Building flexible resource aggregation demand response method and system

Abstract

The invention provides a building flexible resource aggregation demand response method and system, and belongs to the technical field of building load management in a power system. Individual and overall flexibility of the constant-temperature control load is defined by constructing a nonlinear switching model of the constant-temperature control load and converting the nonlinear switching model into a linear system model; introducing a virtual battery model to carry out overall flexibility modeling on the constant temperature control load; optimizing parameters of the virtual battery model based on an optimization algorithm; and representing the optimized virtual battery model by adopting a geometric polyhedron set, searching the optimal scaling and translation of the optimal virtual battery model by adopting an optimization algorithm based on linear programming, and dynamically adjusting the response margin of the constant temperature control load TCLs based on the obtained optimal scaling and translation results. According to the method, aggregation demand response coordination among building flexible resources can be realized, and scientific guidance is provided for constructing a more intelligent, efficient and low-carbon building energy system.

Description

Technical Field

[0001] The present invention belongs to the technical field of building load management in power systems, and in particular relates to a method and system for responding to building flexible resource aggregation demand. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] As renewable energy sources such as wind and solar power penetrate deeper into the power system, energy efficiency management of company-owned buildings faces new challenges. Due to the volatility and uncertainty of renewable energy generation, how to maintain the balance between building energy demand and supply becomes more complicated. To meet this challenge, company-owned buildings need to have sufficient energy flexibility to achieve an efficient balance between load demand and renewable energy generation.

[0004] In owned buildings, demand response is a key technology that helps the grid maintain balance by adjusting the operation of electrical equipment. Temperature-controlled loads (TCLs) in buildings, such as air conditioning systems, heat pumps, and water heaters, have become important targets for optimizing energy efficiency management due to their flexibility in electricity consumption. By coordinating these TCLs, owned buildings can not only respond to grid demand and provide auxiliary services such as frequency regulation and peak-shaving, but also better integrate renewable energy to achieve self-regulation and energy efficiency optimization.

[0005] However, due to the large number of TCLs in buildings and their different thermodynamic properties, how to efficiently aggregate these heterogeneous TCLs and establish a concise and accurate model to characterize their overall flexibility has become a key challenge to improve building energy management. Traditionally, statistical methods (such as Markov chains and partial differential equations) are usually used to describe the dynamic behavior of TCLs groups, but this statistical method is only suitable for describing the situation where the structure and equipment in the building are relatively simple; it is not ideal when used to characterize the load characteristics of specific complex buildings, so it is difficult to achieve effective aggregation demand response coordination between building flexible resources. Summary of the invention

[0006] In order to overcome the shortcomings of the above-mentioned prior art, the present invention provides a method and system for responding to aggregated demand of building flexible resources, which can realize the coordination of aggregated demand response among building flexible resources and provide scientific guidance for building a more intelligent, efficient and low-carbon building energy system.

[0007] To achieve the above objectives, one or more embodiments of the present invention provide the following technical solutions:

[0008] A first aspect of the present invention provides a method for responding to building flexible resource aggregation demand.

[0009] A building flexible resource aggregation demand response method, comprising:

[0010] Constructing a nonlinear switching model of a thermostatically controlled load TCLs in a building, and converting the nonlinear switching model into a linear system model according to the nonlinear characteristics of the nonlinear switching model;

[0011] The individual flexibility of the thermostatically controlled loads TCLs is defined based on the linear system model, and the overall flexibility is obtained based on the individual flexibility; the overall flexibility is used to represent the response margin of the thermostatically controlled loads TCLs, and the response margin is introduced into the virtual battery model to model the overall flexibility of the thermostatically controlled loads TCLs; the parameters of the virtual battery model introduced with the response margin are optimized by an optimization algorithm based on geometric information to obtain an optimized virtual battery model;

[0012] The optimized virtual battery model is represented by a set of geometric polyhedrons, and an optimization algorithm based on linear programming is used to find the optimal scaling and optimal translation of the virtual battery model represented by a set of geometric polyhedrons. The response margin of the thermostatically controlled load TCLs is dynamically adjusted based on the obtained optimal scaling and optimal translation to achieve the aggregated demand response coordination of building flexible resources.

[0013] Furthermore, according to the nonlinear characteristics of the nonlinear switching model, the nonlinear switching model is converted into a linear system model, including: obtaining the temperature change of each individual thermostatically controlled load in the thermostatically controlled loads TCLs, and obtaining the nonlinear characteristics of the nonlinear switching model based on the temperature change of each individual thermostatically controlled load in the thermostatically controlled loads TCLs and the constraints of the local control rules of each thermostatically controlled load TCLs; according to the obtained nonlinear characteristics, the nonlinear switching model is converted into a linear system model.

[0014] Further, obtaining the temperature change of each individual thermostatically controlled load in the thermostatically controlled loads TCLs includes: using a discrete time switching model to characterize the temperature change of each individual thermostatically controlled load in the thermostatically controlled loads TCLs.

[0015] Furthermore, individual flexibility of thermostatically controlled loads TCLs is defined based on the linear system model, and overall flexibility is obtained based on the individual flexibility, including: defining the individual flexibility of the thermostatically controlled loads as a set of all allowed power curves; and calculating the overall flexibility of the thermostatically controlled loads TCLs based on the obtained individual flexibility and in combination with the Minkowski sum.

[0016] Furthermore, an optimization algorithm based on geometric information optimizes the parameters of the virtual battery model with the response margin introduced. Specifically: first, the geometric characteristics of the geometric polyhedron set are used as the input of the virtual battery model with the response margin introduced, and the geometric polyhedron set is mapped to the virtual battery model parameters; then, the Minkowski distance is used to optimize the parameters.

[0017] Furthermore, the geometric properties of the set of geometric polyhedra include power limitation, energy storage capacity, and charge and discharge rate.

[0018] Furthermore, an optimization algorithm based on linear programming is used to find the optimal scaling and optimal translation of the virtual battery model represented by a set of geometric polyhedrons, including: finding the optimal solution of the sum of scaling factors and the sum of displacement factors by calculating the homeomorphism of convex sets in the virtual battery model represented by the set of geometric polyhedrons; using the optimal solution of the sum of scaling factors as the optimal scaling of the virtual battery model represented by the set of geometric polyhedrons, and using the optimal solution of the sum of displacement factors as the optimal translation of the virtual battery model represented by the set of geometric polyhedrons.

[0019] A second aspect of the present invention provides a building flexible resource aggregation demand response system.

[0020] A building flexible resource aggregation demand response system, comprising:

[0021] The linear system model conversion module is configured to: construct a nonlinear switching model of the constant temperature control load TCLs in the building, and convert the nonlinear switching model into a linear system model according to the nonlinear characteristics of the nonlinear switching model;

[0022] The thermostatically controlled load flexibility modeling module is configured to: define individual flexibility of thermostatically controlled loads TCLs based on the linear system model, and obtain overall flexibility based on the individual flexibility; use the overall flexibility to represent the response margin of the thermostatically controlled loads TCLs, introduce the response margin into a virtual battery model to model the overall flexibility of the thermostatically controlled loads TCLs; optimize the parameters of the virtual battery model introduced with the response margin using an optimization algorithm based on geometric information to obtain an optimized virtual battery model;

[0023] The thermostatic control load flexibility characterization module is configured as follows: using a set of geometric polyhedrons to represent the optimized virtual battery model, and using an optimization algorithm based on linear programming to find the optimal scaling and optimal translation of the virtual battery model represented by the set of geometric polyhedrons; dynamically adjusting the response margin of the thermostatic control load TCLs based on the obtained optimal scaling and optimal translation to achieve aggregated demand response coordination of building flexible resources.

[0024] A third aspect of the present invention provides a computer-readable storage medium having a program stored thereon, which, when executed by a processor, implements the steps in a building flexible resource aggregation demand response method as described in the first aspect of the present invention.

[0025] The fourth aspect of the present invention provides an electronic device, including a memory, a processor, and a program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in a method for responding to building flexible resource aggregation demand as described in the first aspect of the present invention are implemented.

[0026] One or more of the above technical solutions have the following beneficial effects:

[0027] The present invention provides a method for aggregated demand response of building flexible resources, which is used to characterize the overall flexibility of a group of temperature-controlled loads (TCLs) in owned buildings. The power flexibility of a single TCL is modeled as a polyhedron, and its overall flexibility is represented by the Minkowski sum of these polyhedrons. At the same time, the present invention adopts two optimization algorithms based on linear programming (one is through maximum inner approximation, and the other is through minimum outer approximation of affine transformation relative to the prototype set) to approximate the overall flexibility. Therefore, even in the face of more complex building structures, the present invention can well approximate the flexibility set of polyhedrons, realize the aggregated demand response coordination between building flexible resources, and then provide scientific guidance for building a more intelligent, efficient and low-carbon building energy system.

[0028] The present invention selects a virtual battery model as a prototype. This geometric method can extract more flexibility than the statistically dependent algorithm in the prior art. Therefore, the present invention can further improve the energy efficiency management and regulation capabilities of TCLs in owned buildings, and provide more effective regulation services for the power grid by controlling TCLs.

[0029] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0031] Figure 1 This is a method flow chart of a building flexible resource aggregation demand response method in embodiment 1 of the present invention.

[0032] Figure 2 It is a schematic diagram of a structure representing the optimal approximation of overall flexibility in the first embodiment of the present invention.

[0033] Figure 3 Schematic diagram of the characteristics of a suboptimal sufficient virtual battery in Embodiment 1 of the present invention. DETAILED DESCRIPTION

[0034] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.

[0035] It should be noted that the terms used herein are for describing specific embodiments only and are not intended to be limiting of exemplary embodiments according to the present invention.

[0036] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0037] The general idea proposed by the present invention is that a large number of thermostatically controlled loads (TCLs) in owned buildings have great potential to provide a variety of ancillary services to the power grid. The key challenge of integrating these TCLs into system-level operation and control is how to develop a model that is both simple and applicable to different building scenarios, and then accurately capture their aggregate flexibility. In this regard, the present invention proposes a building flexible resource aggregation demand response method to simulate the aggregate flexibility of TCLs in owned buildings, and proves that the allowable power distribution set of TCLs can be expressed as a polyhedron, and its aggregate flexibility is the Minkowski sum of each polyhedron. In order to more intuitively show the aggregate flexibility of these loads in the building and achieve simplified processing, this paper develops an optimization algorithm based on geometric properties and linear programming to approximate the allotropic polyhedron of a given convex set, including, this set is the virtual battery model; and the best approximation is achieved by solving the equivalent linear programming problem. The method provided by the present invention can not only provide theoretical support for the application of temperature-controlled loads of owned buildings in power grid ancillary services, but also provide an effective tool for improving the level of building energy efficiency management and regulation, so that buildings can play a greater role in low-carbon operation.

[0038] Embodiment 1

[0039] This embodiment discloses a method for responding to building flexible resource aggregation demand.

[0040] like Figure 1 As shown, a building flexible resource aggregation demand response method includes:

[0041] Step S1, constructing a nonlinear switching model of a constant temperature control load TCLs in a building, and converting the nonlinear switching model into a linear system model according to the nonlinear characteristics of the nonlinear switching model;

[0042] Step S2, defining individual flexibility of the thermostatically controlled loads TCLs based on the linear system model, and obtaining overall flexibility based on the individual flexibility; using the overall flexibility to represent the response margin of the thermostatically controlled loads TCLs, introducing the response margin into the virtual battery model to model the overall flexibility of the thermostatically controlled loads TCLs; optimizing the parameters of the virtual battery model introduced with the response margin using an optimization algorithm based on geometric information to obtain an optimized virtual battery model;

[0043] Step S3, using a geometric polyhedron set to represent the optimized virtual battery model, and using a linear programming-based optimization algorithm to find the optimal scaling and optimal translation of the virtual battery model represented by the geometric polyhedron set; based on the obtained optimal scaling and optimal translation, dynamically adjust the response margin of the constant temperature control load TCLs to achieve aggregated demand response coordination of building flexible resources.

[0044] Based on the above process, the present invention can achieve the coordination of aggregated demand response between building flexible resources, and provide scientific guidance for building a more intelligent, efficient and low-carbon building energy system. In order to facilitate the understanding of the technical solution of the present invention, the specific implementation methods in the technical solution of the present invention are further explained and illustrated, including:

[0045] In step S1, a nonlinear switching model of the constant temperature control load TCLs in the building is constructed, and is converted into a linear system model according to the nonlinear characteristics of the nonlinear switching model, which can be achieved by the following steps:

[0046] Step S1-1, the discrete time switching model can discretize the time axis, that is, simplify the description of the temperature change of the individual TCL by staged modeling, thereby reducing the complexity of the model; therefore, this embodiment uses a discrete time switching model to characterize the temperature change of each individual thermostatically controlled load TCL in the thermostatically controlled load TCLs. The discrete time switching model is shown in the following formula:

[0047] θ(k)=aθ(k-1)+(1-a)(θ a -bq(k)P m ) (1)

[0048] Where θ(k) represents the TCL temperature at the kth time step, θ(k-1) represents the TCL temperature at the k-1th time step, and θ a Indicates the ambient temperature, ambient temperature θ a The dynamic change of P is much slower than that of θ(k); mrepresents the rated power, q(k)∈{0,1} represents a binary variable, which is used to represent the working state of the system as "off / on", for example, '0' can represent the working state of the system as off, and '1' can represent the working state of the system as on. a represents the energy consumption rate, and b represents the resistance and capacitance coefficient.

[0049] The formula including energy consumption rate and resistance and capacitance coefficient is: a=e-ΔT / (R th C th )≈1-ΔT / (R th C th ), b=R th ·η. Where ΔT represents the sampling time, e represents the base of the natural logarithm; R th is thermal resistance, which is used to represent the resistance characteristics of the TCL thermal system and can affect the thermal energy loss of the system; C th is the heat capacity, which is used to represent the capacitance characteristics of the TCL thermal system and can affect the system's response speed to temperature changes; η is the efficiency factor of TCL, which represents the equipment energy efficiency conversion coefficient and is used to describe the efficiency of the refrigeration equipment in converting electrical energy into thermal energy.

[0050] Without loss of generality, assume that each TCL has a rated power P m >0, TCL switches between the "on" and "off" states subject to the following local control rules, namely:

[0051]

[0052] Among them, q(k) represents the working state of TCL at the kth time step, which is a binary variable with a value of 1 or 0; it can be understood that q(k-1) represents the working state of TCL at the k-1th time step. θ(k-1) represents the TCL temperature at the k-1th time step, that is, the temperature value of the system in the previous time step, θ r is the temperature set point set by the user, Δ is the dead zone, and Δ>0 means half of the dead zone.

[0053] Step S1-2, linear system model can usually simplify the expression of optimization problems, especially when convex optimization methods are used. After the nonlinear switching model is approximated as a linear model, the optimal control problem can be solved more quickly and it can be more easily integrated into optimization scenarios such as demand response and energy management. Therefore, this embodiment chooses to establish a linear system model of TCLs. The nonlinear characteristics of the nonlinear switching model are obtained based on the obtained temperature changes and the constraints of the local control rules of the constant temperature control load TCLs; according to the obtained nonlinear characteristics, the nonlinear switching model is converted into a linear system model, including, combined with the nonlinear characteristics of the above-mentioned switching models (1) and (2), a linear system model can be used to approximate the nonlinear switching model, that is:

[0054] θ(k)=aθ(k-1)+(1-a)(θ a -bP(k)) (3)

[0055] Among them, P(k)∈[0,P m ] is a continuous variable, not a binary input {0,P m};θ a For the ambient temperature, the overall behavior of a large number of TCLs can be accurately approximated by model (3) to model (1) and (2). The continuous power input P(k) can be regarded as the average of the binary power inputs in models (1) and (2) over time.

[0056] Furthermore, for a large number of TCLs, the power of the overall linear system model can be closely matched to the power of the nonlinear switching model; by variable substitution, x(k) = C th (θr-θ(k)) / η and u(k)=P(k)-P0(k), where P0(k)=(θa-θr) / b represents the nominal power when the temperature reaches the set point in model (3), which represents the power required for TCL to maintain the set temperature θr under the ambient temperature θa; this power value is a reference value. When the difference u(k) between the actual power and the nominal power of TCL is zero, the temperature can be stabilized near the set value. The above (3) can be rewritten as:

[0057] x(k)=ax(k-1)+u(k)δ (4)

[0058] Among them, x(k) represents the energy storage state of TCL at time k, that is, the energy level stored in the system; including, x(k) as a state variable, can be used to represent the measure of the current temperature deviation from the set temperature (after proportional adjustment); x(k) uses the heat capacity Cth and the efficiency factor η for scaling, converting the temperature deviation into power-related units, which is convenient for modeling in combination with the energy control quantity u(k). δ is a constant related to the time step and thermal parameters, and δ=(1-a)R th Cth ≈ΔT, which can reflect the time evolution rate of the TCL system. In addition, the model has energy constraints and input constraints, including: the energy constraint is expressed as x(k)∈[-x - ,x + ], the input constraint is expressed as u(k)∈[-u - (k),u + (k)]; and x + =x - =C th Δ / η,u - (k) = P0(k), u + (k) = Pm-P0(k). - and x + Represents the minimum energy constraint and the maximum energy constraint, respectively, -u - (k) and u + (k) represents the minimum input constraint and the maximum input constraint respectively; u(k) represents the power input, and Pm represents the rated power. The maximum power output of TCL is used to describe the peak load capacity of the device. P0(k) represents the minimum power (or no-load power) consumed by TCL at time step k, that is, when TCL is turned off, the device may still consume a small amount of standby power, that is, the power contribution of TCL to the system at time k.

[0059] In step S2, the individual flexibility of the thermostatically controlled loads TCLs is defined based on the linear system model, and the overall flexibility is obtained based on the individual flexibility; the overall flexibility is used to represent the response margin of the thermostatically controlled loads TCLs, and the response margin is introduced into the virtual battery model to model the overall flexibility of the thermostatically controlled loads TCLs; the optimization algorithm based on geometric information optimizes the parameters of the virtual battery model with the response margin, and obtains the optimized virtual battery model. This can be achieved specifically through the following process:

[0060] Consider a heterogeneous thermostatically controlled load group TCLs described by model (4), which contains N TCLs. The parameters of each TCL are given by Ω i Indicates, specifically, in, It represents the thermal resistance of the ith TCL, reflecting the thermal insulation performance of the equipment; that is, the larger the thermal resistance, the slower the heat exchange between the TCL and the external environment, thereby reducing the frequency of temperature changes and reducing energy consumption. represents the thermal capacity of the ith TCL, reflecting the buffering ability of the device to temperature changes. Represents the set temperature of the i-th TCL, that is, the target temperature to be maintained by the device; this temperature determines the switch state adjustment of the device to ensure that the internal temperature remains within the range expected by the user.i is the temperature bandwidth or temperature tolerance of the i-th TCL, indicating the allowable deviation range of the temperature around the set value. i is the efficiency factor of the i-th TCL, which represents the energy efficiency of the device; this parameter measures the efficiency of the device in converting electrical energy into heat or cold energy, and is usually directly related to the energy consumption and performance of the device. i (0) is the initial temperature of the ith TCL, that is, the initial state temperature at time t=0; the initial temperature is used as the initial condition for simulation or control. is the rated power of the ith TCL, which represents the maximum power consumption of the device in the “on” state.

[0061] There are many feasible solutions for the total power consumption of the TCL group, which meet all temperature and power constraints. For demand response, the key to non-intrusive control of TCLs is to accurately characterize their overall power flexibility within the considered time horizon T: = {1,2,...m}; therefore, the individual flexibility and overall flexibility of TCLs are first defined, including:

[0062] Definition 1: For each TCLi (i = 1, ···, N), its individual flexibility is defined as the set of all allowed power curves, that is:

[0063]

[0064] Among them, P i represents the individual flexibility of the i-th TCL, i represents the TCL individual; [u i (k)] represents a vector whose kth element is u i (k) P i represents the individual flexibility of the i-th TCL (thermostatically controlled load); this is a collection of power curves that describe the TCL, that is, the range of power values ​​that the TCL can reach within a certain period of time. i represents the number of the TCL, and the value range is i=1,…,N, where N is the total number of TCL groups. u i (k) represents the control input or power regulation of the i-th TCL at time step k; it is a vector, each element u i (k) Describe the amount of flexible power regulation that TCL can provide at that moment. i (k) represents the state variable of the i-th TCL at time step k, which is generally related to physical states such as temperature or energy reserve; this state evolves over time and is controlled by the control input u i (k) Impact. i Represents the state update coefficient of TCL, controlling the state variable x i (k) The rate of change in a discrete time step. δ represents the discrete time step, i.e., the state variable x i(k) The length of the update interval at each time step. and denote the minimum and maximum allowed control inputs of the ith TCL at time step k, respectively; these values ​​limit the range of ui(k) and ensure that the power regulation of the TCL is within the actual physical constraints. and They represent the minimum and maximum state boundaries of the ith TCL at all time steps, respectively; they define the allowable range of state xi(k) to ensure that the temperature of the TCL does not exceed the safe or comfortable range.

[0065] From the obtained individual flexibilities, the overall flexibility of the thermostatically controlled loads TCLs is calculated based on the Minkowski sum, including, the overall flexibility of the TCL group is a set of power curves that satisfy the following conditions, namely:

[0066]

[0067] Among them, P represents the overall flexibility, so the overall flexibility can be further expressed as:

[0068]

[0069] Where U represents the Minkowski sum. i represents the power input of the ith TCL device, and is used to represent the power consumption of the ith device at a certain moment. m is a real number. The set P contains all acceptable overall power curves for the TCL population. However, the set P is difficult to integrate into the operation and control of the power system. In order to represent the overall flexibility in an intuitive way, a simple and portable virtual battery model is defined to describe the overall flexibility of TCL.

[0070] Overall flexibility refers to the set of all power curves that thermostatically controlled loads (TCLs) can provide within a certain time range, satisfying the physical constraints of the equipment (such as temperature range, power range) and the comfort requirements of users. The response margin is based on overall flexibility, further characterizing the actual range of power that the system can adjust at a certain moment or under a certain demand. It is a refinement and quantification of overall flexibility, which can be expressed in the following ways:

[0071]

[0072] Among them, P i is the individual flexibility of the ith TCL, defined as the set of powers that satisfy the device constraints (e.g., temperature range and power limit). is the Minkowski sum, which is used to represent the aggregation of multiple individual flexibilities. The response margin can be expressed as the power range that the system can actually provide at a specific moment or demand situation is called the response margin, which can be extracted from the overall flexibility set:

[0073] S 响应裕度 ={U(k)∈P|U(k)};

[0074] in, represents a subset of the total flexibility set, and U(k) is the variable of input power. The total flexibility is represented by a polyhedron, and the response margin is a part of the polyhedron, which describes the power regulation capability of TCLs in all time steps, that is:

[0075] P={U(k)∈R m |FU(k)≤H};

[0076] Among them, F is the constraint matrix, which defines the constraints such as power limit and temperature limit. H is the corresponding constraint vector, which contains the upper and lower limits of power and state variables; R m Represents a matrix consisting of all possible power adjustment combinations. The overall flexibility describes all possible power adjustment sets, while the response margin is a subset that is further reduced for the current moment and the adjustment target. The response margin can be further specified through two sets of constraints:

[0077] S 响应裕度 = {U(k)∈R m FU(k)≤H,GU(k)≤J};

[0078] Among them, G represents the additional constraint matrix related to the target; J represents the boundary value of the regulation demand. Based on the first set of constraints (FU(k)≤H), the constraints of overall flexibility are defined, and based on the second set of constraints (GU(k)≤J), the additional constraints of specific regulation targets are defined: By adding the above two sets of constraints, the overall flexibility (equipment physical limitations) can be converted into actual available flexibility (response margin), making it more in line with the demand response regulation target, so as to more accurately control the flexibility space and improve the optimization efficiency.

[0079] In the power reduction scenario, limit the current power to U(k)≤P max The power reduction range is to reduce the load pressure or avoid overload operation. In the power increase scenario, the current power is limited to U(k) ≥ P min Increase the range to fill the load gap or optimize power utilization. Finally, the response margin set S is expressed as:

[0080] S 响应裕度 =P∩Q;

[0081] Where P = {U(k)∈R m |FU(k)≤H} is the total flexibility set, Q={U(k)∈R m}|GU(k)≤J} is a set of additional constraints for the adjustment target.

[0082] In order to describe and apply the overall flexibility of thermostatically controlled loads (TCLs), this embodiment first defines individual flexibility as the set of all possible power inputs of a single TCL within a time range; the overall flexibility is characterized by the Minkowski sum of individual flexibility to describe the power regulation capability of the entire TCLs group within a specific time range. Since the overall flexibility set is difficult to apply directly to optimization and control, a virtual battery model is introduced to simplify the overall flexibility set into an easy-to-operate form. The virtual battery model mathematically approximates the overall flexibility set, which can provide a clear model framework in complex TCLs group regulation problems to represent power regulation and energy storage characteristics. Specifically, an m-dimensional discrete-time virtual battery model can be expressed by Definition 2, namely:

[0083] Definition 2: An m-dimensional discrete-time virtual battery model is a set of power curves that satisfy the following conditions, namely:

[0084]

[0085] Furthermore, the introduction of the virtual battery model requires the following steps:

[0086] First, determine the upper and lower power limits, that is, extract the upper and lower power limits from the response margin set S:

[0087]

[0088] Among them, D _ (k) and D + (k) represent the lower and upper limits of the regulation power of the thermostatically controlled load TCLs at time step k, respectively.

[0089] Then, the energy capacity is determined, that is, the power range of the cumulative response margin set, and the upper and lower limits of energy storage are obtained:

[0090]

[0091] Among them, E _ and E + The above power constraints and energy constraints are extracted from the parameter set φ to control the evolution of the virtual battery state.

[0092] The state X of the virtual battery is determined by the charge and discharge power U(k). In this embodiment, the evolution of the charge state is defined by a dynamic state equation, namely:

[0093]

[0094] Among them, U(k) is the dynamic charge and discharge power, which is subject to the upper and lower power limits D + (k) and D - (k). X(k) is subject to the upper and lower limits of energy storage capacity E + and E - a is the energy decay factor (for example, representing self-discharge loss); δ is the time step used to convert the power U(k) into energy increment.

[0095] The parameter set φ is the core of the virtual battery model and is used to define the constraints of the model. The mapping relationship with the state equation can be established in the following ways, including:

[0096] The virtual battery model can be specified by the following parameters:

[0097]

[0098] Among them, φ represents the parameter set of the virtual battery model, and the virtual battery model is concisely expressed as In addition, if Then B(φ) is called a sufficient condition; if Then B(φ) is called a necessary condition. The definition of the virtual battery model naturally comes from the definition of individual flexibility. U(k) is regarded as the power consumption of the battery, and X(k) is regarded as the charging state of the battery, which is used to represent the level of energy stored in the battery; X(0) represents the initial charging state of the battery at the initial time t=0, which represents the energy level stored in the battery at the beginning of the model and is the initial condition for the operation of the model. k represents the discrete time step, that is, the index of the current time; including, each k value corresponds to a time point, and as k increases, the system state will be gradually updated according to the time step. Number D + (k) and D - (k) represents the time-varying charge / discharge rate limit, while E + and E - represent the upper and lower limits relative to the nominal energy level of the battery, respectively, and parameter a represents the self-discharge rate of the battery, which is caused by the heat exchange between the air and the environment.

[0099] Furthermore, after mapping the set of geometric polyhedrons to the parameters of the virtual battery model, the Minkowski distance can be used to optimize the parameters. Specifically, by using the Minkowski distance to optimize the capacity, charge and discharge rate, efficiency, response time and flexibility range of the virtual battery, the comprehensive optimization of each parameter can be achieved under multi-dimensional constraints. In this method, reasonable weights are first set for each parameter to reflect their contribution to the overall performance of the virtual battery model. Then, the Minkowski distance is used to measure the relative differences of these parameters, and the total distance is minimized by adjusting the values ​​of the parameters, thereby achieving a balance between system operation efficiency, flexibility and economy. Specifically, the optimization of capacity and charge and discharge rate can ensure that the battery does not consume excessive energy or damage the battery when meeting grid load fluctuations; efficiency optimization reduces energy loss and improves economy; and the optimization of response time and flexibility range can ensure that the virtual battery responds to grid demand in a timely manner and provides sufficient regulation capabilities.

[0100] Based on step S2, this embodiment, on the one hand, models the power flexibility of a single TCL as a polyhedron through polyhedron modeling and Minkowski sum calculation, which can accurately capture the power regulation range of TCL; and calculates the aggregate flexibility of multiple TCLs through Minkowski sum operation, proving that the feasible power trajectory set of TCL is a polyhedron, and its aggregation characteristics can be accurately described by the Minkowski sum of these polyhedrons, providing an accurate mathematical basis for power regulation; this innovative method provides a more intuitive and efficient solution for the flexibility modeling of the company's own buildings in the power grid, helping to build a more intelligent, efficient and low-carbon building energy system. On the other hand, this embodiment simplifies the representation of TCL aggregation flexibility by introducing a virtual battery model. The model uses parameters such as charge and discharge power limit, energy capacity limit and self-discharge rate to vividly characterize the aggregation characteristics of TCL, that is, through this model, the temperature control load of the owned building can be simplified into a manageable virtual battery, which can greatly improve the flexibility and energy efficiency management level of the building.

[0101] In step S3, a geometric polyhedron set is used to represent the optimized virtual battery model, and an optimization algorithm based on linear programming is used to find the optimal scaling and translation of the virtual battery model; based on the obtained optimal scaling and translation results, the response margin of the thermostatic control load TCLs is dynamically adjusted to achieve the aggregated demand response coordination of the building flexible resources. This can be achieved through the following process:

[0102] Step S3-1. Polyhedral interpretation of overall flexibility. A polyhedron Q is the solution set of a finite linear inequality system, i.e., Q: = {U∈R m |FU≤H}; where the symbol ≤ is used to indicate element-wise inequalities. Contains R m If there is an interior point in , then it is called full-dimensional. The overall flexibility of TCL can be represented by a polyhedron. i =[x i (k)] and U i =[u i (k)] represent the state and input vector respectively, so the model (4) can be rewritten as:

[0103] A i X i =B i U i +C i (10)

[0104] Among them, A i =I m +diag(-a i ; -1) represents the discretized state matrix, which is a lower tridiagonal matrix. This matrix is ​​used to describe the state variable X in the TCL system. i The update relationship of the lower tridiagonal matrix A i The main diagonal is 1, and the elements on the secondary diagonal are (-a i ,-1),a i Represents the self-discharge rate or temperature attenuation coefficient of the TCL system, the lower tridiagonal matrix A i Can help define how the system state changes over time. i The inverse of can be used with respect to a i The polynomial of is clearly expressed, denoted by Λ i This means that Λ i is a function used to calculate A i The auxiliary matrix of the inverse matrix. B i Represents the discretized input matrix, which is used to represent the control input U in the system i For system status X i The influence of B i =δI m ; where δ represents the time step; I m represents an m-dimensional identity matrix that can represent the gradual cumulative effect of the input on the state variables. i represents the initial state or constant term vector of the system, which defines the fixed offset or external influence of the system at the initial moment, including, C i =[a i x i (0),0,,0] T .

[0105] In addition, X i and U i The set of constraints is:

[0106]

[0107] Among them, X i represents the state vector of the i-th TCL. The state vector X i contains the values of physical quantities such as temperature or charge state at different time steps, and is used to describe the evolution process of the TCL system state. U i represents the input vector of the i-th TCL. The input vector U i contains the control input values at different time steps, such as power consumption or switch state, and is used to control the behavior of the TCL system in this embodiment. and 1 / m is an m-dimensional column vector with all elements being 1.

[0108] Through equations (5) and (6), the individual flexibility Pi of the i-th TCL i can be expressed as:

[0109] Pi i = Ui i ∩ Xi i (12)

[0110]

[0111] Since -∞ < -ui < ui < +∞, it can be directly proved that Ui i is a full-dimensional hyper-rectangle (and thus a polyhedron). Similarly, since A i and B i are invertible and -∞ < -xi < xi < +∞, it can be proved that Xi i is also a full-dimensional polyhedron, and their intersection Pi i = Ui i ∩ Xi i is a polyhedron (if ). In addition, it can be proved that the Minkowski sum P = ∑i=1n Pi of polyhedra is also a polyhedron.

[0112] For each TCL, its individual flexibility Pi i can be determined by equations (5)-(7). However, when the number of TCLs is large, the numerical complexity of calculating their Minkowski sum is extremely high. In fact, when two sets Q1 and Q2 are polyhedra specified by faces, calculating their Minkowski sum is NP-hard because the number of faces of the resulting polyhedron can grow exponentially with the number of faces of Q1 and Q2. Therefore, an alternative method is adopted to find its maximum inner approximation (subset approximation) and minimum outer approximation (superset approximation).

[0113] Step S3-2: Optimal approximation of overall flexibility. In this section, we find the set P s and P n , so that Such a set P s and P n They are called sufficient approximation and necessary approximation respectively. Given a power curve U, if It can be concluded that U is not an acceptable overall power curve for TCL. On the other hand, if U∈P s , then there exists a decomposition of U such that U = ∑N i =1U i , and U i ∈P i This holds true for all i=1,...,N.

[0114] Given a compact convex set P o , we call P o The homeomorphism of P o expansion and displacement, where β i >0 is the scaling factor, t i ∈R m is the displacement factor. Since all P i With the same structure (7), for each P i With respect to a given set P o Perform inner and outer approximation. Set P o It will be referred to as the prototype set in the following. Specifically, in P o Find each P in the homeomorphism i The best approximation of . Figure 2 As shown, a two-dimensional example (i.e., the time range is m=2) is used to illustrate this idea, where β +i P o +t +i \ is P i The minimum outer approximation of -i P i + t-i \ is P i The maximum inner approximation of is:

[0115] (β 1 Q+t 1 )U(β 2 Q+t 2 )=(β 1 +β 2 )Q+(t 1 +t 2 ) ← (14)

[0116] The Minkowski sum of homeomorphisms of convex sets can be simplified to the sum of scaling factors and the sum of displacement factors. Therefore, once P of the individual flexible polyhedra is obtained o The homeomorphism approximation makes it very easy to calculate the approximation of the overall flexibility. 1 , β 2 is the scaling factor, t 1 , t 2 is the displacement factor and Q is the set of individual thermostatically controlled load (TCL) flexibility polyhedra.

[0117] For choosing the best approximation P i P o Homeomorphism, including, for each i=1,...,N i =1,...,N, hoping to find the i The maximum P in o Homeomorphism, which can be expressed by the following mathematical expression:

[0118]

[0119] Among them, β i is the optimization variable associated with the ith TCL, usually representing a weight or scaling factor in a scheduling or control strategy; it may control some characteristic of the ith TCL, such as priority, weight, or some operating parameter, to achieve a balance or constraint during the optimization process. i is another optimization variable of the ith TCL, usually representing time-related parameters such as start or end time, duty cycle, etc. o It is a reference flexibility set or a polyhedron that represents a set of allowed operating states or power sets. It is a mathematical representation of the flexibility or power requirements of the TCL system and is used to define the operational region or acceptable state of the TCL device. i is the power set or flexibility set of the ith TCL, representing the power requirement or state set of the TCL device.

[0120] Contains P i The minimum P o Homeomorphism can be expressed by the following mathematical expression:

[0121]

[0122] In the above optimization problem, optimality is based on the inclusion relation, that is, if (β*i,t*i,) is an optimal solution, then there is no other P o The homeomorphism can be between (β*i,P o +t*i) and P iThe optimal solutions of formula (9) and (10) are called maximum inner approximation (MIA) and minimum outer approximation (MOA), respectively. i ,t i )=MIA(P i ,P o ) and (β i ,t i )=MOA(P i ,P o ) represents the solution to these problems. Where Pi is the target polyhedron and Po is the reference polyhedron.

[0123] In order to make the MIA and MOA problems solvable, P must be specified in advance. o Since each P i is a polyhedron. In order to achieve a good approximation to it, it is natural to choose P o is also a polyhedron. Furthermore, we show that under this choice, the MIA and MOA problems can be solved efficiently via equivalent linear programming problems. Solving (β i ,t i )=MIA(P i ,P o ) and (β i ,t i )=MOA(P i ,P o ) is as follows. According to formulas (5)–(7), the individual flexibility polyhedron can be written as P i = {U i ∈R m ∣F i U i ≤H i}, where F i and H i are the matrices and vectors that define the polyhedron, including:

[0124] F i =(I m ,-I m ,Λ i B i ,-Λ i B i ) (17)

[0125]

[0126] Among them, I m is an m-dimensional co-diagonal matrix.

[0127] Step S3-3, flexibility characterization based on virtual battery. Assume that the battery parameter H is predetermined and fixed (for example, taking the average value of all TCL parameters), and focus on estimating its power limit, energy capacity and initial energy state. Therefore, the parameter Represented as the parameters of a virtual battery, where the vector symbols are C = [aX(0), 0, ..., 0] T , D =[D - (k)], E =[E - (k)], Use B * As a virtual battery B(φ * ), where * represents the subscripts o, s, and n, which stand for the prototype, sufficient, and necessary virtual batteries, respectively.

[0128] in, D =[D - (k)] and Respectively represent the lower and upper limits of the virtual battery power, E =[E - (k)], They represent the lower and upper limits of the virtual battery energy respectively; C is the initial condition vector of the virtual battery, and H represents the dynamic parameters of the battery.

[0129] A. Optimal virtual battery. Select P o With P i, s parameter of the same form Ω o , which is the parameter of all TCL parameters Ω i, s. Obviously, the P o This is also a virtual battery model. To emphasize this point, this prototype virtual battery is represented as B o .

[0130] Prototype Virtual Battery B o Parameters is obtained from the average value of the TCL parameters. According to formulas (5)–(7), let the prototype virtual battery be represented by B o = {U∈R m :FU≤H}, where F=(I m ,-I m ,AB,-AB) and

[0131] Then, order and To prove that βB o +t is a sufficient or necessary battery, depending on (β i ,ti ) comes from formula (11) or (12).

[0132] Theorem 2: B * :=βB o +t is a sufficient condition if (β i ,t i )=MIA(P i ,B o )) and is a necessary condition if (β i ,t i )=MOA(P i ,B o ), battery parameters It can be given by the following formula:

[0133]

[0134] In addition, for all The feasible power curve for a single device is given by:

[0135]

[0136] B. Suboptimal virtual battery. In order to further reduce the computational complexity, a suboptimal method is proposed. In the above optimal method, P i =U i ∩X i Relative to B o Alternatively, its components U can be approximated separately i and X i While this is usually more conservative than the optimal approach, it significantly reduces the numerical complexity since the corresponding linear programming problem has a smaller dimension.

[0137] Including, first consider the full battery characterization. To simplify the notation, B o =U o ∩X o It is recorded as a virtual battery model with the structure of formula (7). This method is mainly divided into three steps. Figure 3 As shown (i.e., time range m=2). Assume that X o As B o Given a component of , in the first step, each X i Relative to X o By approximation, we get (β i , t i )=MIA(X i , X o ); Next, in order to ensure B o and P i The homeomorphic transformation between them can be enforced by Or equivalently, ensure that This constraint ensures that the power curve U i Can be included in the virtual battery model U o This approach treats each component U independently. i and X i Although more conservative, this approach significantly reduces the computational complexity and makes the algorithm more efficient when processing a large number of TCLs.

[0138]

[0139] Because U i are all hyperrectangular, so the value on the right side of equation (15) can be calculated accurately, as shown in equation (21). Obviously, this is the maximum U that satisfies equation (20). o The last step is to get B o For U o ∩X o , then through βB o +t to get B s , Bs is the sufficient / necessary virtual battery model. Next, it is proved that Bo obtained by this method makes βBo+t a sufficient condition.

[0140] Theorem 3: First, set the equation (β i ,t i )=MIA(X i ,X o ), and U o It can be given by the following formula: Where R m is the real number space, that is:

[0141]

[0142] min is element-by-element, then B s :=βB o +t is a full battery. Battery parameters Given by formula (13). In addition, for , the individual allowed power configuration can be obtained by equation (14).

[0143] Theorem 4: Let (β i ,t i )=MOA(X i ,X o )and Then B n :=(βX o +t)∩U o is a necessary battery. The battery parameters are given by the following formula:

[0144]

[0145] From equation (23), it can be seen that the way to calculate the power boundary of the suboptimal necessary battery is to sum up each power boundary. i X o If we further abandon these optimization schemes (i.e., MIA (X in Theorem 3 and Theorem 4) i ,X o ) and MOA(X i ,X o ), and using the matrix norm inequality to obtain the energy bound from (5) will yield the corresponding method in the discrete-time finite-time case. In addition, the optimal method proposed in Theorem 2 directly applies to P i approximation and therefore provides the best performance among all these methods.

[0146] Obviously, the main calculation of the suboptimal method only involves solving MIA(X i ,X o ) or MOA(X i ,X o ), while the constraints involved are reduced by about 50% and the decision variables are reduced by about 75% compared to the optimal method. Therefore, the suboptimal method can be solved faster than the optimal method.

[0147] After finding the optimal scaling and optimal translation of the virtual battery model, the response margin of the thermostatic control load TCLs is dynamically adjusted based on the optimal scaling and translation results to achieve the aggregated demand response coordination of building flexible resources, including: The core purpose of the optimal scaling and translation results is to adjust the response margin set S 响应裕度 Adjustments are made to better meet the grid's demand response goals while maintaining the operating comfort of the building's internal thermostatically controlled loads. The response margin set is linearly transformed by optimizing the optimal scaling factor α and translation factor β. The specific adjustment formula is as follows:

[0148]

[0149] in, Represents the response margin set after scaling and translation adjustment. α represents the scaling factor of the response margin set, which is used to adjust the amplitude of the power adjustment range; β represents the translation factor of the response margin set, which is used to adjust the reference point of power adjustment. Adjusted response margin set It is used to guide the dynamic adjustment power U(k) of TCLs, which requires further dynamic adjustment steps, namely:

[0150] At each time step k, the power output U(k) of TCLs is dynamically updated according to the optimal adjustment result, and its constraints are:

[0151] D-(k) 调整 ≤U(k)≤D + (k) 调整 ;

[0152] Among them, D - (k) 调整 =αD - (k)+β is the lower power limit after scaling and translation, D + (k) 调整 =αD + (k)+β is the power upper limit after scaling and translation. The control strategy of TCLs dynamically adjusts the power output U(k) according to the above constraints to achieve real-time response to the grid regulation demand. The adjusted response margin set Aggregate the flexibility of multiple thermostatically controlled loads TCLs and achieve demand response coordination through the following steps:

[0153] The overall adjustment target U 目标 (k) is proportionally allocated to the response margin set of each TCL. Within the allocated range, the optimal power regulation U(k) is solved by linear programming or heuristic algorithm to satisfy:

[0154]

[0155] Real-time monitoring of the operating status of the thermostatically controlled load TCLs and environmental conditions (such as outdoor temperature changes, etc.), and updating the scaling factor α and the translation factor β based on these real-time data, and the response margin set S 响应裕度 Perform dynamic optimization and adjustment. The optimized response margin set Used to guide the power regulation behavior of TCLs, ensuring that the safety of equipment operation and user comfort are maintained while meeting the needs of the grid, and achieving rapid response to the dual needs of the grid and buildings

[0156] Based on step S3, this embodiment further develops an optimization algorithm based on geometric information, which optimizes the parameters of the virtual battery model so that it can better approximate the aggregate flexibility of the TCL, improve the computational efficiency and the operability of the system. At the same time, a linear programming-based optimization algorithm is designed to achieve the best internal and external approximation of a single TCL flexibility polyhedron by finding the optimal scaling and translation of a prototype set (such as a virtual battery model); wherein the optimal scaling refers to scaling the geometric volume of the virtual battery model so that the model can maximize the efficiency according to the specific requirements of the demand response system. By adjusting the scaling factor, the "capacity" or "flexibility" of the virtual battery can be made more suitable for the requirements of the system. The scaling factor here usually controls the proportional change of the virtual battery model in terms of numerical values, thereby ensuring the overall balance and optimization effect of the system. The optimal translation refers to ensuring that the virtual battery model can achieve the best scheduling and resource allocation within a certain specific spatial range by adjusting the position of the geometric model. The translation factor usually affects the power output or response interval of the model at a specific time, ensuring that the behavior and response of the virtual battery can be connected with external demand (such as grid load) to maximize the role of its flexibility resources. The specific optimization steps of the scaling factor and translation factor are as follows: (1) Initialize parameters and set the initial values ​​of the scaling factor S and the translation factor T; (2) Calculate the objective function: Calculate the objective function value based on the current scaling factor and translation factor; (3) Update parameters: Update the scaling factor and translation factor according to the optimization algorithm (such as genetic algorithm or particle swarm optimization); (4) Iterate optimization: Continuously adjust the scaling factor and translation factor until the objective function converges to the optimal solution. The algorithm can effectively calculate the aggregated flexibility, which not only improves the mathematical accuracy of TCL scheduling, but also provides a practical tool for the optimal scheduling of large-scale flexibility resources in the power grid.

[0157] The above method provided by the present invention provides a new path for flexibility modeling and optimization of self-owned buildings, which helps buildings participate in grid ancillary services, improves overall energy efficiency and supports the deep integration of renewable energy. It can be understood that the relevance of TCL flexibility to building energy efficiency management and grid ancillary services is mainly reflected in three aspects, including:

[0158] As for the building energy efficiency management level, by flexibly adjusting the operating status of TCL equipment such as air conditioners and heat pumps, the overall energy consumption performance of the building can be dynamically optimized. During the peak period of energy demand, the peak load of the building can be reduced by reducing the operating power of TCL equipment; during the trough period of energy demand, the excess power generation of low-cost electricity or renewable energy can be effectively utilized by increasing the operating power of TCL equipment. Therefore, the present invention accurately models and optimizes the flexibility of TCLs based on the virtual battery model, which can support more sophisticated energy consumption regulation, thereby improving the energy management level of the building.

[0159] As for the auxiliary services of the power grid, the present invention can effectively deal with the volatility and uncertainty problems brought about by the access of renewable energy to the power grid by coordinating the flexibility of the thermostatically controlled load TCLs, and provide the power grid with auxiliary functions such as frequency regulation, peak shaving and valley filling, and emergency load regulation. Including, the flexibility of TCLs can help the power grid maintain stable operation by dynamically adjusting the power response signal; cutting peak power or filling valley loads according to load changes; and quickly cutting non-essential power when electricity is tight. Therefore, the present invention can significantly improve the efficiency of TCLs participating in the auxiliary services of the power grid, effectively enhance the operational stability of the power grid and reduce the overall operating costs.

[0160] The present invention characterizes the flexibility of TCL as a battery-like charging and discharging behavior through a virtual battery model. This representation not only helps buildings optimize internal energy scheduling, but also directly provides valuable regulation resources for the power grid. Specifically, the virtual battery model enables buildings to better participate in grid demand response. By flexibly adjusting the power output of TCL, buildings can maximize support for grid stability without affecting user comfort.

[0161] Embodiment 2

[0162] This embodiment discloses a building flexible resource aggregation demand response system.

[0163] A building flexible resource aggregation demand response system, comprising:

[0164] The linear system model conversion module is configured to: construct a nonlinear switching model of the constant temperature control load TCLs in the building, and convert the nonlinear switching model into a linear system model according to the nonlinear characteristics of the nonlinear switching model;

[0165] The thermostatically controlled load flexibility modeling module is configured to: define individual flexibility of thermostatically controlled loads TCLs based on the linear system model, and obtain overall flexibility based on the individual flexibility; use the overall flexibility to represent the response margin of the thermostatically controlled loads TCLs, introduce the response margin into a virtual battery model to model the overall flexibility of the thermostatically controlled loads TCLs; optimize the parameters of the virtual battery model introduced with the response margin using an optimization algorithm based on geometric information to obtain an optimized virtual battery model;

[0166] The thermostatic control load flexibility characterization module is configured as follows: using a set of geometric polyhedrons to represent the optimized virtual battery model, and using an optimization algorithm based on linear programming to find the optimal scaling and optimal translation of the virtual battery model represented by the set of geometric polyhedrons; dynamically adjusting the response margin of the thermostatic control load TCLs based on the obtained optimal scaling and optimal translation to achieve aggregated demand response coordination of building flexible resources.

[0167] Embodiment 3

[0168] The purpose of this embodiment is to provide a computer-readable storage medium.

[0169] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps in a building flexible resource aggregation demand response method as described in the first embodiment of the present disclosure.

[0170] Embodiment 4

[0171] The purpose of this embodiment is to provide an electronic device.

[0172] An electronic device includes a memory, a processor, and a program stored in the memory and executable on the processor. When the processor executes the program, the steps in a method for responding to building flexible resource aggregation demand as described in the first embodiment of the present disclosure are implemented.

[0173] The steps involved in the apparatuses of the above embodiments 2, 3 and 4 correspond to the method embodiment 1, and the specific implementation methods can refer to the relevant description part of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0174] Those skilled in the art should understand that the modules or steps of the present invention described above can be implemented by a general-purpose computer device, or alternatively, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.

[0175] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.

Claims

1. A building flexible resource aggregation demand response method, characterized in that: include: Constructing a nonlinear switching model of a thermostatically controlled load TCLs in a building, and converting the nonlinear switching model into a linear system model according to the nonlinear characteristics of the nonlinear switching model; The individual flexibility of the thermostatically controlled loads TCLs is defined based on the linear system model, and the overall flexibility is obtained based on the individual flexibility; the overall flexibility is used to represent the response margin of the thermostatically controlled loads TCLs, and the response margin is introduced into the virtual battery model to model the overall flexibility of the thermostatically controlled loads TCLs; the parameters of the virtual battery model introduced with the response margin are optimized by an optimization algorithm based on geometric information to obtain an optimized virtual battery model; The optimized virtual battery model is represented by a set of geometric polyhedrons, and an optimization algorithm based on linear programming is used to find the optimal scaling and optimal translation of the virtual battery model represented by a set of geometric polyhedrons. The response margin of the thermostatically controlled load TCLs is dynamically adjusted based on the obtained optimal scaling and optimal translation to achieve the aggregated demand response coordination of building flexible resources.

2. A building flexible resource aggregation demand response method as claimed in claim 1, characterized in that: According to the nonlinear characteristics of the nonlinear switching model, the nonlinear switching model is converted into a linear system model, including: obtaining the temperature change of each individual thermostatically controlled load in the thermostatically controlled loads TCLs, and obtaining the nonlinear characteristics of the nonlinear switching model based on the temperature change of each individual thermostatically controlled load in the thermostatically controlled loads TCLs and the constraints of the local control rules of each thermostatically controlled load TCLs; according to the obtained nonlinear characteristics, the nonlinear switching model is converted into a linear system model.

3. A building flexible resource aggregation demand response method as claimed in claim 2, characterized in that: Obtaining the temperature change of each individual thermostatically controlled load in the thermostatically controlled loads TCLs includes: using a discrete time switching model to characterize the temperature change of each individual thermostatically controlled load in the thermostatically controlled loads TCLs.

4. A building flexible resource aggregation demand response method as claimed in claim 1, characterized in that: The individual flexibility of the thermostatically controlled loads TCLs is defined based on the linear system model, and the overall flexibility is obtained based on the individual flexibility, including: defining the individual flexibility of the thermostatically controlled loads as a set of all allowed power curves; and calculating the overall flexibility of the thermostatically controlled loads TCLs based on the obtained individual flexibility and in combination with the Minkowski sum.

5. A building flexible resource aggregation demand response method as claimed in claim 1, characterized in that: An optimization algorithm based on geometric information is used to optimize the parameters of the virtual battery model with the response margin introduced. Specifically: the geometric characteristics of the geometric polyhedron set are used as the input of the virtual battery model with the response margin introduced, and the geometric polyhedron set is mapped to the virtual battery model parameters; the Minkowski distance is used to optimize the parameters.

6. A building flexible resource aggregation demand response method as claimed in claim 5, characterized in that: The geometric properties of the collection of geometric polyhedra include power limitations, energy storage capacity, and charge and discharge rates.

7. A building flexible resource aggregation demand response method as claimed in claim 1, characterized in that: An optimization algorithm based on linear programming is used to find the optimal scaling and optimal translation of a virtual battery model represented by a set of geometric polyhedrons, including: finding the optimal solution of the sum of scaling factors and the sum of displacement factors by calculating the homeomorphic Minkowski sum of convex sets in the virtual battery model represented by the set of geometric polyhedrons; using the optimal solution of the sum of scaling factors as the optimal scaling of the virtual battery model represented by the set of geometric polyhedrons, and using the optimal solution of the sum of displacement factors as the optimal translation of the virtual battery model represented by the set of geometric polyhedrons.

8. A building flexible resource aggregation demand response system, characterized by: include: The linear system model conversion module is configured to: construct a nonlinear switching model of the constant temperature control load TCLs in the building, and convert the nonlinear switching model into a linear system model according to the nonlinear characteristics of the nonlinear switching model; The thermostatically controlled load flexibility modeling module is configured to: define individual flexibility of thermostatically controlled loads TCLs based on the linear system model, and obtain overall flexibility based on the individual flexibility; use the overall flexibility to represent the response margin of the thermostatically controlled loads TCLs, introduce the response margin into a virtual battery model to model the overall flexibility of the thermostatically controlled loads TCLs; optimize the parameters of the virtual battery model introduced with the response margin using an optimization algorithm based on geometric information to obtain an optimized virtual battery model; The thermostatic control load flexibility characterization module is configured as follows: using a set of geometric polyhedrons to represent the optimized virtual battery model, and using an optimization algorithm based on linear programming to find the optimal scaling and optimal translation of the virtual battery model represented by the set of geometric polyhedrons; dynamically adjusting the response margin of the thermostatic control load TCLs based on the obtained optimal scaling and optimal translation to achieve aggregated demand response coordination of building flexible resources.

9. A computer-readable storage medium having a program stored thereon, characterized in that: When the program is executed by a processor, the steps in the building flexible resource aggregation demand response method as described in any one of claims 1 to 7 are implemented.

10. An electronic device comprising a memory, a processor, and a program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the building flexible resource aggregation demand response method as described in any one of claims 1-7 are implemented.

Images