Green energy-saving building optimization design method and system based on BIM and storage medium

Abstract

The invention relates to the technical field of building optimization analysis, in particular to a BIM-based green energy-saving building optimization design method and system and a storage medium, and the method comprises the following steps: collecting user behavior data in a building through an intelligent device, including use data of each region in the building, hot water use data, electric appliance use frequency and time, and temperature setting preference; based on the user behavior data, establishing a user behavior prediction model by using a machine learning algorithm, and predicting the activity intensity of the user in each area in the building; dynamically adjusting an energy distribution scheme of the building according to a prediction result of the user behavior prediction model; and inputting the activity intensity yi, t of the user behavior prediction model and the real-time environment data into the BIM model, and updating the BIM model of the building. According to the invention, the comfort level of the user is improved, the energy use efficiency is also remarkably improved, and the waste of resources is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of building optimization analysis, and in particular to a green energy-saving building optimization design method, system and storage medium based on BIM. Background Art

[0002] With the progress of society and economic development, the proportion of building energy consumption in total energy consumption is increasing, and the demand for green and energy-saving buildings is also growing. Building energy management and optimization design have become an important part of energy conservation and emission reduction. Traditional building energy management systems usually rely on fixed schedules and preset control strategies, which are difficult to respond to user behavior and environmental changes in real time, resulting in energy waste and inefficient management. In addition, traditional systems have deficiencies in data integration and intelligent analysis, and cannot make full use of historical data and real-time data for dynamic adjustment and future optimization.

[0003] In recent years, the application of Building Information Modeling (BIM) technology has provided a powerful tool for building management and design. BIM supports data management and decision-making throughout the building life cycle by digitally representing the physical and functional characteristics of the building. However, traditional BIM models are usually based on static data and are difficult to update and adjust dynamically in real time, which limits their potential for application in smart building energy management.

[0004] In order to solve the above problems, the existing technology has proposed an intelligent building management system that combines sensor networks, the Internet of Things and big data analysis, and uses real-time data for energy management and optimization design. These systems have improved the energy efficiency of buildings to a certain extent, but there are still some limitations. For example, the integration and analysis of user behavior data and environmental data are insufficient, and it is impossible to accurately predict user needs and dynamically adjust energy distribution; equipment status monitoring and management are not efficient enough, and it is difficult to respond to and handle abnormal situations in a timely manner; there is a lack of intelligent scene prediction and optimization solutions, which makes it difficult to comprehensively improve the green energy-saving effect of buildings.

[0005] In view of these technical difficulties, the present invention proposes a green energy-saving building optimization design method based on BIM. Summary of the invention

[0006] Based on the above objectives, the present invention provides a green energy-saving building optimization design method, system and storage medium based on BIM.

[0007] A green energy-saving building optimization design method based on BIM includes the following steps:

[0008] S100, user behavior data collection: collect user behavior data in the building through smart devices, including usage data of each area in the building, hot water usage data, frequency and time of appliance usage (usage of lighting, air conditioning, television, etc.), and temperature setting preferences (users' temperature setting preferences for different time periods and different areas);

[0009] S200, intelligent scene prediction: Based on user behavior data, a user behavior prediction model is established using machine learning algorithms to predict the activity intensity of users in various areas within the building. i,t ;

[0010] S300, Adaptive Energy Management: Dynamically adjust the building's energy allocation plan based on the prediction results of the user behavior prediction model, including dynamic adjustment of the lighting system, dynamic adjustment of the air conditioning system, and dynamic adjustment of the hot water system;

[0011] S400, BIM model dynamic update: the activity intensity of the user behavior prediction model y i,t And real-time environmental data is input into the BIM (Building Information Model) model to update the BIM model of the building, including energy consumption, equipment status and environmental parameters. Through the visualization and data analysis functions of the BIM model, a comprehensive understanding of the operating status of the building can be achieved.

[0012] Optionally, the user behavior data collection in S100 specifically includes:

[0013] S110, install smart sensor devices in various areas of the building to monitor the usage of each area, including door and window sensors and infrared motion sensors to record the frequency and time of use of each area;

[0014] S120, install smart water meters and smart faucets to record the hot water usage in various areas of the building, including water consumption and usage time;

[0015] S130, through smart meters, records the usage frequency and time of each electrical device, including lighting equipment, air conditioners, and office equipment;

[0016] S140, through the intelligent thermostat, records the user's temperature setting preferences in different time periods and different areas;

[0017] S150, the collected data is transmitted to the central data processor via a wireless network for storage and analysis, and the collected data is pre-processed in the central data processor, including data cleaning, denoising and normalization, so as to be used for subsequent intelligent scene prediction and adaptive energy management.

[0018] Optionally, the intelligent scene prediction in S200 specifically includes:

[0019] S210, obtaining pre-processed user behavior data, and performing feature engineering to extract behavior features, where the behavior features include usage frequency, time distribution, and spatial distribution;

[0020] S220, inputting the extracted behavior features into an improved temporal convolutional network (TCN), wherein the improvement includes:

[0021] Combined with Graph Neural Network (GNN): Introducing the GNN module into the temporal convolutional network to model the spatial relationship between different areas in the building. By constructing regional nodes and regional connection edges, it captures the activity patterns and mutual influence of users between different areas.

[0022] Construction of regional relationship graph: Based on the building floor plan and historical usage data, a relationship graph between each area is constructed. The nodes represent each area, and the edges represent the connectivity and interaction between each area. The regional relationship graph construction includes constructing an undirected graph G = (V, E), where V is a node set representing each building area, and E is an edge set representing the connectivity and interaction between each area. Each node v i ∈V has a feature vector x i , including the usage frequency, time distribution, and spatial distribution characteristics of the area, and the connectivity relationship between nodes is represented by the adjacency matrix A. ij =1 indicates that there is a connectivity relationship between node i and node j, otherwise it is 0;

[0023] Spatiotemporal joint modeling: The spatial relationship features output by the graph neural network module are integrated with the time series features captured by the temporal convolutional network to achieve spatiotemporal joint modeling;

[0024] S230 uses an improved temporal convolutional network model for training. During the training process, a cross-validation method is used to evaluate the model performance and perform hyperparameter optimization. The trained improved temporal convolutional network model is used to predict the user's activity scenarios in different time periods and different areas based on the real-time collected data.

[0025] Optionally, the graph neural network is used for message transmission and node feature update, and the node feature update formula of each layer is:

[0026] in, is the feature vector of node i in the kth layer, N(i) represents the set of neighbor nodes of node i, and W (k) is the weight matrix of the kth layer, σ is the activation function;

[0027] In the temporal convolutional network, for each node v i , whose eigenvector h i The calculation after the temporal convolution layer is:

[0028] Among them, z t is the output of the temporal convolutional layer, W k is the weight of the kth convolution kernel, x t-k is the value of the input sequence at time step tk, and b is the bias;

[0029] The spatiotemporal joint modeling includes converting the spatial relationship features output by the graph neural network into Time series features captured by temporal convolutional networks Fusion is performed, and the fusion is expressed as:

[0030] Among them, concat represents the feature concatenation operation, f i,t is the final joint feature representation;

[0031] For the joint feature f i,t Perform full connection layer calculation to extract high-order features, expressed as:

[0032] y i,t =σ(W f f i,t +b f ), where W f and b f are the weights and biases of the fully connected layer, σ is the activation function, where y i,t Represents the activity intensity of region i in time period t.

[0033] Optionally, the dynamic adjustment of the lighting system in S300 includes calculating the lighting demand L of each area in different time periods. i,t :

[0034] Among them, α L , β L , γ L is the weight coefficient, is the spatial relationship feature of node i, is the time series feature of time step t, according to the calculated lighting demand L i,t ,y i,t It is the predicted activity intensity that dynamically adjusts the lighting intensity of each area;

[0035] The dynamic adjustment of the air conditioning system in S300 includes calculating the temperature requirements T of each area in different time periods. i,t :

[0036] Among them, α T , β T , γ T , δ T is the weight coefficient, xi is the temperature setting preference of node i, according to the calculated temperature demand T i,t , dynamically adjust the air conditioning temperature settings of each area;

[0037] The dynamic adjustment of the hot water system in S300 includes calculating the hot water demand H of each area in different time periods. i,t : Among them, α H , β H , γ H is the weight coefficient, according to the calculated hot water demand H i,t , dynamically adjust the operating status of the hot water supply system.

[0038] Optionally, the real-time environmental data in S400 includes indoor and outdoor temperature, humidity, air quality, light intensity, and activity intensity y i,t And real-time environmental data is input into the BIM model, dynamically updating the BIM model of the building, including dynamically updating the energy consumption data of each area:

[0039] E i,t =α·y i,t +β·T out,t +γ·H out,t +δ·AQ out,t , where E i,t represents the energy consumption of each area i in time period t, y i,t is the user activity prediction strength, T out,t is the outdoor temperature,

[0040] H out,t is the outdoor humidity, AQ out,t is the outdoor air quality, α, β, γ, δ are weight coefficients;

[0041] It also includes dynamically updating the operating status data of each device and dynamically updating the environmental parameter data.

[0042] Optionally, the dynamic update of the running status data of each device is expressed as:

[0043] S i,t =f(y i,t ,T out,t ,H out,t ,AQ out,t ), where S i,t represents the device status of each area i in time period t, and f is the status update function, which updates the device status based on user behavior and environmental data.

[0044] Optionally, the dynamically updated environmental parameter data is expressed as:

[0045] Among them, P i,t represents the environmental parameters of each region i in time period t. UpdateParams is the environmental parameter update function, which is updated based on real-time environmental data. out,t is the light intensity.

[0046] A green energy-saving building optimization design system based on BIM is used to implement the above-mentioned green energy-saving building optimization design method based on BIM, and includes the following modules:

[0047] User behavior data collection module: collects user behavior data in the building through smart devices, including usage data of each area in the building, hot water usage data, frequency and time of appliance usage, and temperature setting preferences;

[0048] Intelligent scene prediction module: Based on user behavior data, the key features of user behavior data are extracted, including usage frequency, time distribution, and spatial distribution. Through time convolutional networks and graph neural networks, spatiotemporal joint modeling is performed to establish a user behavior prediction model to predict the activity intensity of users in various areas within the building.

[0049] Adaptive Energy Management Module: Dynamically adjusts the building's energy allocation plan based on the prediction results of the user behavior prediction model, including:

[0050] Dynamically adjust the lighting system submodule: adjust the lighting intensity of each area according to the predicted activity intensity;

[0051] Dynamically adjust the air conditioning system submodule: adjust the air conditioning temperature settings of each area according to the predicted temperature demand;

[0052] Dynamic adjustment of hot water system submodule: adjust hot water supply according to predicted hot water demand;

[0053] BIM model dynamic update module: input the activity intensity and real-time environmental data of the user behavior prediction model into the BIM model, update the BIM model of the building, and understand the operating status of the building through the visualization and data analysis functions of the BIM model.

[0054] A storage medium, which can be read by a computer and stores a computer program. When the computer program is loaded by a processor, the above-mentioned BIM-based green energy-saving building optimization design method is executed.

[0055] Beneficial effects of the present invention:

[0056] The present invention, by introducing an improved temporal convolutional network (TCN) and a graph neural network (GNN), constructs an intelligent scene prediction model that can accurately predict the user's activity scenes in different time periods and different areas in the office building. The intelligent scene prediction can accurately foresee the user's behavior pattern and energy demand. For example, by predicting the user's activity intensity and time distribution in the office area, the operating status of the lighting, air conditioning and hot water systems is adjusted in advance to ensure that the optimal working conditions have been achieved before the user arrives, which not only improves the user's comfort, but also significantly improves the energy efficiency and avoids the waste of resources. At the same time, the intelligent scene prediction solution can also respond to emergencies in a timely manner, flexibly adjust the energy management strategy, ensure the stability and efficient operation of the building's internal environment, and achieve more intelligent and precise energy scheduling, which comprehensively improves the green energy-saving effect of the building.

[0057] The present invention establishes a user behavior prediction model by collecting and analyzing user behavior data, and inputs the prediction results and environmental data into the BIM model, dynamically updating the building's energy consumption, equipment status and environmental parameters. Through these dynamic updates and real-time adjustments, the operating status of each system inside the building can be accurately controlled, thereby significantly improving energy efficiency. For example, by dynamically adjusting the lighting system, air-conditioning system and hot water system, it is ensured that sufficient resources are provided in areas with more user activities, and energy consumption is reduced in areas with less activity, thereby achieving a green energy-saving effect.

[0058] The present invention introduces a status update function and an environmental parameter update function. Based on user behavior prediction and environmental data, the equipment operation status and environmental parameters of each office area are updated in real time, various equipment inside the building are effectively monitored and managed, abnormal conditions are identified and handled in time, and unnecessary energy waste is avoided. For example, the status update function can dynamically adjust the on and off status of the equipment to ensure that the equipment operates in the best state; the environmental parameter update function can adjust the indoor temperature, humidity, air quality and light intensity according to real-time environmental data, thereby maintaining a comfortable and efficient working environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0060] Figure 1 A schematic diagram of the optimization design method flow of an embodiment of the present invention;

[0061] Figure 2 A schematic diagram of intelligent scene prediction according to an embodiment of the present invention;

[0062] Figure 3 Schematic diagram of the functional modules of the optimization design system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. At the same time, it is explained here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments, and those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0064] It should be noted that the references to "one embodiment", "an embodiment", "an exemplary embodiment", "some embodiments" and the like in the specification indicate that the embodiments described may include specific features, structures or characteristics, but not every embodiment may include the specific features, structures or characteristics. In addition, when a specific feature, structure or characteristic is described in conjunction with an embodiment, it should be within the knowledge of a person skilled in the art to implement such feature, structure or characteristic in conjunction with other embodiments (whether or not explicitly described).

[0065] In general, a term can be understood, at least in part, from its use in context. For example, depending, at least in part, on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in the singular sense, or can be used to describe a combination of features, structures, or characteristics in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but can instead, depending, at least in part, on the context, allow for the presence of other factors that are not necessarily explicitly described.

[0066] like Figure 1-Figure 2 As shown, a green energy-saving building optimization design method based on BIM includes the following steps:

[0067] S100, user behavior data collection: collect user behavior data in the building through smart devices, including usage data of each area in the building, hot water usage data, frequency and time of appliance usage (usage of lighting, air conditioning, television, etc.), and temperature setting preferences (users' temperature setting preferences for different time periods and different areas);

[0068] S200, intelligent scene prediction: Based on user behavior data, a user behavior prediction model is established using machine learning algorithms to predict the activity intensity of users in various areas within the building. i,t ;

[0069] S300, Adaptive Energy Management: Dynamically adjust the building's energy allocation plan based on the prediction results of the user behavior prediction model, including dynamic adjustment of the lighting system, dynamic adjustment of the air conditioning system, and dynamic adjustment of the hot water system;

[0070] S400, BIM model dynamic update: the activity intensity of the user behavior prediction model y i,t And real-time environmental data is input into the BIM (Building Information Model) model to update the BIM model of the building, including energy consumption, equipment status and environmental parameters. Through the visualization and data analysis functions of the BIM model, a comprehensive understanding of the operating status of the building can be achieved.

[0071] The user behavior data collection in S100 specifically includes:

[0072] S110, install smart sensor devices in various areas of the building to monitor the usage of each area, including door and window sensors and infrared motion sensors to record the frequency and time of use of each area;

[0073] S120, install smart water meters and smart faucets to record the hot water usage in various areas of the building, including water consumption and usage time;

[0074] S130, through smart meters, records the usage frequency and time of each electrical device, including lighting equipment, air conditioners, and office equipment;

[0075] S140, through the intelligent thermostat, records the user's temperature setting preferences in different time periods and different areas;

[0076] S150, the collected data is transmitted to the central data processor via a wireless network for storage and analysis, and the collected data is pre-processed in the central data processor, including data cleaning, denoising and normalization, so as to be used for subsequent intelligent scene prediction and adaptive energy management.

[0077] Data cleaning: Identify and delete incomplete, erroneous or duplicate data records to ensure data accuracy and completeness;

[0078] Denoising: Use a mean filter to remove noise from the data and improve the quality of the data;

[0079] Normalization: Normalize the data and convert data of different dimensions to the same scale range (such as 0 to 1) to facilitate subsequent analysis and processing.

[0080] The intelligent scene prediction in S200 specifically includes:

[0081] S210, obtaining pre-processed user behavior data, and performing feature engineering to extract behavior features, where the behavior features include usage frequency, time distribution, and spatial distribution;

[0082] Frequency of use: Calculate the frequency with which each area is used over a period of time, such as the number of times it is used per day.

[0083] Time distribution: Statistics on the usage of each area in different time periods, such as the number of times used in the morning, noon, afternoon, and evening.

[0084] Spatial distribution: Count the mutual use of different office areas, such as how often employees go from the office to the meeting room;

[0085] S220, inputting the extracted behavior features into an improved temporal convolutional network (TCN), wherein the improvements include:

[0086] Combined with Graph Neural Network (GNN): Introducing the GNN module into the temporal convolutional network to model the spatial relationship between different areas in the building. By constructing regional nodes and regional connection edges, it captures the activity patterns and mutual influence of users between different areas.

[0087] Construction of regional relationship graph: Based on the building floor plan and historical usage data, a relationship graph between each area is constructed. The nodes represent each area, and the edges represent the connectivity and interaction between each area. The construction of the regional relationship graph includes constructing an undirected graph G = (V, E), where V is a node set representing each building area, and E is an edge set representing the connectivity and interaction between each area. Each node v i ∈V has a feature vector x i , including the usage frequency, time distribution, and spatial distribution characteristics of the area, and the connectivity relationship between nodes is represented by the adjacency matrix A. ij =1 indicates that there is a connectivity relationship between node i and node j, otherwise it is 0;

[0088] Spatiotemporal joint modeling: The spatial relationship features output by the graph neural network module are integrated with the time series features captured by the temporal convolutional network to achieve spatiotemporal joint modeling and improve the accuracy of predicting user behavior.

[0089] S230, using the improved temporal convolutional network model for training, using a cross-validation method to evaluate model performance during the training process, and performing hyperparameter optimization, using the trained improved temporal convolutional network model to predict user activity scenarios in different time periods and different areas based on real-time collected data;

[0090] The prediction results are input into the adaptive energy management system to dynamically adjust the building's energy allocation plan to optimize energy efficiency.

[0091] The graph neural network is used for message passing and node feature update. The node feature update formula of each layer is:

[0092] in, is the feature vector of node i in the kth layer, N(i) represents the set of neighbor nodes of node i, and W (k) is the weight matrix of the kth layer, σ is the activation function, and the activation function adopts the ReLU activation function;

[0093] In the temporal convolutional network, for each node v i , whose eigenvector h i The calculation after the temporal convolution layer is:

[0094] Among them, z t is the output of the temporal convolutional layer, W k is the weight of the kth convolution kernel, x t-k is the value of the input sequence at time step tk, and b is the bias;

[0095] Spatiotemporal joint modeling involves transforming the spatial relationship features output by the graph neural network into Time series features captured by temporal convolutional networks Fusion is performed, and the fusion is expressed as:

[0096] Among them, concat represents the feature concatenation operation, f i,t is the final joint feature representation;

[0097] For the joint feature f i,t Perform full connection layer calculation to extract high-order features, expressed as:

[0098] y i,t =σ(W f f i,t +b f ), where W f and b f are the weights and biases of the fully connected layer, σ is the activation function, where y i,t Represents the activity intensity of region i in time period t. As above, the ReLU activation function is selected.

[0099] Using the above joint feature y i,t Model training is carried out. During the training process, the cross-validation method is used to evaluate the model performance and perform hyperparameter optimization. The trained improved TCN model can predict user behavior based on the real-time collected data and predict the user's activity scenarios in different time periods and different areas.

[0100] The dynamic adjustment of the lighting system in S300 includes calculating the lighting demand L of each area at different time periods. i,t :

[0101] Among them, α L , βL , γ L is the weight coefficient, is the spatial relationship feature of node i, is the time series feature of time step t, according to the calculated lighting demand L i,t ,y i,t It is the predicted activity intensity that dynamically adjusts the lighting intensity of each area to ensure adequate lighting in areas with more user activity, while reducing lighting in areas with less activity to save energy;

[0102] The dynamic adjustment of the air conditioning system in S300 includes calculating the temperature requirements T of each area at different time periods. i,t :

[0103] Among them, α T , β T , γ T , δ T is the weight coefficient, x i is the temperature setting preference of node i, according to the calculated temperature demand T i,t , dynamically adjusts the air conditioning temperature settings in each area to ensure comfortable temperatures in areas with more user activity, while raising or lowering the temperature in areas with less activity to save energy;

[0104] The dynamic adjustment of the hot water system in S300 includes calculating the hot water demand H of each area in different time periods. i,t : Among them, α H , β H , γ H is the weight coefficient, according to the calculated hot water demand H i,t , dynamically adjust the operating status of the hot water supply system to ensure sufficient hot water is provided in areas with more user activities, while reducing hot water supply in areas with less activity to save energy;

[0105] Through the building's energy management system, the lighting demand, temperature demand and hot water demand calculated above are transmitted to the controllers of each subsystem, and the operating parameters of each subsystem are adjusted in real time to achieve dynamic energy allocation and management.

[0106] The weight coefficients mentioned above are estimated using multivariate linear regression combined with the least squares method. Take lighting demand as an example:

[0107] Assume that the target variable is the lighting demand L i,t , use the multiple linear regression model to fit:

[0108] Among them, is the error term, and the coefficient that needs to be determined is α L,β L ,γ L ;

[0109] The least squares method is used to estimate the regression coefficients, and the formula is as follows:

[0110] Where X is the feature matrix, including y i,t , y is the target variable L i,t The vector of are the estimated regression coefficients, including α L ,β L ,γ L , verify the model performance on the pre-prepared test set. If the model performs well, the determined weight coefficient can be used. If the model performance is poor, it is necessary to adjust the features and increase the amount of data.

[0111] The real-time environmental data in S400 includes indoor and outdoor temperature, humidity, air quality, light intensity, activity intensity i,t And real-time environmental data is input into the BIM model, dynamically updating the BIM model of the building, including dynamically updating the energy consumption data of each area:

[0112] E i,t =α·y i,t +β·T out,t +γ·H out,t +δ·AQ out,t , where E i,t represents the energy consumption of each area i in time period t, y i,t is the user activity prediction strength, T out,t is the outdoor temperature,

[0113] H out,t is the outdoor humidity, AQ out,t is the outdoor air quality, α, β, γ, δ are weight coefficients;

[0114] It also includes dynamically updating the operating status data of each device and dynamically updating the environmental parameter data;

[0115] Through the visualization function of the BIM model, the operating status of the building can be viewed, including energy consumption distribution, equipment operating efficiency, and environmental conditions, providing data support for optimized design.

[0116] Dynamically update the operating status data of each device as follows:

[0117] S i,t =f(y i,t ,T out,t ,H out,t ,AQ out,t ), where Si,t represents the device status of each area i in time period t, and f is the status update function, which updates the device status based on user behavior and environmental data;

[0118] The state update function f updates the operating state of the device based on user behavior prediction and real-time environmental data. The function is further expressed as:

[0119] S i,t =f(y i,t ,T out,t ,H out,t ,AQ out,t )=σ

[0120] (α s ·y i,t +β s ·T out,t +γ s ·H out,t +δ s ·AQ out,t +b s , where S i,t is the device state of each region i in time period t (e.g., on or off), σ is the activation function (Sigmoid function), which is used to map the result to the interval [0, 1], indicating the probability of the device state, α s ,β s ,γ s ,δ s is the weight coefficient, which indicates the influence of different factors on the equipment status, b s is the bias term.

[0121] Dynamically updated environmental parameter data is represented as:

[0122] Among them, P i,t represents the environmental parameters of each region i in time period t. UpdateParams is the environmental parameter update function, which is updated based on real-time environmental data. out,t is the light intensity.

[0123] The parameter relationship in the environment parameter update function is specifically expressed as follows:

[0124] T i,t =α T ·T out,t +β T ·T in,t-1 +b T ;

[0125] H i,t =α H ·H out,t +β H ·Hin,t-1 +b H ;

[0126] AQ i,t =α AQ ·AQ out,t +β AQ ·AQ in,t-1 +b AQ ;

[0127] LI i,t =α LI ·LI out,t +β LI ·LI in,t-1 +b LI ;

[0128] Among them, α T , β T , α H , β H , α AQ , β AQ , α LI , β LI is the weight coefficient, T in,t-1 , H in,t-1 AQ in,t-1 , LI in,t-1 are the indoor temperature, humidity, air quality and light intensity in the previous time period, b T , b H , b AQ , b LI is the bias term.

[0129] like Figure 3 As shown, a BIM-based green energy-saving building optimization design system is used to implement the above-mentioned BIM-based green energy-saving building optimization design method, including the following modules:

[0130] User behavior data collection module: collects user behavior data in the building through smart devices, including usage data of each area in the building, hot water usage data, frequency and time of appliance usage, and temperature setting preferences;

[0131] Intelligent scene prediction module: Based on user behavior data, the key features of user behavior data are extracted, including usage frequency, time distribution, and spatial distribution. Through time convolutional networks and graph neural networks, spatiotemporal joint modeling is performed to establish a user behavior prediction model to predict the activity intensity of users in various areas within the building.

[0132] Adaptive Energy Management Module: Dynamically adjusts the building's energy allocation plan based on the prediction results of the user behavior prediction model, including:

[0133] Dynamically adjust the lighting system submodule: adjust the lighting intensity of each area according to the predicted activity intensity;

[0134] Dynamically adjust the air conditioning system submodule: adjust the air conditioning temperature settings of each area according to the predicted temperature demand;

[0135] Dynamic adjustment of hot water system submodule: adjust hot water supply according to predicted hot water demand;

[0136] BIM model dynamic update module: input the activity intensity and real-time environmental data of the user behavior prediction model into the BIM model, update the BIM model of the building, and understand the operating status of the building through the visualization and data analysis functions of the BIM model.

[0137] It also includes a storage medium, which can be read by a computer and stores a computer program. When the computer program is loaded by a processor, it executes the above-mentioned BIM-based green energy-saving building optimization design method.

[0138] The present invention covers any substitution, modification, equivalent method and scheme made on the essence and scope of the present invention. In order to make the public have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention, but those skilled in the art can fully understand the present invention without the description of these details. In addition, in order to avoid unnecessary confusion about the essence of the present invention, well-known methods, processes, procedures, components and circuits are not described in detail.

[0139] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A green energy-saving building optimization design method based on BIM, characterized in that: The following steps are involved: S100, user behavior data collection: collect user behavior data in the building through smart devices, including usage data of each area in the building, hot water usage data, appliance usage frequency and time, and temperature setting preferences; S200, intelligent scene prediction: Based on user behavior data, a user behavior prediction model is established using machine learning algorithms to predict the activity intensity of users in various areas within the building. i,t ; S300, Adaptive Energy Management: Dynamically adjust the building's energy allocation plan based on the prediction results of the user behavior prediction model, including dynamic adjustment of the lighting system, dynamic adjustment of the air conditioning system, and dynamic adjustment of the hot water system; S400, BIM model dynamic update: the activity intensity of the user behavior prediction model is y i,t And real-time environmental data is input into the BIM model to update the BIM model of the building, including energy consumption, equipment status and environmental parameters.

2. The green energy-saving building optimization design method based on BIM according to claim 1 is characterized in that: The user behavior data collection in S100 specifically includes: S110, install smart sensor devices in various areas of the building to monitor the usage of each area, including door and window sensors and infrared motion sensors to record the frequency and time of use of each area; S120, install smart water meters and smart faucets to record the hot water usage in various areas of the building, including water consumption and usage time; S130, through smart meters, records the usage frequency and time of each electrical device, including lighting equipment, air conditioners, and office equipment; S140, through the intelligent thermostat, records the user's temperature setting preferences in different time periods and different areas; S150, transmitting the collected data to a central data processor via a wireless network for storage and analysis, and preprocessing the collected data in the central data processor, including data cleaning, denoising and normalization.

3. The green energy-saving building optimization design method based on BIM according to claim 1 is characterized in that: The intelligent scene prediction in S200 specifically includes: S210, obtaining pre-processed user behavior data, and performing feature engineering to extract behavior features, where the behavior features include usage frequency, time distribution, and spatial distribution; S220, inputting the extracted behavior features into an improved temporal convolutional network, wherein the improvement includes: Combined with graph neural network: Introduce graph neural network module into the temporal convolutional network to model the spatial relationship between different areas in the building. By constructing regional nodes and regional connection edges, it captures the activity patterns and mutual influence of users between different areas. Construction of regional relationship graph: Based on the building floor plan and historical usage data, a relationship graph between each area is constructed. The nodes represent each area, and the edges represent the connectivity and interaction between each area. The regional relationship graph construction includes constructing an undirected graph G = (V, E), where V is a node set representing each building area, and E is an edge set representing the connectivity and interaction between each area. Each node v i ∈V has a feature vector x i , including the usage frequency, time distribution, and spatial distribution characteristics of the area, and the connectivity relationship between nodes is represented by the adjacency matrix A. ij =1 indicates that there is a connectivity relationship between node i and node j, otherwise it is 0; Spatiotemporal joint modeling: The spatial relationship features output by the graph neural network module are integrated with the time series features captured by the temporal convolutional network to achieve spatiotemporal joint modeling; S230 uses an improved temporal convolutional network model for training. During the training process, a cross-validation method is used to evaluate the model performance and perform hyperparameter optimization. The trained improved temporal convolutional network model is used to predict the user's activity scenarios in different time periods and different areas based on the real-time collected data.

4. The green energy-saving building optimization design method based on BIM according to claim 3 is characterized in that: The graph neural network is used for message transmission and node feature update. The node feature update formula of each layer is: in, is the feature vector of node i in the kth layer, N(i) represents the set of neighbor nodes of node i, and W (k) is the weight matrix of the kth layer, σ is the activation function; In the temporal convolutional network, for each node v i , whose eigenvector h i The calculation after the temporal convolution layer is: Among them, z t is the output of the temporal convolutional layer, W k is the weight of the kth convolution kernel, x t-k is the value of the input sequence at time step tk, and b is the bias; The spatiotemporal joint modeling includes converting the spatial relationship features output by the graph neural network into Time series features captured by temporal convolutional networks Fusion is performed, and the fusion is expressed as: Among them, concat represents the feature concatenation operation, f i,t is the final joint feature representation; For the joint feature f i,t Perform full connection layer calculation to extract high-order features, expressed as: y i,t =σ(W f f i,t +b f ), where W f and b f are the weights and biases of the fully connected layer, σ is the activation function, where y i,t Represents the activity intensity of region i in time period t.

5. The green energy-saving building optimization design method based on BIM according to claim 4 is characterized in that: The dynamic adjustment of the lighting system in S300 includes calculating the lighting demand L of each area in different time periods. i,t : Among them, α L , β L , γ L is the weight coefficient, is the spatial relationship feature of node i, is the time series feature of time step t, according to the calculated lighting demand L i,t ,y i,t It is the predicted activity intensity that dynamically adjusts the lighting intensity of each area; The dynamic adjustment of the air conditioning system in S300 includes calculating the temperature requirements T of each area in different time periods. i,t : Among them, α T , β T , γ T , δ T is the weight coefficient, x i is the temperature setting preference of node i, according to the calculated temperature demand T i,t , dynamically adjust the air conditioning temperature settings of each area; The dynamic adjustment of the hot water system in S300 includes calculating the hot water demand of each area in different time periods. Among them, α H , β H , γ H is the weight coefficient, according to the calculated hot water demand H i,t , dynamically adjust the operating status of the hot water supply system.

6. The green energy-saving building optimization design method based on BIM according to claim 5 is characterized in that: The real-time environmental data in S400 include indoor and outdoor temperature, humidity, air quality, light intensity, and activity intensity y i,t And real-time environmental data is input into the BIM model to dynamically update the BIM model of the building, including dynamically updating the energy consumption data of each area: E i,t =α·y i,t +β·T out,t +γ·H out,t +δ·AQ out,t , where E i,t represents the energy consumption of each area i in time period t, y i,t is the user activity prediction strength, T out,t is the outdoor temperature, H out,t is the outdoor humidity, AQ out,t is the outdoor air quality, α, β, γ, δ are weight coefficients; It also includes dynamically updating the operating status data of each device and dynamically updating the environmental parameter data.

7. The green energy-saving building optimization design method based on BIM according to claim 6 is characterized in that: The dynamic update of the running status data of each device is expressed as: S i,t =f(y i,t ,T out,t ,H out,t ,AQ out,t ), where S i,t represents the device status of each area i in time period t, and f is the status update function, which updates the device status based on user behavior and environmental data.

8. The green energy-saving building optimization design method based on BIM according to claim 7 is characterized in that: The dynamically updated environmental parameter data is expressed as: Among them, P i,t represents the environmental parameters of each region i in time period t. UpdateParams is the environmental parameter update function, which is updated based on real-time environmental data. out,t is the light intensity.

9. A BIM-based green energy-saving building optimization design system, used to implement a BIM-based green energy-saving building optimization design method as claimed in any one of claims 1 to 8, characterized in that: Includes the following modules: User behavior data collection module: collects user behavior data in the building through smart devices, including usage data of each area in the building, hot water usage data, frequency and time of appliance usage, and temperature setting preferences; Intelligent scene prediction module: Based on user behavior data, the key features of user behavior data are extracted, including usage frequency, time distribution, and spatial distribution. Through time convolutional networks and graph neural networks, spatiotemporal joint modeling is performed to establish a user behavior prediction model to predict the activity intensity of users in various areas within the building. Adaptive Energy Management Module: Dynamically adjusts the building's energy allocation plan based on the prediction results of the user behavior prediction model, including: Dynamically adjust the lighting system submodule: adjust the lighting intensity of each area according to the predicted activity intensity; Dynamically adjust the air conditioning system submodule: adjust the air conditioning temperature settings of each area according to the predicted temperature demand; Dynamic adjustment of hot water system submodule: adjust hot water supply according to predicted hot water demand; BIM model dynamic update module: input the activity intensity and real-time environmental data of the user behavior prediction model into the BIM model, update the BIM model of the building, and understand the operating status of the building through the visualization and data analysis functions of the BIM model.

10. A storage medium, characterized in that: The storage medium can be read by a computer, and the storage medium stores a computer program. When the computer program is loaded by a processor, the BIM-based green energy-saving building optimization design method described in any one of claims 1 to 8 is executed.

Images