Building energy-saving control method and building energy-saving control device

By building HVAC model and simulation environment in the building, combining reinforcement learning model, using BIM and IOT data to optimize HVAC system control, the problems of high energy consumption and poor control effect in the existing technology are solved, and intelligent energy saving and heat saving are achieved A balance of comfort.

CN119065235APending Publication Date: 2024-12-03HITACHI LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202310639043.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-05-31
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

The existing building HVAC system has high energy consumption and is difficult to achieve the optimal energy-saving strategy through simple prior knowledge or linear analytical methods, especially when indoor and outdoor environment changes, the control effect is poor.

Method used

By building the HVAC model and simulation environment of the building, combining the reinforcement learning model, based on indoor and outdoor environment information and the HVAC system equipment module model, optimize the control actions to generate the optimal energy-saving strategy, and use BIM and IOT data fusion to achieve the HVAC system Intelligent control.

Benefits of technology

The balance between energy saving and thermal comfort is achieved, the energy efficiency of the HVAC system is improved, the optimal energy saving strategy is provided, and the indoor and outdoor environment changes are adapted to reduce energy consumption.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119065235A_ABST
    Figure CN119065235A_ABST
Patent Text Reader

Abstract

The invention provides a building energy-saving control method and a building energy-saving control device. The method comprises the following steps: constructing an HVAC model of a target building based on a building room model of the target building and an equipment module model corresponding to each equipment module in the HVAC system; based on the HVAC model of the target building, the outdoor weather information and the indoor dynamic environment information, constructing an HVAC simulation environment of the target building, and building an HVAC simulation model comprising the HVAC simulation environment and a reinforcement learning model; through the HVAC simulation model, simulation results of the target building under different control actions are obtained; based on the simulation result, obtaining optimal control actions in different states and generating an HVAC control optimization model; and controlling operation of a real HVAC system of the target building based on the HVAC control optimization model. According to the invention, energy-saving control can be carried out based on indoor and outdoor environments, and balance between energy saving and thermal comfort is achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of heating, ventilation and air conditioning (HVAC) of buildings, and in particular to a building energy-saving control method and a building energy-saving control device. Background Art

[0002] According to relevant statistics, China's total building operation energy consumption accounts for about 22% of the total social energy consumption, of which the energy consumption of heating, ventilation and air conditioning (HVAC) systems accounts for about 50% of the total building energy consumption. Therefore, energy saving in HVAC systems is of great significance to achieving the dual carbon goals.

[0003] At present, most of the control methods for central air-conditioning systems in commercial buildings are based on timetables or fixed rules. Most of the control methods based on fixed rules rely on prior knowledge. When the indoor and outdoor environments change, the control rules remain unchanged, so the effect is usually not the most energy-efficient. In addition, there are many factors affecting the thermal process of buildings, and the action process is complex. Simple prior knowledge or conventional linear analytical methods are difficult to give the optimal strategy. For example, it is difficult to adjust the energy-saving strategy by sensing the number of people in the building in real time. Summary of the invention

[0004] At least one embodiment of the present application provides a building energy-saving control method and a building energy-saving control device, which can perform energy-saving control based on indoor and outdoor environments and achieve a balance between energy saving and thermal comfort.

[0005] According to one aspect of the present application, at least one embodiment provides a building energy-saving control method, including:

[0006] Constructing an HVAC model of the target building based on a building room model of the target building and equipment module models corresponding to each equipment module in the HVAC system, wherein the HVAC system includes at least one equipment module selected from the group consisting of a water-cooling unit, a cooling tower, a cooling water pump, a chilled water pump, a fresh air unit, a fan coil unit, a gas mixing valve, a liquid mixing valve, and a temperature control valve;

[0007] Based on the HVAC model of the target building, outdoor weather information, and indoor dynamic environment information, construct an HVAC simulation environment of the target building, and build an HVAC simulation model including the HVAC simulation environment and a reinforcement learning model;

[0008] By using the HVAC simulation model, simulation results of the target building under different control actions are obtained; based on the simulation results, optimal control actions under different states are obtained and an HVAC control optimization model is generated;

[0009] The operation of the real HVAC system of the target building is controlled based on the HVAC control optimization model.

[0010] Optionally, before constructing the HVAC model of the target building, the method further includes:

[0011] Generate a BIM model of the target building, wherein the BIM model includes a plurality of building room entities in IFC format;

[0012] The building room entity in the IFC format is converted into a building room entity in the IDF format, and a parameter acquisition interface is set to obtain a building room model of the target building, wherein the parameter acquisition interface is used to obtain at least one of building static information, outdoor weather information and indoor dynamic environment information.

[0013] Optionally, the building static information includes at least one of building wall window material parameters, ventilation permeability coefficient, personnel work and rest schedule, equipment thermal disturbance information, room location information, and heat gain coefficient corresponding to the building room model;

[0014] The outdoor weather information includes at least one of temperature, humidity and light intensity;

[0015] The indoor dynamic environment information includes at least one of indoor temperature, indoor humidity, indoor air volume, carbon dioxide concentration, regional crowd density, thermal comfort and conference room reservation time corresponding to the building room model.

[0016] Optionally, constructing an HVAC model of the target building based on the building room model of the target building and the equipment module models corresponding to each equipment module in the HVAC system includes:

[0017] linking a building room model of the target building and a device module model of the HVAC system according to a flow relationship of a heat transfer medium of the HVAC system;

[0018] A result output module and an interactive interface module are set up, wherein the result output module is used to output outdoor weather information and the operation results of a preset equipment module model, and the interactive interface module is used to receive the control action output by the reinforcement learning model and provide it to the corresponding equipment module model.

[0019] Optionally, obtaining simulation results of the target building under different control actions through the HVAC simulation model includes:

[0020] The reinforcement learning model calculates a reward for a previous time step based on a control action outputted at a previous time step and a state outputted by the HVAC simulation environment at a current time step, and determines a control action for a current time step based on the reward for the previous time step and inputs the control action into the HVAC simulation environment;

[0021] The HVAC simulation environment simulates the energy consumption of the target building based on the control action of the current time step, obtains the state of the next time step and inputs it into the reinforcement learning model, so that the reinforcement learning module calculates the reward of the current time step based on the control action of the current time step and the state of the next time step, and determines the control action of the next time step according to the reward of the current time step and inputs it into the HVAC simulation environment;

[0022] The conditions include indoor temperature, regional crowd density and thermal comfort.

[0023] According to another aspect of the present application, at least one embodiment provides a building energy-saving control device, including:

[0024] A first construction module is used to construct an HVAC model of the target building based on a building room model of the target building and equipment module models corresponding to each equipment module in the HVAC system, wherein the HVAC system includes at least one equipment module selected from the group consisting of a water-cooling unit, a cooling tower, a cooling water pump, a chilled water pump, a fresh air unit, a fan coil unit, a gas mixing valve, a liquid mixing valve, and a temperature control valve;

[0025] A second construction module is used to construct an HVAC simulation environment of the target building based on the HVAC model of the target building, outdoor weather information and indoor dynamic environment information, and to build an HVAC simulation model including the HVAC simulation environment and the reinforcement learning model;

[0026] A simulation module, used to obtain simulation results of the target building under different control actions through the HVAC simulation model; based on the simulation results, obtain the optimal control actions under different states and generate an HVAC control optimization model;

[0027] A control module is used to control the operation of the real HVAC system of the target building based on the HVAC control optimization model.

[0028] Optionally, the above device further comprises:

[0029] A first generating module, configured to generate a BIM model of the target building, wherein the BIM model includes a plurality of building room entities in an IFC format;

[0030] The second generation module is used to convert the building room entity in the IFC format into the building room entity in the IDF format, and set a parameter acquisition interface to obtain the building room model of the target building, wherein the parameter acquisition interface is used to obtain at least one of the building static information, outdoor weather information and indoor dynamic environment information.

[0031] Optionally, the building static information includes at least one of building wall window material parameters, ventilation permeability coefficient, personnel work and rest schedule, equipment thermal disturbance information, room location information, and heat gain coefficient corresponding to the building room model;

[0032] The outdoor weather information includes at least one of temperature, humidity and light intensity;

[0033] The indoor dynamic environment information includes at least one of indoor temperature, indoor humidity, indoor air volume, carbon dioxide concentration, regional crowd density, thermal comfort and conference room reservation time corresponding to the building room model.

[0034] Optionally, the first building module is further used to:

[0035] linking a building room model of the target building and a device module model of the HVAC system according to a flow relationship of a heat transfer medium of the HVAC system;

[0036] A result output module and an interactive interface module are set up, wherein the result output module is used to output outdoor weather information and the operation results of a preset equipment module model, and the interactive interface module is used to receive the control action output by the reinforcement learning model and provide it to the corresponding equipment module model.

[0037] Optionally, the simulation module is further used for:

[0038] The reinforcement learning model calculates a reward for a previous time step based on a control action outputted at a previous time step and a state outputted by the HVAC simulation environment at a current time step, and determines a control action for a current time step based on the reward for the previous time step and inputs the control action into the HVAC simulation environment;

[0039] The HVAC simulation environment simulates the energy consumption of the target building based on the control action of the current time step, obtains the state of the next time step and inputs it into the reinforcement learning model, so that the reinforcement learning module calculates the reward of the current time step based on the control action of the current time step and the state of the next time step, and determines the control action of the next time step according to the reward of the current time step and inputs it into the HVAC simulation environment;

[0040] The conditions include indoor temperature, regional crowd density and thermal comfort.

[0041] According to another aspect of the present application, at least one embodiment provides a computer-readable storage medium, on which a program is stored. When the program is executed by a processor, the steps of the method described above are implemented.

[0042] Compared with the prior art, the building energy-saving control method and the building energy-saving control device provided in the embodiments of the present application realize a building HVAC energy-saving solution based on BIM and IOT, which can quickly integrate BIM and IOT data to build a digital twin HVAC system, and provide important building environment perception data (indoor temperature, regional crowd density and thermal comfort, etc.) for HVAC energy-saving simulation, thereby obtaining the optimal energy-saving strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present application. Also, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0044] Figure 1 A schematic diagram of a flow chart of a building energy-saving control method according to an embodiment of the present application;

[0045] Figure 2 This is an example diagram of the control process of the building HVAC energy-saving system based on BIM and IOT;

[0046] Figure 3 A schematic diagram of the process of converting IFC format data into IDF format data;

[0047] Figure 4 A flow chart example for the HVAC model of central air conditioning;

[0048] Figure 5 Schematic diagram of the HVAC simulation model built;

[0049] Figure 6 Schematic diagram of the interaction process between the HVAC simulation environment and the AI ​​reinforcement learning model;

[0050] Figure 7 An example diagram of the information synchronization process between the HVAC simulation environment and the reinforcement learning model;

[0051] Figure 8 A schematic diagram of a reinforcement learning model;

[0052] Fig. 9 The following is an example diagram of the architecture of the Q network and the target Q network;

[0053] Fig.10 Schematic diagram of the model training process for fan coil unit strategy optimization;

[0054] Fig.11 A schematic diagram of the structure of a building energy-saving control device according to an embodiment of the present application;

[0055] Fig.12 Another structural schematic diagram of the building energy-saving control device according to an embodiment of the present application. DETAILED DESCRIPTION

[0056] The exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided in order to enable a more thorough understanding of the present application and to fully convey the scope of the present application to those skilled in the art.

[0057] It should be understood that the "one embodiment" or "an embodiment" mentioned throughout the specification means that the specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in one embodiment" appearing throughout the specification may not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices. "And / or" in the specification and claims represents at least one of the connected objects.

[0058] In the various embodiments of the present application, it should be understood that the size of the serial numbers of the following processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0059] The following description provides examples and does not limit the scope, applicability, or configuration set forth in the claims. Changes may be made to the functions and arrangements of the elements discussed without departing from the spirit and scope of the present disclosure. Various examples may appropriately omit, replace, or add various procedures or components. For example, the described methods may be performed in an order different from that described, and various steps may be added, omitted, or combined. In addition, features described with reference to certain examples may be combined in other examples.

[0060] The embodiment of the present application provides a building energy-saving control method, which can integrate Building Information Modeling (BIM) and IOT data through IOT data collected by Internet of Things (IOT) devices to build a digital twin HVAC system, providing important building environment perception data (such as indoor temperature, regional crowd density and thermal comfort) for HVAC energy-saving simulation, so as to achieve a balance between energy saving and thermal comfort and obtain the optimal energy-saving strategy. In addition, based on the method provided in the embodiment of the present application, other digital twin scenarios of smart buildings and smart parks can be further expanded to provide more smart services.

[0061] Please refer to Figure 1 , a building energy-saving control method provided in an embodiment of the present application includes:

[0062] Step 11, constructing an HVAC model of the target building based on the building room model of the target building and the equipment module models corresponding to each equipment module in the HVAC system, wherein the HVAC system includes at least one equipment module selected from the group consisting of a water-cooling unit, a cooling tower, a cooling water pump, a chilled water pump, a fresh air unit, a fan coil, a gas mixing valve, a liquid mixing valve and a temperature control valve.

[0063] Here, the target building usually includes multiple building rooms, and the embodiment of the present application pre-establishes a building room model corresponding to each building room. Then, according to the flow relationship of the heat transfer medium (such as refrigerant and / or heat medium, etc.) of the HVAC system, the building room model of the target building and the equipment module model of the HVAC system are linked. In addition, a result output module and an interactive interface module are set, wherein the result output module is used to output outdoor weather information and the operation results of the preset equipment module model, and the interactive interface module is used to receive the control action output by the reinforcement learning model and provide it to the corresponding equipment module model, thereby controlling the operation of the corresponding equipment module model.

[0064] Step 12: constructing an HVAC simulation environment of the target building based on the HVAC model of the target building, outdoor weather information, and indoor dynamic environment information, and building an HVAC simulation model including the HVAC simulation environment and a reinforcement learning model.

[0065] Step 13: obtaining simulation results of the target building under different control actions through the HVAC simulation model; based on the simulation results, obtaining optimal control actions under different states and generating an HVAC control optimization model.

[0066] Here, the reinforcement learning model calculates the reward of the previous time step based on the control action outputted at the previous time step and the state of the HVAC simulation environment outputted at the current time step, and determines the control action of the current time step according to the reward of the previous time step and inputs it into the HVAC simulation environment;

[0067] The HVAC simulation environment simulates the energy consumption of the target building based on the control action of the current time step, obtains the state of the next time step and inputs it into the reinforcement learning model, so that the reinforcement learning module calculates the reward of the current time step based on the control action of the current time step and the state of the next time step, and determines the control action of the next time step according to the reward of the current time step and inputs it into the HVAC simulation environment. Here, the previous time step refers to the previous time step of the current time step, and the next time step refers to the next time step of the current time step.

[0068] Among them, the state includes indoor temperature, regional crowd density and thermal comfort. These parameters can be collected by IOT devices. In the embodiment of the present application, the state is determined based on the three parameters of indoor temperature, regional crowd density and thermal comfort in each building room of the target building. For example, the values ​​of the above three parameters of each building room are formed into an array, which corresponds to the state, and different arrays correspond to different states. For ease of implementation, multiple corresponding numerical intervals can be set for each parameter; based on the numerical intervals to which the values ​​of the above three parameters of each building room belong, an array is formed, which corresponds to the state, so that the total number of states can be reduced. In this way, when the first state is different from the second state, it means that between the first state and the second state, there is at least one building room with a different numerical interval to which at least one parameter belongs. If the first state is the same as the second state, the numerical interval to which any parameter in any building room in the first state belongs is the same as the numerical interval to which any parameter in any building room in the second state belongs.

[0069] In step 13, the embodiment of the present application obtains the optimal control action under different states based on the simulation results of the target building under different control actions, thereby generating an HVAC control optimization model. The HVAC control optimization model is used to determine the current control action based on the state of the target building to control the operation of related equipment modules.

[0070] Step 14: Control the operation of the real HVAC system of the target building based on the HVAC control optimization model.

[0071] Here, the embodiment of the present application can control the operation of the real HVAC system based on the HVAC control optimization model. For example, the state of the target building is input into the HVAC control optimization model. The control action output by the HVAC control optimization model is obtained and provided to the building automation control system (Building Automation System, BAS), thereby controlling the operation of the HVAC system of the target building.

[0072] Through the above steps, the embodiment of the present application can control the operation of the HVAC system of the target building based on predefined conditions (indoor temperature, regional crowd density and thermal comfort) and outdoor weather information, and achieve a balance between energy saving and thermal comfort.

[0073] Before the above step 11, the embodiment of the present application can also obtain the building room model of the target building in the following manner. Specifically, a BIM model of the target building is first generated, and the BIM model includes a plurality of building room entities in the Industry Foundation Classes (IFC) format. Then, the building room entity in the IFC format is converted into a building room entity in the Instrument Definition File (IDF) format using a two-dimensional text file with a defined data structure, and a parameter acquisition interface is set to obtain the building room model of the target building, wherein the parameter acquisition interface is used to obtain at least one of the building static information, outdoor weather information, and indoor dynamic environment information.

[0074] Here, the building static information includes at least one of the building wall window material parameters, ventilation permeability coefficient, personnel work and rest schedule, equipment thermal disturbance, room location information, and heat gain coefficient corresponding to the building room model. The outdoor weather information includes at least one of temperature, humidity, and light intensity. The indoor dynamic environment information includes at least one of the indoor temperature, indoor humidity, indoor air volume, carbon dioxide concentration, regional crowd density (which can be represented by the number of people in the building room area), thermal comfort, and conference room reservation time (used to represent the use time of the building room model as a conference room) corresponding to the building room model.

[0075] To help better understand the above embodiments, the above method is described below through a more specific example. The following example is described using an HVAC system of a central air conditioner based on a water-cooled unit as an example. It should be noted that the embodiments of the present application are not limited to the above scenarios, but can also be applied to HVAC systems based on multi-split units, etc.

[0076] Please refer to Figure 2 , Figure 2 This example provides the control process of the building HVAC energy-saving system based on BIM and IOT, including:

[0077] A) Low-code platforms

[0078] The low-code platform allows users to quickly customize BIM models, diagrams, and layouts. Users can use simple configuration to determine objects that need to be managed, such as building structures, hierarchical structures, and digital twin devices, as well as campus building structures composed of buildings, to improve the convenience of interaction in 3D scenes.

[0079] B) Simulation model conversion

[0080] When the building energy consumption model uses static building information (geometric information), the common method is to use related tools (such as SketchUp) to build a 3D building model to obtain the building's geometric information. This method is labor-intensive and inaccurate. This example uses the program's automatic conversion method to convert the original 3D geometric information in the building digital twin into a format that can be recognized by the building energy consumption model, saving man-hours and improving efficiency and accuracy.

[0081] If the data format in the digital twin is to be converted into a data format that can be recognized by the HVAC simulation model, it needs to go through the simulation model data conversion. At present, the IFC data standard has become the mainstream standard for expressing BIM data in digital twins, so this example uses the IFC data format to express the BIM model in the digital twin.

[0082] At present, mainstream building energy consumption software such as Energy-plus or Trnsys all use the IDF format. In order to enable the IFC data in the digital twin environment to be used by the energy consumption software, this example provides a set of IFC format data to IDF format data conversion process, refer to Figure 3 .

[0083] Figure 3 This is the process of converting IFC format data into IDF format data. The IFC entity based on the IFC format is processed by the process in the dotted box and output as an IDF entity in IDF format. Specifically:

[0084] 301: Get the relationship between components and spaces in IFC entities;

[0085] 302: Calculate the geometric information of the component;

[0086] 303: Calculate point set information of related components;

[0087] 304: Obtain corresponding building components;

[0088] 305: Get component name;

[0089] 306: The cutting point set becomes an element of the component;

[0090] 307: Match component elements with component names and assemble them into an integral component;

[0091] 308: Define the ZONE of each room;

[0092] 309: Check whether adjacent points, lines and surfaces are coupled when drawing the geometric model;

[0093] 310: Model output is IDF entity.

[0094] C) Simulation model parameter setting

[0095] Through the simulation model parameter setting, set the interface so that the simulation model of the relevant module can obtain outdoor weather information, building static information and indoor dynamic environment information. The model converted from the above format is a geometric model with only geometric parameters. The following parameters need to be set through the setting interface:

[0096] (1) Static building information: building wall and window material parameters, ventilation permeability coefficient, personnel work and rest schedule, equipment thermal disturbance information, room location information and heat gain coefficient, etc.

[0097] (2) Outdoor weather information: temperature, humidity, light intensity, etc.

[0098] (3) Indoor dynamic environment information: indoor temperature, indoor humidity, indoor air volume, carbon dioxide concentration, regional population density, thermal comfort, and meeting room reservation time, etc. The number of people in the area can be detected through WIFI. For details, please refer to the Chinese patent application with publication number CN115134843A.

[0099] D) HVAC model building

[0100] Take the HVAC of central air conditioner based on water cooling unit as an example to build HVAC model. The construction process of HVAC model based on multi-split is similar to that of water cooling unit, but the components are slightly different. Figure 4 This is an example of a process for building an HVAC model for central air conditioning.

[0101] The equipment modules of the HVAC system include: water-cooled units, cooling towers, cooling water pumps, chilled water pumps, fresh air units, fan coils, gas mixing valves, liquid mixing valves, temperature control valves, and building rooms. The building rooms are the content of the simulation model parameter settings in C). Building rooms, fresh air units, and cooling towers need to read outdoor weather information.

[0102] The module link includes: mixing the air processed by the fan coil unit and the air processed by the fresh air unit and delivering them to the building room, that is, inputting the air processed by the fan coil unit and the air processed by the fresh air unit into the gas mixing valve, and connecting the output of the gas mixing valve to the input of the building room.

[0103] Result output module: The result output module is used to output outdoor weather information and the operation results of the preset equipment module model, for example, output files of outdoor weather information, fan coil operation result files, and water chiller operation result files, etc.

[0104] Interactive interface module: used to receive the control actions output by the reinforcement learning model and provide them to the corresponding equipment module model, for example, to control the parameters of the fan coil unit and the water chiller, and to realize the connection between the reinforcement learning algorithm and the simulation system.

[0105] E) Simulation model construction

[0106] Figure 5 This is a schematic diagram of the HVAC simulation model built in this example, which mainly includes the HVAC simulation environment and the AI ​​reinforcement learning model. The HVAC simulation environment mainly includes the following three types of data:

[0107] (1) Building Room

[0108] Static information of building rooms, including building wall and window material parameters, ventilation permeability coefficient, personnel work and rest schedule, equipment thermal disturbance, room location information, heat gain coefficient, etc., is transmitted to the simulation environment through the interface file.

[0109] (2) HVAC equipment system

[0110] The HVAC equipment system includes information on chillers, cooling towers, cooling water pumps, refrigeration water pumps, fresh air units, fan coils, gas mixing valves, liquid mixing valves and temperature control valves.

[0111] (3) Indoor dynamic environment information

[0112] Indoor dynamic environmental information includes regional crowd density, thermal comfort, indoor temperature and humidity, carbon dioxide concentration, and conference room reservation time.

[0113] Figure 6 The following is a schematic diagram of the interaction process between the HVAC simulation environment and the AI ​​reinforcement learning model. Because there is data exchange between the HVAC simulation environment and the reinforcement learning model, this example uses a file-based data exchange method. The HVAC simulation environment output file is the simulation result of the HVAC, and the operation control command input file is the control action output by the AI ​​reinforcement learning model. Data interaction is achieved in this way. Input file reading is responsible for reading the operation control command and passing the data into the simulation environment.

[0114] When exchanging data, it is necessary to ensure that the HVAC simulation environment and the AI ​​reinforcement learning model are synchronized, otherwise there will be a problem of mismatch in file reading content. This example uses file locks to achieve inter-process communication and data synchronization.

[0115] The information synchronization process between the HVAC simulation environment and the reinforcement learning model is as follows: Figure 7 As shown. Within a time step, the HVAC simulation environment simulates the building environment and system operation energy consumption according to the control action given by the reinforcement learning model at the previous moment, and inputs the simulation results into the reinforcement learning model. The reinforcement learning model calculates the control actions such as the fan speed and chilled water outlet set temperature at the next moment based on the room temperature and humidity, number of people, comfort, Co2 concentration, outdoor temperature and humidity, number of chillers, chilled water return temperature, and cold station state parameters at the current and historical moments, and then enters the simulation at the next moment. This cycle can obtain the simulation results of the building environment and system operation under the given control strategy throughout the year.

[0116] F) Reinforcement Learning Model

[0117] Reinforcement learning is a kind of unlabeled learning. Its core idea is that the intelligent agent interacts with the environment continuously and conducts trial and error, and optimizes the decision through the feedback of the environment. Specifically, the reward function is used to judge whether an action is good or bad under a certain state. The learning process is to change the strategy of executing the action through the reward signal, and the final result is to form a strategy that maximizes the reward.

[0118] Figure 8 is a schematic diagram of a reinforcement learning model. A is the action set, S is the state set, and R is the reward calculated from action A and state S. At each decision time t, the agent chooses to perform action a t (a t ∈A), after a time step Δt, the environment changes from state s t becomes s t+1 At time t+1, the agent calculates the reward R for this time step according to the defined reward function t .

[0119] The goal of reinforcement learning is to obtain a state-related policy function a=f(s), that is, to output actions according to the state, and to maximize the cumulative reward by executing the policy function. This policy function is the optimal control strategy learned by the agent. The calculation of the cumulative reward can be in the form of γ-discounted cumulative reward. The specific method for calculating the cumulative reward value Total Reward is shown in Formula 1.

[0120]

[0121] Among them, E is the mathematical expectation.

[0122] The Deep Q-network (DQN) algorithm is often used to fit the Q value (cumulative reward value) in reinforcement learning. However, the traditional DQN has two limitations: 1. The memory limits the storage of experience data. 2. It needs to perceive complete observation information. The reinforcement learning algorithm in this example uses DRQN, and its iterative object is the value function corresponding to the state-action pair, that is, the Q value function, represented by Q(s,a), which means the cumulative reward value that the system will obtain when performing action a in state s. The Q-Learning algorithm stores the Q values ​​corresponding to a finite number of value pairs (state & action) in the neural network and uses the DRQN ​​network to fit the optimal Q value. Fig. 9 A schematic diagram of the neural network architecture in DRQN ​​is provided.

[0123] G) Model training process

[0124] In this example, there are two main optimization objectives for the HVAC system, namely, the strategy optimization of the fan coil unit and the strategy optimization of the outlet water temperature of the water chiller. The following takes the strategy optimization of the fan coil unit as an example to illustrate the model training process. The outlet water temperature of the water chiller is similar to the strategy optimization of the fan coil unit, so it will not be repeated here to save space. Fig.10 This is the model training process for fan coil unit strategy optimization, including:

[0125] 101: Action definition: Assume that the gears of the fan coil are divided into off, low, medium and high. The low gear opens 50% of the rated air volume, the medium gear opens 75% of the rated air volume, and the high gear opens the rated air volume. The relationship between the fan coil gear and the control command is shown in Table 1.

[0126] Table 1

[0127] Gear action Order upscale 3 1 Mid-range 2 0.75 Low Range 1 0.5 closure 0 0

[0128] 102: State definition: This example uses indoor air temperature, regional crowd density and thermal comfort as states.

[0129] 103: The definition of the reward function is shown in Formula 2

[0130]

[0131] Among them, a t Represents the energy consumption item, for example, when the high gear is turned on, a t =3, when low gear is turned on, a t =1, such a setting can make the training of the agent move towards energy saving. i is a penalty term. The specific form is shown in formulas 3 to 5. When the indoor air temperature, regional crowd density and thermal comfort exceed the upper or lower limit, the agent will be punished. i It is the weight, and the corresponding value is set according to the importance of the impact of population density and thermal comfort in the air temperature area on energy consumption.

[0132]

[0133] In formula 3, T t Represents the real-time temperature of the area, T lower bound =25℃, T upper bound =27℃.

[0134]

[0135] In formula 4, C t Represents the real-time crowd density of the area.

[0136] C lower bound =0.8people / m 2 C upper bound =3people / m 2 .

[0137]

[0138] In formula 5, S t Represents the regional thermal comfort, S lower bound =-1, S upper bound=1.

[0139] 104: Building Q network: Building Q network and target Q network

[0140] Q network and target Q network architecture reference Fig. 9 This example uses a DRQN ​​network, which includes one LSTM layer and two hidden layers. The number of neurons in each layer is 128, 64, and 32 respectively. The activation function of the first two layers is RELU, and the last layer uses a linear function as the activation function.

[0141] 105: Setting the Exploration Strategy: Exploring the Actions Selected by the Agent

[0142] This example uses the ε-greedy exploration strategy. Let ε0 = 0.2, and ε decreases by 0.0003 for each step. Each time an action is selected, a random number is generated. If the random number is less than ε, the agent randomly selects an action, otherwise the agent uses the Q network to select an action.

[0143] 106: Setting Network Goals and Update Strategies

[0144] When the experience pool is full, the average value of the reward within the latest 100 steps is taken. If the average value is greater than the pre-set critical value, the existing model will continue to be used for control, otherwise the Q network will be trained. In addition, every 100 times the Q network is trained, its parameters are copied to the target Q network.

[0145] 107: Setting network hyperparameters

[0146] DRQN network hyperparameters include learning rate Lr, penalty item weight Wi, discount factor γ, experience pool capacity, pre-trained network update frequency, batch size, etc.

[0147] H) HVACHVAC control optimization model controls the real HVAC system

[0148] After the model training process, the optimal HVAC control optimization model is obtained, which can output the best fan coil control strategy and chiller outlet water temperature control strategy to the real HVAC control system according to different external conditions, so as to achieve the purpose of energy saving.

[0149] Please refer to Fig.11 , a structure of a building energy-saving control device provided in an embodiment of the present application includes:

[0150] A first construction module 111 is used to construct an HVAC model of the target building based on a building room model of the target building and equipment module models corresponding to each equipment module in the HVAC system, wherein the HVAC system includes at least one equipment module selected from the group consisting of a water-cooling unit, a cooling tower, a cooling water pump, a chilled water pump, a fresh air unit, a fan coil unit, a gas mixing valve, a liquid mixing valve, and a temperature control valve;

[0151] A second construction module 112 is used to construct an HVAC simulation environment of the target building based on the HVAC model of the target building, outdoor weather information and indoor dynamic environment information, and to build an HVAC simulation model including the HVAC simulation environment and the reinforcement learning model;

[0152] The simulation module 113 is used to obtain simulation results of the target building under different control actions through the HVAC simulation model; based on the simulation results, obtain the optimal control actions under different states and generate an HVAC control optimization model;

[0153] The control module 114 is used to control the operation of the real HVAC system of the target building based on the HVAC control optimization model.

[0154] Through the above modules, the embodiment of the present application can perform energy-saving control based on the indoor and outdoor environments, and achieve a balance between energy saving and thermal comfort.

[0155] Optionally, the above device further comprises:

[0156] A first generating module, configured to generate a BIM model of the target building, wherein the BIM model includes a plurality of building room entities in an IFC format;

[0157] The second generation module is used to convert the building room entity in the IFC format into the building room entity in the IDF format, and set a parameter acquisition interface to obtain the building room model of the target building, wherein the parameter acquisition interface is used to obtain at least one of the building static information, outdoor weather information and indoor dynamic environment information.

[0158] Optionally, the building static information includes at least one of building wall window material parameters, ventilation permeability coefficient, personnel work and rest schedule, equipment thermal disturbance information, room location information, and heat gain coefficient corresponding to the building room model;

[0159] The outdoor weather information includes at least one of temperature, humidity and light intensity;

[0160] The indoor dynamic environment information includes at least one of indoor temperature, indoor humidity, indoor air volume, carbon dioxide concentration, regional crowd density, thermal comfort and conference room reservation time corresponding to the building room model.

[0161] Optionally, the first building module 111 is further used to:

[0162] linking a building room model of the target building and a device module model of the HVAC system according to a flow relationship of a heat transfer medium of the HVAC system;

[0163] A result output module and an interactive interface module are set up, wherein the result output module is used to output outdoor weather information and the operation results of a preset equipment module model, and the interactive interface module is used to receive the control action output by the reinforcement learning model and provide it to the corresponding equipment module model.

[0164] Optionally, the simulation module 113 is further used for:

[0165] The reinforcement learning model calculates a reward for a previous time step based on a control action outputted at a previous time step and a state outputted by the HVAC simulation environment at a current time step, and determines a control action for a current time step based on the reward for the previous time step and inputs the control action into the HVAC simulation environment;

[0166] The HVAC simulation environment simulates the energy consumption of the target building based on the control action of the current time step, obtains the state of the next time step and inputs it into the reinforcement learning model, so that the reinforcement learning module calculates the reward of the current time step based on the control action of the current time step and the state of the next time step, and determines the control action of the next time step according to the reward of the current time step and inputs it into the HVAC simulation environment;

[0167] The conditions include indoor temperature, regional crowd density and thermal comfort.

[0168] It should be noted that the various systems provided in the above embodiments are devices corresponding to the above building energy-saving control method, and the implementation methods in the above embodiments are applicable to the embodiments of the device, and can also achieve the same technical effects. The above device provided in the embodiment of the present application can implement all the method steps implemented in the above method embodiment, and can achieve the same technical effects. The parts and beneficial effects that are the same as those in the method embodiment will not be specifically described here.

[0169] Please refer to Fig.12 , a schematic diagram of the structure of another building energy-saving control device provided in an embodiment of the present application, the device includes: a processor 1201, a transceiver 1202, a memory 1203, a user interface 1204 and a bus interface.

[0170] In the embodiment of the present application, the device also includes: a program stored in the memory 1203 and executable on the processor 1201 .

[0171] The transceiver 1202 is used to send and receive data under the control of the processor;

[0172] The processor 1201 is configured to read the computer program in the memory and perform the following operations:

[0173] Constructing an HVAC model of the target building based on a building room model of the target building and equipment module models corresponding to each equipment module in the HVAC system, wherein the HVAC system includes at least one equipment module selected from the group consisting of a water-cooling unit, a cooling tower, a cooling water pump, a chilled water pump, a fresh air unit, a fan coil unit, a gas mixing valve, a liquid mixing valve, and a temperature control valve;

[0174] Based on the HVAC model of the target building, outdoor weather information, and indoor dynamic environment information, construct an HVAC simulation environment of the target building, and build an HVAC simulation model including the HVAC simulation environment and a reinforcement learning model;

[0175] By using the HVAC simulation model, simulation results of the target building under different control actions are obtained; based on the simulation results, optimal control actions under different states are obtained and an HVAC control optimization model is generated;

[0176] The operation of the real HVAC system of the target building is controlled based on the HVAC control optimization model.

[0177] It can be understood that in the embodiment of the present application, when the computer program is executed by the processor 1201, the various processes of the above-mentioned building energy-saving control method embodiment can be implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0178] exist Fig.12 In the embodiment, the bus architecture may include any number of interconnected buses and bridges, specifically one or more processors represented by processor 1201 and various circuits of memory represented by memory 1203 are linked together. The bus architecture may also link together various other circuits such as peripherals, voltage regulators, and power management circuits, which are well known in the art and are therefore not further described herein. The bus interface provides an interface. The transceiver 1202 may be a plurality of components, namely, a transmitter and a receiver, providing a unit for communicating with various other devices on a transmission medium. For different user devices, the user interface 1204 may also be an interface capable of externally and internally connecting required devices, and the connected devices include but are not limited to a keypad, a display, a speaker, a microphone, a joystick, and the like.

[0179] The processor 1201 is responsible for managing the bus architecture and general processing, and the memory 1203 can store data used by the processor 1201 when performing operations.

[0180] It should be noted that the device in this embodiment is a device corresponding to the above-mentioned building energy-saving control method, and the implementation methods in the above-mentioned embodiments are all applicable to the embodiments of the device, and can also achieve the same technical effects. In the device, the transceiver 1202 and the memory 1203, as well as the transceiver 1202 and the processor 1201 can be connected through a bus interface communication, and the function of the processor 1201 can also be implemented by the transceiver 1202, and the function of the transceiver 1202 can also be implemented by the processor 1201. It should be noted that the above-mentioned device provided in the embodiment of the present application can implement all the method steps implemented in the above-mentioned method embodiment, and can achieve the same technical effects, and the parts and beneficial effects that are the same as the method embodiment in this embodiment will not be specifically repeated here.

[0181] In some embodiments of the present application, a computer-readable storage medium is further provided, on which a program is stored, and when the program is executed by a processor, the following steps are implemented:

[0182] Constructing an HVAC model of the target building based on a building room model of the target building and equipment module models corresponding to each equipment module in the HVAC system, wherein the HVAC system includes at least one equipment module selected from the group consisting of a water-cooling unit, a cooling tower, a cooling water pump, a chilled water pump, a fresh air unit, a fan coil unit, a gas mixing valve, a liquid mixing valve, and a temperature control valve;

[0183] Based on the HVAC model of the target building, outdoor weather information, and indoor dynamic environment information, construct an HVAC simulation environment of the target building, and build an HVAC simulation model including the HVAC simulation environment and a reinforcement learning model;

[0184] By using the HVAC simulation model, simulation results of the target building under different control actions are obtained; based on the simulation results, optimal control actions under different states are obtained and an HVAC control optimization model is generated;

[0185] The operation of the real HVAC system of the target building is controlled based on the HVAC control optimization model.

[0186] When the program is executed by the processor, it can implement all the implementation methods of the above-mentioned building energy-saving control method and achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0187] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0188] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0189] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0190] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0191] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0192] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard drives, ROM, RAM, magnetic disks, or optical disks.

[0193] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A building energy-saving control method, characterized in that: include: Constructing an HVAC model of the target building based on a building room model of the target building and equipment module models corresponding to each equipment module in the HVAC system, wherein the HVAC system includes at least one equipment module selected from the group consisting of a water-cooling unit, a cooling tower, a cooling water pump, a chilled water pump, a fresh air unit, a fan coil unit, a gas mixing valve, a liquid mixing valve, and a temperature control valve; Based on the HVAC model of the target building, outdoor weather information, and indoor dynamic environment information, construct an HVAC simulation environment of the target building, and build an HVAC simulation model including the HVAC simulation environment and a reinforcement learning model; Obtaining simulation results of the target building under different control actions through the HVAC simulation model; Based on the simulation results, obtaining optimal control actions under different states and generating an HVAC control optimization model; The operation of the real HVAC system of the target building is controlled based on the HVAC control optimization model.

2. The method according to claim 1, characterized in that Before constructing the HVAC model of the target building, the method further includes: Generate a BIM model of the target building, wherein the BIM model includes a plurality of building room entities in IFC format; The building room entity in the IFC format is converted into a building room entity in the IDF format, and a parameter acquisition interface is set to obtain a building room model of the target building, wherein the parameter acquisition interface is used to obtain at least one of building static information, outdoor weather information and indoor dynamic environment information.

3. The method according to claim 2, characterized in that The building static information includes at least one of building wall window material parameters, ventilation permeability coefficient, personnel work and rest schedule, equipment thermal disturbance information, room location information, and heat gain coefficient corresponding to the building room model; The outdoor weather information includes at least one of temperature, humidity and light intensity; The indoor dynamic environment information includes at least one of indoor temperature, indoor humidity, indoor air volume, carbon dioxide concentration, regional crowd density, thermal comfort and conference room reservation time corresponding to the building room model.

4. The method according to claim 1, characterized in that Based on the building room model of the target building and the equipment module models corresponding to each equipment module in the HVAC system, an HVAC model of the target building is constructed, including: linking a building room model of the target building and a device module model of the HVAC system according to a flow relationship of a heat transfer medium of the HVAC system; A result output module and an interactive interface module are set up, wherein the result output module is used to output outdoor weather information and the operation results of a preset equipment module model, and the interactive interface module is used to receive the control action output by the reinforcement learning model and provide it to the corresponding equipment module model.

5. The method according to claim 1, characterized in that Through the HVAC simulation model, simulation results of the target building under different control actions are obtained, including: The reinforcement learning model calculates a reward for a previous time step based on a control action outputted at a previous time step and a state outputted by the HVAC simulation environment at a current time step, and determines a control action for a current time step based on the reward for the previous time step and inputs the control action into the HVAC simulation environment; The HVAC simulation environment simulates the energy consumption of the target building based on the control action of the current time step, obtains the state of the next time step and inputs it into the reinforcement learning model, so that the reinforcement learning module calculates the reward of the current time step based on the control action of the current time step and the state of the next time step, and determines the control action of the next time step according to the reward of the current time step and inputs it into the HVAC simulation environment; The conditions include indoor temperature, regional crowd density and thermal comfort.

6. A building energy-saving control device, characterized in that: include: A first construction module is used to construct an HVAC model of the target building based on a building room model of the target building and equipment module models corresponding to each equipment module in the HVAC system, wherein the HVAC system includes at least one equipment module selected from the group consisting of a water-cooling unit, a cooling tower, a cooling water pump, a chilled water pump, a fresh air unit, a fan coil unit, a gas mixing valve, a liquid mixing valve, and a temperature control valve; A second construction module is used to construct an HVAC simulation environment of the target building based on the HVAC model of the target building, outdoor weather information and indoor dynamic environment information, and to build an HVAC simulation model including the HVAC simulation environment and the reinforcement learning model; A simulation module, used to obtain simulation results of the target building under different control actions through the HVAC simulation model; Based on the simulation results, obtaining optimal control actions under different states and generating an HVAC control optimization model; A control module is used to control the operation of the real HVAC system of the target building based on the HVAC control optimization model.

7. The device according to claim 6, characterized in that Also includes: A first generating module, configured to generate a BIM model of the target building, wherein the BIM model includes a plurality of building room entities in an IFC format; The second generation module is used to convert the building room entity in the IFC format into the building room entity in the IDF format, and set a parameter acquisition interface to obtain the building room model of the target building, wherein the parameter acquisition interface is used to obtain at least one of the building static information, outdoor weather information and indoor dynamic environment information.

8. The device according to claim 7, characterized in that The building static information includes at least one of building wall window material parameters, ventilation permeability coefficient, personnel work and rest schedule, equipment thermal disturbance information, room location information, and heat gain coefficient corresponding to the building room model; The outdoor weather information includes at least one of temperature, humidity and light intensity; The indoor dynamic environment information includes at least one of indoor temperature, indoor humidity, indoor air volume, carbon dioxide concentration, regional crowd density, thermal comfort and conference room reservation time corresponding to the building room model.

9. The device according to claim 6, characterized in that The first building block is further used for: linking a building room model of the target building and a device module model of the HVAC system according to a flow relationship of a heat transfer medium of the HVAC system; A result output module and an interactive interface module are set up, wherein the result output module is used to output outdoor weather information and the operation results of a preset equipment module model, and the interactive interface module is used to receive the control action output by the reinforcement learning model and provide it to the corresponding equipment module model.

10. The device according to claim 6, characterized in that The simulation module is also used for: The reinforcement learning model calculates a reward for a previous time step based on a control action outputted at a previous time step and a state outputted by the HVAC simulation environment at a current time step, and determines a control action for a current time step based on the reward for the previous time step and inputs the control action into the HVAC simulation environment; The HVAC simulation environment simulates the energy consumption of the target building based on the control action of the current time step, obtains the state of the next time step and inputs it into the reinforcement learning model, so that the reinforcement learning module calculates the reward of the current time step based on the control action of the current time step and the state of the next time step, and determines the control action of the next time step according to the reward of the current time step and inputs it into the HVAC simulation environment; The conditions include indoor temperature, regional crowd density and thermal comfort.

Citation Information

Cited By

  • Indoor ecological environment design method and system based on digital twinning and storage medium

    CN120103727A

  • Multi-area air conditioner control method and system based on iterative simulation and fine-tuning large model

    CN122170510A