the coordinates of all vertices of the inner approximate polyhedron refer to the position vectors of each vertex in the decision variable space. Since the inner approximate polyhedron is a convex polyhedron determined by the center point, the generator matrix, and the upper limit of the scaling factor, its number of vertices is determined by the number of generators and the dimension. The coordinates of each vertex can be calculated by combining the coordinates of the center point with the columns of the generator matrix according to the upper limit of the scaling factor. The operational constraints in the candidate operation interval set refer to all the linear inequality constraints used when constructing the candidate operation interval set, including the equipment output power limit constraint and the temperature state constraint. These constraints together define the feasible power region of the controllable equipment under a single deviation state. All points being within the corresponding candidate operation interval set means that the inner approximate polyhedron as a whole is contained within the candidate operation interval set. Since the inner approximate polyhedron is a convex polyhedron, when all vertices of the convex polyhedron satisfy the linear constraints in the candidate operation interval set, based on the convex combination property of the convex polyhedron, any point on the convex polyhedron can be represented as a convex combination of vertices. Therefore, the entire convex polyhedron is located within the candidate operation interval set.

[0070] It is understandable that this implementation obtains the coordinates of all vertices of the inner approximate polyhedron and determines whether each vertex satisfies the operational constraints in the candidate operation interval set. The satisfaction of all vertices is used as a sufficient condition for determining that the entire inner approximate polyhedron is within the candidate operation interval set. Essentially, it utilizes the convex combination relationship between the vertices and the entire polyhedron to transform the determination of the inclusion relationship of infinite points into a constraint satisfaction check of finite vertices. Since practical aggregation applications require the aggregated feasible region to be strictly within the true feasible regions of each individual unit to ensure the safety of scheduling instructions, this implementation verifies the inclusion relationship of the inner approximate polyhedron through a vertex verification mechanism. This ensures that the inner approximation process does not introduce any infeasible power points located outside the candidate operation interval set, thereby guaranteeing the safety of the aggregated feasible region. Furthermore, vertex verification involves only a finite number of vertex coordinate calculations and linear inequality checks, making the computational load controllable and not significantly affecting the overall efficiency of the inner approximation process.

[0071] In one feasible implementation, before the step of obtaining the operating parameter information of each controllable device to be aggregated and the uncertainty information of the external environment in which the controllable devices are located, the method further includes: Obtain a pre-configured list of device clusters, which contains the device identifiers of each controllable device to be aggregated; Based on the device identifier, the inherent parameters of each controllable device are read as the operating parameter information, and the statistical characteristic values ​​of the historical measurement data of the environmental sensors at the location of each controllable device are read as the uncertainty information.

[0072] It should be noted that the device cluster list refers to a pre-configured list of devices to be aggregated, used to clearly define the scope of each controllable device involved in the current aggregation operation. The device cluster list includes the device identifier for each controllable device; this identifier is a unique code used to identify each controllable device. Device inherent parameters refer to the physical characteristic parameters of the controllable device determined at the factory, including but not limited to thermal resistance, heat capacity, and energy efficiency ratio. These parameters do not change with the device's operating status or external environment. Environmental sensors refer to sensing devices deployed at the location of each controllable device for real-time monitoring of external environmental parameters, including but not limited to temperature and humidity sensors. Historical measurement data refers to the sequence of measurement values ​​collected and recorded by environmental sensors over a past period, used to characterize the historical changes in the external environment at that location. Statistical characteristic values ​​refer to quantitative indicators obtained after statistical analysis of historical measurement data, including but not limited to the mean and standard deviation of historical measurement data.

[0073] Understandably, this implementation pre-configures a device cluster list before acquiring operational parameter information and uncertainty information, and reads the inherent parameters of each controllable device and historical measurement data from environmental sensors accordingly. The inherent parameters of the devices are used as operational parameter information, and the statistical characteristic values ​​of the historical measurement data are used as uncertainty information. This achieves the conversion of multi-source data into a unified information format, allowing subsequent steps to directly read the required information from the configured data structure. This eliminates the need to repeatedly perform device discovery and data acquisition operations during each aggregation process, reducing the online computational burden. Furthermore, since statistical characteristic values ​​reflect the historical fluctuation patterns of environmental parameters at the location of each controllable device, the external environmental deviation state values ​​generated based on these statistical characteristic values ​​more accurately characterize the uncertainty distribution in the actual operating scenario compared to directly using a single predicted value, thus improving the rationality of subsequent feasible domain construction.

[0074] For example, to help understand the implementation flow of the feasible domain aggregation method obtained by combining the above embodiments, please refer to... Figure 2 , Figure 2 A simplified flowchart of a feasible region aggregation method is provided, specifically: Step S101: Construct an accurate linear feasible region for a single temperature-controlled load using a second-order equivalent thermal parameter model, and construct a robust feasible region considering uncertainties by using the intersection of constraints from multiple outdoor temperature scenarios. Step S101 includes: for a single temperature-controlled load in a winter heating scenario, a second-order equivalent thermal parameter model is used to describe the thermal dynamics between its building walls, indoor air, and thermal mass, and the discretized state equation is expressed as: in, for The state vector at the end of the time period, where the superscript T indicates the transpose of a vector or matrix; This is the state vector at the end of time interval t-1; Indoor air temperature (°C); Internal thermal mass temperature (°C); for Time period heat pump input power (kW); for Outdoor temperature (°C) during the time period; Here is the discretized coefficient matrix, where This is the state transition matrix, used to describe the influence of the previous time period's state on the current time period's state; This is the power input matrix, used to describe the impact of the electrical power in the previous time period on the state in the current time period; The outdoor temperature input matrix describes the impact of the outdoor temperature from the previous time period on the current time period's conditions. This matrix is ​​derived from the building's thermal resistance. Heat capacity Heat pump energy efficiency ratio and the length of the scheduling period It was calculated using the Euler method.

[0075] The equipment must meet the constraints of upper and lower power limits and the user's temperature comfort range during operation. in, for Time period heat pump input power (kW); Rated thermal power (kW) for the heat pump; These are the upper and lower limits (°C) of the user-set comfortable temperature.

[0076] The indoor temperature can be determined using the discretized state equations described above. Represented as decision variables Linear combination: Wherein, the subscript T represents the total number of time periods included in the scheduling period, while the superscript T represents the vector transpose; for right The influence coefficient; From the initial state The constant term is determined by the outdoor temperature sequence. Substituting the above relationship into the temperature control constraint, the flexibility feasible region of a single temperature control load can be uniformly described as relating to... The system of linear inequalities: in, It is a constraint matrix that includes power constraints and temperature constraint coefficients after conversion; The constraint constant vector for dimension matching.

[0077] To mitigate the uncertainty in outdoor temperature forecasting, a scenario-based approach is employed to construct a robust feasible region. Assume the outdoor temperature... The predicted mean is The standard deviation is ,generate One possible scenario: in, This represents the outdoor temperature value at time t in the s-th scenario; This is a random perturbation factor that follows a standard normal distribution. For each scenario... Calculate the corresponding constraint set . Let be the constraint coefficient matrix corresponding to the s-th outdoor temperature scenario. This represents the constraint constant vector for this scenario. Robust feasible region. Defined as the intersection of constraints across all scenarios: This intersection is a conservative subset of the original deterministic feasible region, which can guarantee that the scheduling solution strictly satisfies the physical constraints under any considered temperature scenario.

[0078] Step S102: By constructing a generator set parallel to the constraint hyperplane, a high-precision inner approximation is performed on the robust feasible region to obtain a single temperature-controlled load's Chino polyhedron approximation model. Step S102 includes: to achieve efficient aggregation, using the Kino polyhedron to robust feasible regions. Perform an inner approximation. One dimension( Kino polyhedron Z The parameterization is defined as follows: Where N is the number of dimensions in the decision variable space, and here N = T, that is, the number of dimensions is equal to the total number of scheduling periods T; Center point; Let be the generator matrix, where Let G be the first generation of the generator matrix. Column, i.e., the first One generator; Let be the total number of generators, and ; The vector of scaling coefficients; This is the upper limit vector of the scaling factor.

[0079] The goal of inner approximation is to maximize the approximate similarity. That is, the average of the ratios of the diameters of the original polyhedron and the Chino polyhedron in random directions, while simultaneously satisfying strict inner approximation constraints. This constraint is equivalent to: in, This represents taking the absolute value of the matrix elements. Traditional methods suffer from significant volume loss because the generator cannot be parallel to the constraint hyperplane. This invention proposes: for the constraint matrix... Hyperplane constraints represented by any row ,in For the coefficient vector, For the decision variable vector, For the constant scalar corresponding to the constraint, the superscript T denotes the vector transpose, and the entire expression... Let represent a linear inequality constraint, corresponding to a hyperplane in the decision variable space. Construct a hyperplane parallel to this hyperplane. The generator, whose first generator The generator takes the form of: in, The Each component is , No. Each component is The remaining components are zero. This construction allows the Chino polyhedron to form complete surface contact rather than degenerate point contact in the corresponding constraint direction. That is, one face of the Chino polyhedron is parallel to and closely fitted with the constraint hyperplane of the original feasible region, thereby maximizing the approximate volume and significantly reducing conservatism while strictly satisfying the inner approximation constraint.

[0080] Step S103: Based on the Minkowski and operational closure properties of the Chino polyhedron, the Chino polyhedron approximation models of each individual temperature-controlled load are efficiently aggregated to obtain the aggregated feasible region model of the temperature-controlled load cluster.

[0081] For including A resource cluster with temperature-controlled loads, whose aggregated feasible region is the Minkowski sum of the feasible regions of each individual unit: Since the feasible region of each individual is approximately a Chino polyhedron The aggregation process can be directly completed by simple splicing and addition of the parameters of the Chino polyhedron, reducing the computational complexity from NP-hard to zero. : in, This represents the total number of devices included in the temperature-controlled load cluster. For the j-th device, an approximate model of the Chino polyhedron; The center point of the aggregated Chino polyhedron; This is the aggregated generator matrix; This is the aggregated scaling coefficient vector; is the upper bound vector of the aggregation scaling coefficients; ⊕ represents the Minkowski sum operation.

[0082] The aggregated parameters are calculated using the following formula: in, The first The center point of the Chino polyhedron corresponding to each temperature-controlled load, the generator matrix, and the upper limit of the scaling factor; where... This indicates that the generator matrices are horizontally concatenated along the column direction; This indicates that the upper bound vectors of each scaling factor are vertically concatenated along the row direction. (Aggregation model) That is, the compact mathematical description of the overall flexible and adjustable range of the temperature-controlled load cluster can be directly embedded as a linear constraint into the power grid optimization scheduling model.

[0083] For example, please refer to Figure 3 , Figure 3 This is a schematic diagram illustrating the construction and internal approximation principle of the improved Zonotope generator in the embodiments of this application. Figure 3 With a two-dimensional decision variable space (in) and Using the coordinate axis as an example, the core principle of approximation processing in this application is intuitively demonstrated.

[0084] In the diagram, This represents the set of candidate operation intervals constructed based on uncertainty information, i.e., the original robust feasible region. This represents the improved Cino polyhedron generated by approximating the original polyhedron in this application. Wherein, The chino polyhedron generator vector, which is parallel to the original feasible region boundary hyperplane, is a key element for achieving high-precision inner approximation.

[0085] The key point is that the generator construction method proposed in this application is no longer limited to the traditional limitation that generators can only be parallel to the coordinate axes. For example... Figure 3 As shown, the generated generator vector is constructed to be the same as the original robust feasible region. The constrained hyperplanes are parallel. This surface-contact construction method results in the generation of the Chino polyhedron. The side surface can be compared with the original polyhedron The boundaries fit together tightly.

[0086] Figure 3 The marked and These represent the original polyhedra in a specific direction or on a constrained surface. With the generated inner approximate polyhedron The difference between the diameter or the support function. The generator construction rules proposed in this application can minimize this error, making the inner approximate polyhedron... Under the guarantee Under the premise of (i.e., strict inner approximation, without generating practically infeasible power points), the geometric features and flexible volume of the original feasible region are preserved to the greatest extent, and the volume loss caused by geometric degradation is significantly reduced compared with traditional generator methods.

[0087] For example, please refer to Figure 4 , Figure 4 This is a comparison diagram of the robust feasible region of a single unit and its approximation within the Cino polyhedron in the embodiments of this application. The diagram illustrates the feasible region state of a single temperature-controlled load on a two-dimensional power plane. The solid purple line and its inner light purple shaded area represent the original feasible region, i.e., the accurate robust linear feasible region constructed based on the intersection of the second-order equivalent thermal parameter (ETP) model and constraints for each temperature scenario. The dashed orange line and light orange area represent the deterministic Cino polyhedron, i.e., the approximate result within the Cino polyhedron constructed using traditional deterministic environmental parameters; while the solid green line and light green area represent the robust Cino polyhedron constructed by the method of this application. It can be clearly observed that the robust Cino polyhedron (green area) is completely enclosed within the original feasible region (purple area), indicating that it strictly satisfies the physical constraints under all possible outdoor temperature deviation scenarios, such as the indoor temperature comfort range, thereby achieving robust resistance to uncertainty. Meanwhile, compared to the deterministic Cino polyhedron, the robust Cino polyhedron constructed in this application, although considering a certain degree of conservatism to cope with environmental deviations, has a boundary that fits the original polyhedron boundary more closely, proving that the method in this application can effectively reduce the loss of flexibility caused by excessive conservatism in traditional robust methods while ensuring scheduling safety.

[0088] For example, please refer to Figure 5 , Figure 5This is a comparison chart of the Minkowski similarity and aggregation similarity of the cluster aggregation feasible domains in the embodiments of this application. The chart illustrates the performance of the method in this application when applied to large-scale device cluster aggregation. The horizontal axis represents the number of scheduling periods (from 1 to 25 periods), and the vertical axis represents the similarity of the feasible domains of individual resources, reflecting the degree to which the aggregated feasible domains retain the geometric features of the individual real feasible domains. The chart contains two main curves and their corresponding... Error bands (shaded areas). Orange circles represent deterministic similarity and their confidence intervals, while green squares represent robust similarity and their confidence intervals. Observation Figure 5 It can be seen that as the number of scheduling periods increases, the similarity of the aggregations all show a slow decreasing trend, which is consistent with the geometric characteristics of high-dimensional polyhedral aggregation. Comparing the two curves, it can be found that although the similarity of the robust aggregation method (green line) in this application is slightly lower than that of the deterministic method (orange line), it is due to the introduction of environmental prediction errors. The deviation band ensures, in engineering terms, that the aggregation results strictly meet the physical constraints of each individual device even in scenarios where there are deviations in outdoor temperature predictions, thereby avoiding the risk of scheduling instructions exceeding limits that may be caused by deterministic models.

[0089] More importantly, Figure 5 The actual data verified that the robust aggregation framework based on the improved Zonotope proposed in this application can maintain high geometric fidelity even in high-dimensional spaces with dozens of scheduling periods. Moreover, the aggregation process is based on the linear operation characteristics of Minkowski sums, which can achieve fast aggregation at the millisecond level, overcoming the curse of dimensionality problem faced by traditional precise polyhedral aggregation.

[0090] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the feasible domain aggregation method of this application. Any simple transformations based on this technical concept are within the protection scope of this application.

[0091] This application also provides a feasible domain aggregation apparatus, please refer to... Figure 6 The feasible domain aggregation device includes: The acquisition module 10 is used to acquire the operating parameter information of each controllable device to be aggregated and the uncertainty information of the external environment in which the controllable devices are located; The construction module 20 is used to construct a candidate operation interval set for each controllable device based on the operating parameter information and the uncertainty information. Each element in the candidate operation interval set corresponds to an operation constraint under a deviation state of the external environment. The generation module 30 is used to perform inner approximation processing on each of the candidate operation interval sets to generate an inner approximate polyhedron that is parallel to the boundary hyperplane of the candidate operation interval set. The aggregation module 40 is used to sum and aggregate the internal approximate polyhedra of each controllable device to obtain the overall aggregated feasible region of the controllable device cluster.

[0092] The feasible domain aggregation device provided in this application, employing the feasible domain aggregation method in the above embodiments, can solve the technical problem of poor robustness in feasible domain aggregation of temperature-controlled loads in real-world scenarios. Compared with the prior art, the beneficial effects of the feasible domain aggregation device provided in this application are the same as those of the feasible domain aggregation method provided in the above embodiments, and other technical features in the feasible domain aggregation device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0093] This application provides a feasible domain aggregation device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the feasible domain aggregation method in Embodiment 1 above.

[0094] The following is for reference. Figure 7 The diagram illustrates a structural schematic of a feasible domain aggregation device suitable for implementing embodiments of this application. The feasible domain aggregation device in the embodiments of this application may include, but is not limited to, mobile terminals such as mobile phones, tablets, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital televisions and desktop computers. Figure 7 The feasible domain aggregation device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0095] like Figure 7As shown, the feasible domain aggregation device may include a processing unit 1001, which can perform various appropriate actions and processes according to a program stored in ROM 1002 or a program loaded from storage device 1003 into RAM 1004. RAM 1004 also stores various programs and data required for the operation of the feasible domain aggregation device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via bus 1005. I / O interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touch screens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the feasible domain aggregation device to exchange data with other devices wirelessly or via wired communication. While the figures illustrate feasible domain aggregation devices with various systems, it should be understood that implementation or having all of the systems shown is not required. More or fewer systems may be implemented alternatively.

[0096] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0097] The feasible domain aggregation device provided in this application, employing the feasible domain aggregation method in the above embodiments, can solve the technical problem of poor robustness in feasible domain aggregation of temperature-controlled loads in real-world scenarios. Compared with the prior art, the beneficial effects of the feasible domain aggregation device provided in this application are the same as those of the feasible domain aggregation method provided in the above embodiments, and other technical features in this feasible domain aggregation device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0098] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0099] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0100] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the feasible domain aggregation method in the above embodiments.

[0101] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0102] The aforementioned computer-readable storage medium may be included in the feasible domain aggregation device; or it may exist independently and not assembled into the feasible domain aggregation device.

[0103] The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the feasible domain aggregation device, cause the feasible domain aggregation device to: Obtain the operating parameter information of each controllable device to be aggregated, as well as the uncertainty information of the external environment in which the controllable devices are located; Based on the operating parameter information and the uncertainty information, a candidate operation interval set is constructed for each controllable device. Each element in the candidate operation interval set corresponds to an operation constraint under a deviation state of the external environment. The candidate operation interval sets are subjected to internal approximation processing to generate an internal approximate polyhedron that is parallel to the boundary hyperplane of the candidate operation interval sets. The approximate polyhedra of each controllable device are summed and aggregated to obtain the aggregated feasible region of the entire controllable device cluster.

[0104] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0105] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0106] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0107] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described feasible region aggregation method, which can solve the technical problem of poor robustness in feasible region aggregation of temperature-controlled loads in real-world scenarios. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as the beneficial effects of the feasible region aggregation method provided in the above embodiments, and will not be repeated here.

[0108] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the feasible domain aggregation method described above.

[0109] The computer program product provided in this application can solve the technical problem of poor robustness in feasible region aggregation of temperature-controlled loads in real-world scenarios. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the feasible region aggregation method provided in the above embodiments, and will not be repeated here.

[0110] All acquisition of signals, information, or actions in this application are carried out in compliance with the relevant data protection laws and policies of the country where the application is located, and with the authorization of the relevant device owner.

[0111] The above description is only a part of the embodiments of this application and does not limit the scope of protection of this application. All equivalent structural transformations made under the technical concept of this application and using the content of this application specification and drawings, or direct / indirect applications in other related technical fields, are included in the scope of protection of this application.

Claims

1. A feasible domain aggregation method, characterized in that, The method includes: Obtain the operating parameter information of each controllable device to be aggregated, as well as the uncertainty information of the external environment in which the controllable devices are located; Based on the operating parameter information and the uncertainty information, a candidate operation interval set is constructed for each controllable device. Each element in the candidate operation interval set corresponds to an operation constraint under a deviation state of the external environment. The candidate operation interval sets are subjected to internal approximation processing to generate an internal approximate polyhedron that is parallel to the boundary hyperplane of the candidate operation interval sets. The approximate polyhedra of each controllable device are summed and aggregated to obtain the aggregated feasible region of the entire controllable device cluster.

2. The method as described in claim 1, characterized in that, The step of constructing a candidate operating range set for each controllable device based on the operating parameter information and the uncertainty information includes: Based on the operating parameter information and the uncertainty information, calculate the deviation state values ​​of the external environment; Calculate the constraint boundaries corresponding to each of the aforementioned deviation state values; The intersection of the constraint boundaries corresponding to each deviation state value is used as the set of candidate operation intervals.

3. The method as described in claim 2, characterized in that, The step of calculating the deviation state values ​​of the external environment based on the operating parameter information and the uncertainty information includes: Read the pre-configured baseline state value and the pre-configured deviation distribution information, wherein the deviation distribution information includes the magnitude of each deviation and its corresponding probability of occurrence; The deviation state values ​​of the external environment are generated based on the baseline state value and the deviation distribution information.

4. The method as described in claim 1, characterized in that, The operating parameter information of the controllable device includes thermal resistance parameter, heat capacity parameter and energy efficiency ratio parameter. The operating parameter information is described by a second-order equivalent thermal parameter model. The operating constraints include device output power limit constraints and temperature state constraints. The temperature state constraints are obtained by iteratively expanding the state equations from the operating parameter information. The coefficients of the temperature state constraints include the thermal resistance parameter and the heat capacity parameter.

5. The method as described in claim 2, characterized in that, The step of calculating the constraint boundary corresponding to each of the deviation state values ​​includes: For each of the aforementioned deviation state values, corresponding constraint boundaries are generated. Each of the aforementioned constraint boundaries is represented in the form of a system of linear inequalities. The coefficient matrix of each of the aforementioned linear inequalities is calculated by combining the aforementioned deviation state values ​​with the operating parameter information. The constant vector of each of the aforementioned linear inequalities is calculated by combining the aforementioned deviation state values ​​with the initial state of the controllable device.

6. The method as described in claim 1, characterized in that, The step of performing inner approximation processing on each of the candidate operation interval sets to generate an inner approximate polyhedron that is parallel to the boundary hyperplane of the candidate operation interval set includes: Obtain the normal vector of each constraint hyperplane in the candidate operation interval set; Based on each of the normal vectors, a generator set parallel to the corresponding constraint hyperplane is constructed, wherein each generator in the generator set is in a parallel direction to the constraint hyperplane; Construct the inner approximate polyhedron based on the generator set.

7. The method as described in claim 6, characterized in that, The step of constructing a generator set parallel to the corresponding constrained hyperplane based on each of the normal vectors includes: Determine the number of dimensions in the decision variable space of the controllable device; For any of the constrained hyperplanes, the total number of generators is calculated by subtracting one from the dimension number. Each generator corresponding to the constrained hyperplane is constructed, and the position of the non-zero component in each generator is determined alternately based on the adjacent components of the normal vector. Each of the generators is used as the generator set corresponding to the constrained hyperplane.

8. The method as described in claim 1, characterized in that, The step of summing and aggregating the internal approximate polyhedra of each controllable device includes: Obtain the center point, generator matrix, and upper limit of the scaling factor for each of the aforementioned internal approximate polyhedra; The central points are merged into an aggregate central point, the generator matrices are merged into an aggregate generator matrix, and the upper limits of the scaling coefficients are merged into an aggregate scaling coefficient upper limit. The feasible region for aggregation is generated based on the aggregation center point, the aggregation generator matrix, and the upper limit of the aggregation scaling factor.

9. The method as described in claim 8, characterized in that, The steps of merging the various center points into an aggregate center point, merging the various generator matrices into an aggregate generator matrix, and merging the various scaling factor upper limits into an aggregate scaling factor upper limit include: The corresponding dimensional coordinates of each of the aforementioned center points are added together, and the resulting coordinates are used as the coordinates of the aggregate center point. The generator matrices are concatenated sequentially along the column direction, and the resulting matrix is ​​used as the aggregate generator matrix. The upper limits of each of the above-mentioned scaling coefficients are concatenated sequentially in the row direction, and the resulting vector is used as the upper limit of the aggregated scaling coefficient.

10. The method as described in claim 1, characterized in that, After the step of performing inner approximation processing on each of the candidate operation interval sets, the method further includes: Obtain the coordinates of all vertices of each of the aforementioned internal approximate polyhedra; Determine whether the coordinates of each vertex satisfy the operation constraints in the corresponding candidate operation interval set; When all vertex coordinates satisfy the aforementioned operational constraints, it is determined that all points of the inner approximate polyhedron are within the corresponding candidate operation interval set.

11. The method as described in claim 1, characterized in that, Before the step of obtaining the operating parameter information of each controllable device to be aggregated and the uncertainty information of the external environment in which the controllable devices are located, the method further includes: Obtain a pre-configured list of device clusters, which contains the device identifiers of each controllable device to be aggregated; Based on the device identifier, the inherent parameters of each controllable device are read as the operating parameter information, and the statistical characteristic values ​​of the historical measurement data of the environmental sensors at the location of each controllable device are read as the uncertainty information.

12. A feasible domain aggregation apparatus, characterized in that, The device includes: The acquisition module is used to acquire the operating parameter information of each controllable device to be aggregated, as well as the uncertainty information of the external environment in which the controllable devices are located; The construction module is used to construct a candidate operation interval set for each controllable device based on the operating parameter information and the uncertainty information. Each element in the candidate operation interval set corresponds to the operation constraint under a deviation state of the external environment. The generation module is used to perform inner approximation processing on each of the candidate operation interval sets to generate an inner approximate polyhedron that is parallel to the boundary hyperplane of the candidate operation interval set. The aggregation module is used to sum and aggregate the internal approximate polyhedra of each controllable device to obtain the aggregated feasible region of the entire controllable device cluster.

13. A feasible domain aggregation device, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the feasible domain aggregation method as described in any one of claims 1 to 11.

14. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the feasible domain aggregation method as described in any one of claims 1 to 11.

15. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the feasible domain aggregation method as described in any one of claims 1 to 11.