Method and device for modeling thermal dynamic model and electronic device

Abstract

The embodiment of the invention provides a modeling method and device for a thermal dynamic model and an electronic device. The method comprises the steps that a thermal dynamic model representing indoor temperature changes is determined; establishing an optimization problem with the minimum cost as a target, and determining a first gradient of the optimization problem for parameters of the thermal dynamic model; determining a second gradient of the precision loss function for the parameters of the thermal dynamic model; determining a collaborative gradient based on the first gradient and the second gradient; and updating parameters of the thermal dynamic model by iterating the collaborative gradient. According to the embodiment of the invention, the economic cost of building optimization operation can be effectively reduced while the model precision is ensured, and the flexible adjustment potential of building thermal dynamic can be more fully excavated, so that the energy consumption efficiency is improved, and the building energy consumption cost is reduced.

Description

Technical Field

[0001] This invention relates to the field of power and building energy system operation, and more specifically, to a method and device for energy cost-oriented building thermal dynamic control. Background Technology

[0002] Currently, my country is accelerating the construction of a new power system based on new energy sources, vigorously developing new energy, and gradually phasing out traditional energy sources based on the safe and reliable replacement of traditional energy sources, thus promoting a clean energy transition. However, with the high proportion of clean energy input, the uncertainty and volatility of its power generation are becoming more significant, affecting the power system's supply and demand balance and even causing supply gaps. Against this backdrop, improving the load-side's flexible adjustment capability has become an important means to meet supply and demand balance. Building temperature control is not only an important energy-consuming link on the load side but also a highly potential flexible resource. By optimizing the building temperature control process, proactive load reduction or transfer can be achieved, thereby effectively reducing the power supply pressure on the power system and improving the efficiency of building electricity use. The prerequisite for building temperature control is accurate modeling of complex thermal dynamic processes; however, current modeling methods ignore the potential impact of thermal dynamic model errors on optimization, making it difficult to achieve lower optimization costs. Further reduction in optimization costs is needed.

[0003] Therefore, in order to fully explore the flexible adjustment potential of building systems in the temperature control process and reduce energy consumption costs, an improved solution for building temperature control is needed. Summary of the Invention

[0004] In view of this, embodiments of the present invention propose a novel energy-cost-oriented building thermal dynamics modeling technology and equipment. This invention helps to achieve energy conservation and consumption reduction goals in power and building systems.

[0005] This invention provides a method for modeling a thermal dynamic model, comprising: determining a thermal dynamic model characterizing indoor temperature changes; establishing an optimization problem with the objective of minimizing cost, and determining a first gradient of the optimization problem with respect to the parameters of the thermal dynamic model; determining a second gradient of the accuracy loss function with respect to the parameters of the thermal dynamic model; determining a cooperative gradient based on the first and second gradients; and updating the parameters of the thermal dynamic model by iterating the cooperative gradient.

[0006] In one example, determining the thermal dynamics model characterizing changes in indoor temperature involves: determining the thermal dynamics model based on outdoor temperature data, air conditioning power data, solar radiation power data, indoor heat source power data, and indoor temperature data.

[0007] In one example, the thermal dynamics model is represented as Where M represents the thermal dynamics model, Let τ(t) represent the indoor temperature at time t with respect to time, τ(t) represent the indoor temperature data at time t, q(t) represent the air conditioning power data at time t, and x(t) = [τ...]. out (t), q rad (t), q occ [(t)] represents the set of other influencing factors, F(x(t)) represents the influence of the influencing factors in x(t) on the temperature derivative, θ represents the parameters of the neural network, and where a and b represent the linear weighting coefficients of τ(t) and q(t) with respect to the temperature derivative, respectively. out (t) represents the outdoor temperature data at time t, q rad (t) represents the solar radiation power data at time t, q occ (t) represents the indoor heat source power data corresponding to time t.

[0008] In one example, the input data for the thermal dynamics model includes outdoor temperature data τ. out (t), air conditioning power data q(t), solar radiation power data q rad (t) and indoor heat source power data q occ (t), where the output data of the thermal dynamics model includes indoor temperature data τ(t).

[0009] In one example, determining the first gradient includes: determining the gradient g of the optimization problem with respect to the parameters a, b, and θ of the thermal dynamics model. opt .

[0010] In one example, determining the second gradient involves determining the gradient g of the accuracy loss function with respect to the parameters a, b, and θ of the thermal dynamics model. phy .

[0011] In one example, the cooperative gradient is determined as:

[0012]

[0013] Where r is the radius parameter and represents the co-gradient and the second gradient g. phy The degree of deviation between them For the optimal combination gradient, ||g phy ||respectively and g phy The vector magnitude.

[0014] In one example, a thermal dynamics model was deployed on building temperature control equipment.

[0015] This invention also provides an apparatus for modeling a thermal dynamic model, comprising: a model determination component for determining a thermal dynamic model characterizing changes in indoor temperature; a first gradient determination component for establishing an optimization problem with the objective of minimizing cost and determining a first gradient of the optimization problem with respect to the parameters of the thermal dynamic model; a second gradient determination component for determining a second gradient of the accuracy loss function with respect to the parameters of the thermal dynamic model; a cooperative gradient determination component for determining a cooperative gradient based on the first and second gradients; and an update component for updating the parameters of the thermal dynamic model by iterating the cooperative gradient.

[0016] This invention also provides an electronic device, including: a transceiver; and a controller coupled to the transceiver and configured to perform the method described in any of the preceding embodiments.

[0017] Therefore, according to the embodiments of the present invention, while ensuring the accuracy of the model, the economic cost of building optimization operation can be effectively reduced, which is conducive to fully exploring the flexible adjustment potential of building thermal dynamics, thereby improving energy efficiency and reducing building energy consumption costs. Attached Figure Description

[0018] The invention will be more readily understood from the following detailed description with reference to the accompanying drawings, wherein like reference numerals designate units of the same structure, and wherein:

[0019] Figure 1 A schematic flowchart illustrating a method for modeling a thermal dynamics model according to an embodiment of the present invention is shown.

[0020] Figure 2 This illustrates the data acquisition requirements for building thermal dynamics modeling.

[0021] Figure 3 The modeling process for a building thermal dynamics model according to an embodiment of the present invention is shown.

[0022] Figure 4 A schematic block diagram of an apparatus for modeling a thermal dynamic model according to an embodiment of the present invention is shown.

[0023] Figure 5 A schematic block diagram of an electronic device according to an embodiment of the present invention is shown. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] Figure 1 A schematic flowchart illustrating a method 100 for modeling a thermal dynamics model according to an embodiment of the present invention is shown.

[0026] In step 110 of method 100, a thermal dynamic model characterizing changes in indoor temperature is determined.

[0027] In step 120, an optimization problem with the goal of minimizing cost is established, and the first gradient of the optimization problem with respect to the parameters of the thermal dynamic model is determined.

[0028] In step 130, the second gradient of the accuracy loss function with respect to the parameters of the thermal dynamics model is determined.

[0029] In step 140, a cooperative gradient is determined based on the first gradient and the second gradient.

[0030] In step 150, the parameters of the thermal dynamics model are updated by iterating the cooperative gradient.

[0031] Therefore, according to the embodiments of the present invention, while ensuring the accuracy of the model, the economic cost of building optimization operation can be effectively reduced, which is conducive to fully exploring the flexible adjustment potential of building thermal dynamics, thereby improving energy efficiency and reducing building energy consumption costs.

[0032] Figure 2 This illustrates the data acquisition requirements for building thermal dynamics modeling.

[0033] According to embodiments of the present invention, a thermal dynamic model, such as a building thermal dynamic model, can characterize the mapping relationship between different influencing factors and indoor temperature changes. To identify the building thermal dynamic model, the data required for training is first collected, including the input data and corresponding output data of the thermal dynamic model.

[0034] The input data for the thermal dynamics model can include the outdoor temperature data τ corresponding to time t. out (t), air conditioning power data q(t) at time t, solar radiation power data q at time t rad (t) and indoor heat source power data q at time t occ (t). In addition, the output data of the thermal dynamics model is the indoor temperature τ(t).

[0035] Then, the model structure and parameters to be identified for building thermal dynamics are defined.

[0036] The thermal dynamics model can be defined as follows: Equation 1:

[0037]

[0038] Where M represents the thermal dynamics model, Let represent the derivative of indoor temperature with respect to time at time t, and let a and b represent the linear weighting coefficients of the derivatives of τ(t) and q(t) with respect to temperature, respectively. x(t) = [τ... out (t), q rad (t), q occ [x(t)] represents the set of other influencing factors, and F(x(t)) is a high-dimensional function representing the influence of the factors in x(t) on the temperature derivative. θ (x(t)) is represented using a neural network structure. Therefore, the parameters that need to be identified in thermal dynamics modeling include a, b, and F. θ The parameters of the neural network in (x(t)) are represented as θ.

[0039] Figure 3 The modeling process for a building thermal dynamics model according to an embodiment of the present invention is shown.

[0040] like Figure 3 As shown, parameters a and b are the state variable and control variable, respectively, and θ is the disturbance variable.

[0041] According to an embodiment of the present invention, an optimization problem and an accuracy loss function are established respectively, and then two gradient vectors relative to the model parameters are calculated. The two gradient vectors obtained by collaboration are used to realize the gradient update of the model, and the building thermal dynamic model is obtained through iteration.

[0042] First, the optimization problem of the thermal dynamics model can be expressed by the following equations 2 and 3:

[0043]

[0044] Make:

[0045] τ(t+Δt)=τ(t)+Δt·(aτ(t)+bq(t)+F θ (x(t)))

[0046] q(t)=ηp(t)

[0047]

[0048] Among them, C opt Let T be the cost function, and c be the time period for optimization. t Let p(t) be the electricity price at time t, p(t) be the power consumption of the air conditioner at time t, and c be the power consumption of the air conditioner. u c L These represent the penalty costs for temperatures exceeding their upper and lower limits, respectively. u (t), e L (t) represents the upper and lower temperature values ​​at time t, respectively; Δt is the optimized time interval; and η is the conversion efficiency coefficient between electrical power and heat power. p , These are the upper and lower limits of the electrical power, respectively. t t , These represent the upper and lower limits of the thermal comfort range. This represents the maximum value exceeding the limit. The optimal decision variable is v = [p(t), e] U (t), e L The model parameters to be solved are w = [a, b, θ].

[0049] Based on the optimization problem, the gradient of the optimization problem with respect to the parameters of the thermal dynamic model, i.e., the first gradient, is calculated and denoted as:

[0050] In addition, establish the accuracy loss function L phy Equation 4.

[0051]

[0052] in, w represents the actual temperature value at time t. R R(t) is the weighting coefficient, R(t) is the weighting term at time t, and max{a, 0} is the maximum value comparison function. If a>0, the output is a, otherwise it is 0, and so on.

[0053] Based on the accuracy loss function, its gradient with respect to the parameters of the thermal dynamics model, i.e., the second gradient, is calculated and expressed as:

[0054] Then, a collaborative gradient vector is obtained based on the first and second gradients to update the model parameters, and the building thermal dynamics modeling is iteratively realized.

[0055] Specifically, based on two gradient vectors g opt and g phy Calculate the expected final cooperative gradient vector g. * This is used to update the model parameters, where the final co-gradient vector g * It is calculated as Equation 5.

[0056]

[0057] Where r is the defined radius parameter; To find the optimal combination gradient, and g * Related; ||g phy ‖ respectively and g phy The magnitude of the vector. For the radius parameter r, it refers to the magnitude of the vector g. phy Using r as the center and r as the radius, determine the final cooperative gradient vector g within this circle. *Therefore, the radius parameter r can refer to the final cooperative gradient g. * With the first gradient vector g phy The degree / magnitude of the deviation between them is represented by the "final co-gradient g". * It is in g phy The search was conducted within a region centered at r and with radius r.

[0058] Here, parameters marked with a superscript "*" represent the final solution, that is, the final solution that satisfies both the constraints and minimizes the objective function; parameters without a superscript "*" represent variables, which are potential solutions, representing all feasible solutions that satisfy the constraints.

[0059] The calculation method is to solve the optimal solution to the following optimization problem, i.e., Equation 6.

[0060]

[0061] g com Represented as two gradient vectors g opt and g phy The weighted combination, where w is the weight coefficient.

[0062] Equation 6 above can be understood as, in all g... com In the process, find the unique optimal solution that minimizes the objective condition.

[0063] Obtain the gradient g * Then, the defined model parameters are continuously updated iteratively to finally output the building's thermal dynamics model. The actual performance of the final model in reducing energy costs is evaluated through actual building operation.

[0064] The final model is deployed on existing computing devices, such as servers, personal computers, and smart meters. Combined with data acquisition units such as temperature sensors, relevant data is input and simulation analysis is carried out based on the proposed model to achieve real-time temperature control of the building.

[0065] The features and beneficial effects of this invention are as follows: This invention proposes a novel energy cost-oriented building thermal dynamic modeling technology and equipment, which can fully explore the flexible adjustment potential of building systems in the temperature control process and reduce energy costs.

[0066] Figure 4 A schematic block diagram of an apparatus 400 for modeling a thermal dynamic model according to an embodiment of the present invention is shown.

[0067] like Figure 4As shown, the device 400 includes: a model determination component 401 for determining a thermal dynamic model characterizing changes in indoor temperature; a first gradient determination component 402 for establishing an optimization problem with the goal of minimizing cost and determining a first gradient of the optimization problem with respect to the parameters of the thermal dynamic model; a second gradient determination component 403 for determining a second gradient of the accuracy loss function with respect to the parameters of the thermal dynamic model; a cooperative gradient determination component 404 for determining a cooperative gradient based on the first and second gradients; and an update component 405 for updating the parameters of the thermal dynamic model by iterating the cooperative gradient.

[0068] Therefore, according to the embodiments of the present invention, while ensuring the accuracy of the model, the economic cost of building optimization operation can be effectively reduced, which is conducive to fully exploring the flexible adjustment potential of building thermal dynamics, thereby improving energy efficiency and reducing building energy consumption costs.

[0069] In one example, model determining component 401 can determine a thermal dynamic model based on outdoor temperature data, air conditioning power data, solar radiation power data, indoor heat source power data, and indoor temperature data.

[0070] In one example, the thermal dynamics model is represented as

[0071] Where M represents the thermal dynamics model, Let τ(t) represent the indoor temperature at time t with respect to time, τ(t) represent the indoor temperature data at time t, q(t) represent the air conditioning power data at time t, and x(t) = [τ...]. out (t), q rad (t), q occ [x(t)] represents the set of other influencing factors, F(x(t)) represents the influence of the influencing factors in x(t) on the temperature derivative, θ represents the parameters of the neural network, and

[0072] Where a and b represent the linear weighting coefficients of τ(t) and q(t) with respect to the temperature derivative, respectively. out (t) represents the outdoor temperature data at time t, q rad (t) represents the solar radiation power data at time t, q occ (t) represents the indoor heat source power data corresponding to time t.

[0073] In one example, the input data for the thermal dynamics model includes outdoor temperature data τ. out (t), air conditioning power data q(t), solar radiation power data q rad (t) and indoor heat source power data q occ (t), where the output data of the thermal dynamics model includes indoor temperature data τ(t).

[0074] In one example, the first gradient determining component 402 can determine the gradient g of the optimization problem with respect to the parameters a, b, and θ of the thermal dynamics model. opt .

[0075] In one example, the second gradient determination component 403 can determine the gradient g of the accuracy loss function with respect to the parameters a, b, and θ of the thermal dynamics model. phy .

[0076] In one example, the cooperative gradient is determined as:

[0077]

[0078] Where r is the radius parameter and represents the co-gradient and the second gradient g. phy The degree of deviation between them For the optimal combination gradient, ||g phy ||respectively and g phy The vector magnitude.

[0079] In one example, a thermal dynamics model was deployed on building temperature control equipment.

[0080] Figure 5 A schematic block diagram of an electronic device 500 according to an embodiment of the present invention is shown.

[0081] The electronic device 500 includes: a transceiver 501; and a controller 502 coupled to the transceiver and configured to perform the method according to embodiments of the present disclosure.

[0082] It should be noted that, for clarity and brevity, in Figures 1 to 5 Only the parts relevant to embodiments of the present invention are shown, but those skilled in the art should understand that... Figures 1 to 5 The device or apparatus shown may include other necessary units.

[0083] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0084] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0085] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0086] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0087] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0088] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0089] It should also be noted that, in the apparatus and method of the present invention, it is obvious that the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered equivalent solutions of the present invention. Furthermore, the steps of performing the above series of processes can naturally be performed in the order described, but are not necessarily required to be performed in chronological order. Some steps can be performed in parallel or independently of each other.

[0090] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for modeling thermal dynamics, comprising: Determine a thermal dynamic model characterizing changes in indoor temperature; Establish an optimization problem with the goal of minimizing cost, and determine the first gradient of the optimization problem with respect to the parameters of the thermal dynamics model; Determine the second gradient of the accuracy loss function with respect to the parameters of the thermal dynamics model; Determine the cooperative gradient based on the first and second gradients; and The parameters of the thermal dynamics model are updated by iterating the cooperative gradient.

2. The method according to claim 1, wherein, Determining the thermal dynamic model characterizing indoor temperature changes involves: based on outdoor temperature data, air conditioning power data, solar radiation power data, indoor heat source power data, and indoor temperature data, determining the thermal dynamic model.

3. The method according to claim 2, wherein, The thermal dynamics model is represented as Where M represents the thermal dynamics model, Let τ(t) represent the indoor temperature derivative with respect to time t, τ(t) represent the indoor temperature data at time t, q(t) represent the air conditioning power data at time t, and x(t) = [t out (t),q rad (t),q occ [x(t)] represents the set of other influencing factors, F(x(t)) represents the influence of the influencing factors in x(t) on the temperature derivative, q represents the parameters of the neural network, and Where a and b represent the linear weighting coefficients of the temperature derivatives of t(t) and q(t), respectively, and τ out (t) represents the outdoor temperature data at time t, q rad (t) represents the solar radiation power data at time t, q occ (t) represents the indoor heat source power data corresponding to time t.

4. The method according to claim 3, wherein, The input data for the thermal dynamics model includes outdoor temperature data τ. out (t), air conditioning power data q(t), solar radiation power data q rad (t) and indoor heat source power data q occ (t), and The output data of the thermal dynamics model includes indoor temperature data τ(t).

5. The method according to claim 3, wherein, Determining the first gradient includes: determining the gradient g of the optimization problem with respect to the parameters a, b, and q of the thermal dynamics model. opt .

6. The method according to claim 4, wherein, Determining the second gradient includes: determining the gradient g of the accuracy loss function with respect to the parameters a, b, and q of the thermal dynamics model. phy .

7. The method according to claim 6, wherein, The cooperative gradient was determined as follows: Where r is the radius parameter and represents the co-gradient and the second gradient g. phy The degree of deviation between them For the optimal combination gradient, ||g phy ||respectively and g phy The vector magnitude.

8. The method according to any one of claims 1 to 7, wherein, The thermal dynamics model was deployed on the building's temperature control equipment.

9. An apparatus for modeling thermal dynamics, comprising: Model determination component, used to determine the thermal dynamic model characterizing changes in indoor temperature; The first gradient determination component is used to establish an optimization problem with the goal of minimizing cost and to determine the first gradient of the optimization problem with respect to the parameters of the thermal dynamics model. The second gradient determination component is used to determine the second gradient of the accuracy loss function with respect to the parameters of the thermal dynamics model. A cooperative gradient determination component, used to determine a cooperative gradient based on a first gradient and a second gradient; and Update the component to update the parameters of the thermal dynamics model by iterating the cooperative gradient.

10. An electronic device, comprising: transceiver; as well as A controller, coupled to the transceiver and configured to perform the method of any one of claims 1-8.

Images