Building energy consumption analysis method and system based on multi-mode perception

Abstract

The invention relates to the technical field of building energy consumption analysis, and discloses a building energy consumption analysis method and system based on multi-modal perception. According to the method and the system, building environment data, energy consumption data, personnel flow data and equipment state data are collected from a plurality of data sources, the data are cleaned, abnormal values and noise information are eliminated, the data collected at different time points are integrated through a time alignment method, and the data are stored in a database. Data from different sources and dimensions are standardized, feature extraction and modeling are carried out on the preprocessed data, time features, environment features and personnel flow features are extracted, an energy consumption change mathematical model is established, energy consumption analysis is carried out on the extracted features, total energy consumption of a building is calculated for energy efficiency evaluation, and energy efficiency evaluation is carried out. The energy consumption intensity and the energy efficiency ratio are obtained, time sequence analysis is conducted through historical data, the future building energy consumption trend is predicted, the analysis result is visualized, and energy conservation and emission reduction suggestions and optimization decisions are provided based on the visualized analysis result.

Description

Technical Field

[0001] The present invention relates to the technical field of building energy consumption analysis, and in particular to a building energy consumption analysis method and system based on multimodal sensing. Background Art

[0002] Building energy consumption analysis is an in-depth study of energy usage in buildings, covering all aspects of collecting, analyzing and optimizing building energy consumption. By analyzing building environmental data, changes in factors such as temperature, humidity and air quality can be identified. Managers can adjust system settings based on real-time data to improve the comfort of the indoor environment. In addition, good energy efficiency management can ensure the normal operation of building facilities and reduce safety hazards caused by equipment failures. At the same time, it can identify high-energy consumption areas and potential energy-saving opportunities, allowing managers to implement energy-saving measures in a targeted manner.

[0003] Currently, many buildings rely on regular manual data collection and analysis, which makes it difficult to achieve real-time monitoring and dynamic feedback. The lack of real-time data limits the ability to adjust and respond to changes in energy consumption in a timely manner, reducing the energy utilization efficiency of buildings. Summary of the invention

[0004] 1. Technical issues to be solved

[0005] In view of the shortcomings of the prior art, the present invention provides a building energy consumption analysis method and system based on multimodal perception, which has the functions of collecting building environment data, energy consumption data, personnel flow data and equipment status data from multiple data sources, cleaning the above-mentioned collected data, eliminating outliers and noise information, integrating data collected at different time points through a time alignment method, standardizing data from different sources and dimensions, extracting features and modeling the pre-processed data, extracting time features, environmental features and personnel flow features, and establishing a mathematical model of energy consumption changes, performing energy consumption analysis on the extracted features, calculating the total energy consumption of the building for energy efficiency evaluation, and obtaining energy consumption intensity and energy efficiency ratio, using historical data for time series analysis, predicting future building energy consumption trends, visualizing the analysis results, and using charts and heat maps and other intuitive forms to display the changing trends and distribution of energy consumption, proposing energy-saving and emission reduction suggestions and optimization decisions based on the visualized analysis results, and collecting building energy consumption-related data for analysis through intelligent means, and performing real-time monitoring, timely adjusting and optimizing the energy consumption anomalies of the building, thereby improving the utilization efficiency of building energy and solving the above-mentioned problems.

[0006] (II) Technical solution

[0007] To achieve the above object, the present invention provides the following technical solution: a building energy consumption analysis method based on multimodal perception, comprising the following steps:

[0008] S1, collect building environment data, energy consumption data, personnel flow data and equipment status data from multiple data sources;

[0009] S2. Clean the above collected data, remove outliers and noise information, integrate the data collected at different time points through time alignment method, and standardize the data from different sources and dimensions;

[0010] S3. Extract features and model the preprocessed data, extract time features, environmental features and personnel flow features, and establish a mathematical model of energy consumption changes;

[0011] S4. Perform energy consumption analysis on the extracted features, calculate the total energy consumption of the building for energy efficiency evaluation, obtain the energy consumption intensity and energy efficiency ratio, use historical data for time series analysis, and predict future building energy consumption trends;

[0012] S5. Visualize the analysis results and use charts, heat maps and other intuitive forms to show the changing trend and distribution of energy consumption;

[0013] S6. Based on the visual analysis results, make recommendations for energy conservation and emission reduction and optimize decisions.

[0014] Preferably, the formula for denoising the collected data in S2 is as follows:

[0015]

[0016] In the formula, Yd t represents the smoothed value at time t, Kx t-i Represents the value of the original data at time ti, N represents the window size, i represents the count subscript, and the formula represents the removal of data noise through moving average.

[0017] Preferably, the formula for extracting time features in S3 is as follows:

[0018]

[0019] In the formula, St represents the time characteristic value, sin represents the sign of sine, cos represents the sign of cosine, H represents the number of hours in a day, and π represents 3.1415.

[0020] Preferably, the formula for extracting environmental features in S3 is as follows:

[0021]

[0022] In the formula, Tp represents the environmental characteristic value, T represents the temperature data at time, E represents the energy consumption data at time, Cov(T, E) represents the covariance of temperature and energy consumption, and Var(T) represents the variance of temperature data.

[0023] Preferably, the formula for extracting personnel flow characteristics in S3 is as follows:

[0024]

[0025] In the formula, Lr represents the personnel flow characteristics, R in represents the number of inflows in a specific time period, R out Indicates the number of outflows in a specific time period, R total Represents the total number of people in a specific time period.

[0026] Preferably, the energy consumption change mathematical model established in S3 is as follows:

[0027] Ek=k0+k1*St+k2*St sin +k3*St cos +k4*Tp+k5*Lr+∈

[0028] In the formula, Ek represents the energy consumption of the building, k0 represents the intercept term, that is, the energy consumption value predicted by the model when all other features are 0, k1 represents the coefficient of the time feature St, which represents the change in energy consumption Ek for every hour of increase. A positive value indicates that energy consumption increases with the passage of time, and vice versa. k2 represents the coefficient of the daily cycle sinusoidal feature, which is used to capture the periodic changes in energy consumption within a day, especially the impact on energy consumption in a specific period of time. When the coefficient is positive, it is considered that the energy consumption in this period of time is high. k3 represents the coefficient of the daily cycle cosine feature, k4 is the coefficient of the environmental feature, which represents the change in energy consumption for every unit change in temperature. The change in temperature will directly affect the air conditioning or heating demand of the building, thereby affecting energy consumption. k5 is the coefficient of the personnel flow feature, which represents the impact of personnel flow on energy consumption. When the number of staff in the building increases, energy consumption will increase. ∈ is the error term, which represents the part that the model cannot explain, including all other factors not included in the model, measurement errors and random noise.

[0029] Preferably, the formula for calculating the total energy consumption of the building in S4 is as follows:

[0030] Lyzh=Cn+Zl+Zm+Sn

[0031] In the formula, Lyzh represents the total energy consumption of the building, Cn represents heating energy consumption, which refers to the energy consumed to maintain the temperature in the building, Zl represents cooling energy consumption, which refers to the energy consumed to reduce the temperature in the building, Zm represents lighting energy consumption, which refers to the electric energy used for lighting, and Sn represents equipment energy consumption, which refers to the energy consumed by various electrical equipment.

[0032] Preferably, the formula for calculating energy consumption intensity in S4 is as follows:

[0033]

[0034] In the formula, Nqdk represents energy consumption intensity, Lyzh represents total energy consumption of the building, and Jzmj represents the total area of ​​the building.

[0035] Preferably, the formula for calculating the energy efficiency ratio in S4 is as follows:

[0036]

[0037] In the formula, COP represents energy efficiency ratio, Scnl represents output energy, and Srdl represents input energy.

[0038] A building energy consumption analysis system based on multimodal perception, including a data acquisition module, a data preprocessing module, a feature extraction and modeling module, an energy consumption analysis and prediction module, a data visualization module and an optimization module;

[0039] The data acquisition module collects building environment data, energy consumption data, personnel flow data and equipment status data from multiple data sources;

[0040] The data preprocessing module cleans the collected data, removes outliers and noise information, integrates the data collected at different time points through a time alignment method, and standardizes the data from different sources and dimensions;

[0041] The feature extraction and modeling module performs feature extraction and modeling on the preprocessed data, extracts time features, environmental features and personnel flow features, and establishes a mathematical model of energy consumption changes;

[0042] The energy consumption analysis and prediction module performs energy consumption analysis on the extracted features, calculates the total energy consumption of the building for energy efficiency evaluation, obtains energy consumption intensity and energy efficiency ratio, uses historical data for time series analysis, and predicts future building energy consumption trends;

[0043] The data visualization module visualizes the analysis results and uses intuitive forms such as charts and heat maps to display the changing trend and distribution of energy consumption;

[0044] The optimization module proposes energy-saving and emission-reduction suggestions and optimization decisions based on the visual analysis results.

[0045] Compared with the prior art, the present invention provides a building energy consumption analysis method and system based on multimodal perception, which has the following beneficial effects:

[0046] The present invention collects building environment data, energy consumption data, personnel flow data and equipment status data from multiple data sources, cleans the above collected data, removes outliers and noise information, integrates data collected at different time points through a time alignment method, standardizes data from different sources and dimensions, extracts features and models the pre-processed data, extracts time features, environmental features and personnel flow features, and establishes a mathematical model of energy consumption changes, performs energy consumption analysis on the extracted features, calculates the total energy consumption of the building for energy efficiency evaluation, obtains energy consumption intensity and energy efficiency ratio, uses historical data for time series analysis, predicts future building energy consumption trends, visualizes the analysis results, uses charts and heat maps and other intuitive forms to display the changing trends and distribution of energy consumption, proposes energy-saving and emission reduction suggestions and optimization decisions based on the visualized analysis results, collects building energy consumption-related data for analysis through intelligent means, and performs real-time monitoring, timely adjusts and optimizes the energy consumption anomalies of the building, and improves the utilization efficiency of building energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram of the steps of the method of the present invention;

[0048] Figure 2 It is a schematic diagram of the system flow of the present invention. DETAILED DESCRIPTION

[0049] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] Many buildings currently rely on regular manual data collection and analysis, which makes it difficult to achieve real-time monitoring and dynamic feedback. The lack of real-time data will limit the ability to adjust and respond to energy consumption changes in a timely manner, reducing the efficiency of building energy utilization. Therefore, a building energy consumption analysis method based on multimodal perception is proposed. Figure 1 , the method comprises the following steps:

[0051] S1, collect building environment data, energy consumption data, personnel flow data and equipment status data from multiple data sources;

[0052] By using IoT technology, with the help of sensors and smart devices, it is possible to efficiently collect building environment data such as temperature, humidity, air quality, etc., energy consumption data such as electricity, water, gas, etc., personnel flow data through access control systems and video surveillance to obtain personnel entry and exit, and equipment status data such as the operating status of HVAC systems and lighting equipment from multiple data sources. These data are transmitted and stored in real time through wireless networks and cloud computing platforms. Using big data processing and analysis technology, we can ensure the rapid integration and processing of data. By applying data collection and analysis tools, managers can cross-verify information from different data sources to improve the accuracy and reliability of data, thereby laying a solid foundation for subsequent energy consumption analysis and intelligent decision-making;

[0053] S2. Clean the above collected data, remove outliers and noise information, integrate the data collected at different time points through time alignment method, and standardize the data from different sources and dimensions;

[0054] The formula for denoising the collected data is as follows:

[0055]

[0056] Denoising can effectively reduce outliers and noise in the data and improve the accuracy and reliability of the data, which means that the input data based on which the subsequent analysis is based will more truly reflect the actual energy consumption of the building. In the formula, Yd t represents the smoothed value at time t, Kx t-i Represents the value of the original data at time ti, N represents the window size, i represents the count subscript, and the formula represents the removal of data noise through moving average. Denoising can reduce unnecessary data redundancy and noise, and reduce the cost of data storage and processing, which makes resources more efficiently utilized, especially when processing large amounts of real-time data.

[0057] By applying the statistical analysis method Z- sco r e Perform outlier detection on the data set and remove those outliers that exceed the preset threshold to ensure the accuracy and reliability of the data. Use timestamp synchronization technology to integrate data collected at different time points to ensure the consistency of all data in the time dimension, which is convenient for subsequent analysis. In addition, for the standardized processing of data from different sources such as sensors, monitoring systems, energy management platforms, etc. and dimensions such as temperature, humidity, energy consumption, etc., use methods such as normalized Min-Max Scaling to convert data to a unified dimension and eliminate the impact of dimensional differences on analysis results;

[0058] S3. Extract features and model the preprocessed data, extract time features, environmental features and personnel flow features, and establish a mathematical model of energy consumption changes;

[0059] The formula for extracting time features is as follows:

[0060]

[0061] By extracting time features, we can identify the periodic change pattern of energy consumption, such as the difference in energy consumption between daytime and nighttime, weekdays and weekends. This helps managers understand the regularity of energy consumption and formulate more effective energy-saving strategies. In the formula, St represents the time feature value, sin represents the sine sign, cos represents the cosine sign, H represents the number of hours in a day, and π represents 3.1415. Based on time features, we can optimize the operation and scheduling of equipment, such as running high-energy-consuming equipment during low-energy consumption periods, thereby reducing overall energy consumption and costs.

[0062] The formula for extracting environmental features is as follows:

[0063]

[0064] The extraction of environmental characteristics can help analyze its impact on energy consumption and identify the relationship between environmental factors and energy consumption, thus providing a scientific basis for energy efficiency optimization. In the formula, Tp represents the environmental characteristic value, T represents the temperature data at time, E represents the energy consumption data at time, Cov(T, E) represents the covariance of temperature and energy consumption, and Var(T) represents the variance of temperature data. By analyzing environmental characteristics, the operation of the building's HVAC system can be optimized to ensure that energy consumption is reduced and energy efficiency is improved while meeting comfort requirements.

[0065] The covariance calculation formula of temperature and energy consumption is as follows:

[0066]

[0067] In the formula, Cov(T, E) represents the covariance of temperature and energy consumption, T represents temperature, E represents energy consumption, and n represents the number of data points. represents the mean of the temperature data set, represents the mean of the energy consumption data set;

[0068] The formula for extracting personnel flow characteristics is as follows:

[0069]

[0070] The extraction of personnel flow characteristics can help managers optimize the use of space, identify which areas are frequently used and which areas are idle, so as to rationally allocate resources and improve space utilization. In the formula, Lr represents the personnel flow characteristics, Rin represents the number of inflows in a specific time period, R out Indicates the number of outflows in a specific time period, R total Indicates the total number of people in a specific time period. By analyzing the flow characteristics of people, we can understand the usage patterns and behavior habits of people in the building, identify the difference in energy consumption during peak and off-peak hours, and provide a basis for energy consumption management;

[0071] The established mathematical model of energy consumption change is as follows:

[0072] Ek=k0+k1*St+k2*St sin +k3*St cos +k4*Tp+k5*Lr+∈

[0073] In the formula, Ek represents the energy consumption of the building, k0 represents the intercept term, that is, the energy consumption value predicted by the model when all other features are 0, k1 represents the coefficient of the time feature St, which represents the change in energy consumption Ek for every hour of improvement. A positive value indicates that energy consumption increases with the passage of time, and vice versa. k2 represents the coefficient of the daily cycle sine feature, which is used to capture the periodic changes in energy consumption within a day, especially the impact on energy consumption in a specific period of time. When the coefficient is positive, it is considered that the energy consumption in this period of time is high. k3 represents the coefficient of the daily cycle cosine feature, k4 is the coefficient of the environmental feature, which represents the change in energy consumption for every unit change in temperature. The change in temperature will directly affect the air conditioning or heating demand of the building, thereby affecting energy consumption. k5 is the coefficient of the personnel flow feature, which represents the impact of personnel flow on energy consumption. When the number of staff in the building increases, energy consumption will increase. ∈ is the error term, which represents the part that the model cannot explain, including all other factors not included in the model, measurement errors and random noise.

[0074] Mathematical models can provide a quantitative analysis framework, making the interpretation and decision-making of energy consumption data more accurate. They can clearly show the relationship between variables, accurately identify key factors affecting energy consumption such as temperature, humidity, personnel flow, etc., and quantify their impact. At the same time, they can capture the nonlinear relationship between different variables and reveal complex energy consumption behaviors.

[0075] S4. Perform energy consumption analysis on the extracted features, calculate the total energy consumption of the building for energy efficiency evaluation, obtain the energy consumption intensity and energy efficiency ratio, use historical data for time series analysis, and predict future building energy consumption trends;

[0076] The formula for calculating the total energy consumption of a building is as follows:

[0077] Lyzh=Cn+Zl+Zm+Sn

[0078] Calculating total energy consumption can provide managers with comprehensive usage of all equipment in the building and help understand the overall energy consumption level. In the formula, Lyzh represents the total energy consumption of the building, Cn represents heating energy consumption, which refers to the energy consumed to maintain the temperature in the building, Zl represents cooling energy consumption, which refers to the energy consumed to reduce the temperature in the building, Zm represents lighting energy consumption, which refers to the electrical energy used for lighting, and Sn represents equipment energy consumption, which refers to the energy consumed by various electrical equipment. Monitoring the trend of total energy consumption can help managers identify abnormal situations, such as sudden energy consumption surges, so that they can take corresponding measures;

[0079] The formula for calculating energy intensity is as follows:

[0080]

[0081] Energy intensity provides a standardized way to evaluate energy consumption, which enables effective comparison of different buildings and buildings with different uses. In the formula, Nqdk represents energy intensity, Lyzh represents total building energy consumption, and Jzmj represents the total area of ​​the building. By analyzing energy intensity, managers can identify areas with good energy efficiency and potential energy-saving improvement space.

[0082] The formula for calculating the energy efficiency ratio is as follows:

[0083]

[0084] The energy efficiency ratio provides a comprehensive efficiency evaluation index that can reflect the overall performance of the system. Monitoring the energy efficiency ratio can promote technological innovation and process optimization, and help find and apply new energy-saving technologies. In the formula, COP represents the energy efficiency ratio, Scnl represents the output energy, and Srdl represents the input energy. The energy efficiency ratio helps to combine sustainable development goals with actual energy efficiency management, and provides data support for achieving environmental protection standards and energy-saving goals.

[0085] Time series analysis using statistical analysis and autoregressive integrated moving average models can identify seasonal and trend changes in energy consumption;

[0086] S5. Visualize the analysis results and use charts, heat maps and other intuitive forms to show the changing trend and distribution of energy consumption;

[0087] Using a line chart to display the energy consumption forecast for each month can clearly show the fluctuation trend of energy consumption in different time periods. This chart can highlight seasonal changes and facilitate the rapid identification of energy consumption peaks and valleys. In addition, a bar chart is used to compare historical monthly energy consumption with predicted energy consumption to help managers evaluate the actual effect of implementing energy-saving measures. In order to further analyze the data distribution of energy consumption, a heat map is used to correspond time such as month and date to energy consumption values. Combined with the depth of color, it can effectively display the specific time points of high energy consumption and low energy consumption, which is convenient for managers to quickly identify the periods of concentrated and dispersed energy consumption. The dark area in the heat map represents the peak use period of air conditioning, while the light area shows low energy consumption when there are fewer people. With the help of interactive data visualization tools such as Tableau and Power BI, more dynamic chart interactions are achieved, allowing users to freely select time ranges and view specific data points, so as to deeply explore the factors closely related to economic activities and energy consumption behind the data;

[0088] S6. Propose energy-saving and emission-reduction suggestions and optimize decisions based on the visual analysis results;

[0089] Through in-depth analysis of visual charts, peak and trough periods of energy consumption can be identified. This data can guide managers to take proactive energy-saving measures during peak periods, such as adjusting air conditioning temperature settings, limiting the use of non-essential equipment, or adopting demand response mechanisms to balance energy consumption;

[0090] See also Figure 2 , a building energy consumption analysis system based on multimodal perception, including a data acquisition module, a data preprocessing module, a feature extraction and modeling module, an energy consumption analysis and prediction module, a data visualization module and an optimization module;

[0091] The data acquisition module collects building environment data, energy consumption data, personnel flow data and equipment status data from multiple data sources;

[0092] The data preprocessing module cleans the collected data, removes outliers and noise information, integrates the data collected at different time points through a time alignment method, and standardizes the data from different sources and dimensions;

[0093] The feature extraction and modeling module performs feature extraction and modeling on the preprocessed data, extracts time features, environmental features and personnel flow features, and establishes a mathematical model of energy consumption changes;

[0094] The energy consumption analysis and prediction module performs energy consumption analysis on the extracted features, calculates the total energy consumption of the building for energy efficiency evaluation, obtains energy consumption intensity and energy efficiency ratio, uses historical data for time series analysis, and predicts future building energy consumption trends;

[0095] The data visualization module visualizes the analysis results and uses intuitive forms such as charts and heat maps to display the changing trend and distribution of energy consumption;

[0096] The optimization module proposes energy-saving and emission-reduction suggestions and optimization decisions based on the visual analysis results.

[0097] Through the application of the above method and system, the building energy consumption related data is intelligently collected for analysis and real-time monitoring, and the abnormal energy consumption of the building is adjusted and optimized in time, thereby improving the utilization efficiency of the building energy.

[0098] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A building energy consumption analysis method based on multimodal perception, characterized in that: The following steps are involved: S1, collect building environment data, energy consumption data, personnel flow data and equipment status data from multiple data sources; S2. Clean the above collected data, remove outliers and noise information, integrate the data collected at different time points through time alignment method, and standardize the data from different sources and dimensions; S3. Extract features and model the preprocessed data, extract time features, environmental features and personnel flow features, and establish a mathematical model of energy consumption changes; S4. Perform energy consumption analysis on the extracted features, calculate the total energy consumption of the building for energy efficiency evaluation, obtain the energy consumption intensity and energy efficiency ratio, use historical data for time series analysis, and predict future building energy consumption trends; S5. Visualize the analysis results and use charts, heat maps and other intuitive forms to show the changing trend and distribution of energy consumption; S6. Based on the visual analysis results, make recommendations for energy conservation and emission reduction and optimize decisions.

2. The building energy consumption analysis method based on multimodal perception according to claim 1 is characterized in that: The formula for S2 to denoise the collected data is as follows: In the formula, Yd t represents the smoothed value at time t, Kx t-i Represents the value of the original data at time ti, N represents the window size, i represents the count subscript, and the formula represents the removal of data noise through moving average.

3. The building energy consumption analysis method based on multimodal sensing according to claim 2 is characterized in that: The formula for extracting time features of S3 is as follows: In the formula, St represents the time characteristic value, sin represents the sign of sine, cos represents the sign of cosine, H represents the number of hours in a day, and π represents 3.1415.

4. The building energy consumption analysis method based on multimodal sensing according to claim 3 is characterized in that: The formula for extracting environmental features by S3 is as follows: In the formula, Tp represents the environmental characteristic value, T represents the temperature data at time, E represents the energy consumption data at time, Cov(T,E) represents the covariance of temperature and energy consumption, and Var(T) represents the variance of temperature data.

5. The building energy consumption analysis method based on multimodal sensing according to claim 4 is characterized in that: The formula for extracting personnel flow characteristics in S3 is as follows: In the formula, Lr represents the personnel flow characteristics, R in represents the number of inflows in a specific time period, R out Indicates the number of outflows in a specific time period, R total Represents the total number of people in a specific time period.

6. The building energy consumption analysis method based on multimodal sensing according to claim 5 is characterized in that: The energy consumption change mathematical model established by S3 is as follows: Ek=k0+k1*St+k2*St sin +k3*St cos +k4*Tp+k5*Lr+∈ In the formula, Ek represents the energy consumption of the building, k0 represents the intercept term, that is, the energy consumption value predicted by the model when all other features are 0, k1 represents the coefficient of the time feature St, which represents the change in energy consumption Ek for every hour of increase. A positive value indicates that energy consumption increases with the passage of time, and vice versa. k2 represents the coefficient of the daily cycle sinusoidal feature, which is used to capture the periodic changes in energy consumption within a day, especially the impact on energy consumption in a specific period of time. When the coefficient is positive, it is considered that the energy consumption in this period of time is high. k3 represents the coefficient of the daily cycle cosine feature, k4 is the coefficient of the environmental feature, which represents the change in energy consumption for every unit change in temperature. The change in temperature will directly affect the air conditioning or heating demand of the building, thereby affecting energy consumption. k5 is the coefficient of the personnel flow feature, which represents the impact of personnel flow on energy consumption. When the number of staff in the building increases, energy consumption will increase. ∈ is the error term, which represents the part that the model cannot explain, including all other factors not included in the model, measurement errors and random noise.

7. The building energy consumption analysis method based on multimodal sensing according to claim 6 is characterized by: The formula for calculating the total energy consumption of a building in S4 is as follows: Lyzh=Cn+Zl+Zm+Sn In the formula, Lyzh represents the total energy consumption of the building, Cn represents heating energy consumption, which refers to the energy consumed to maintain the temperature in the building, Zl represents cooling energy consumption, which refers to the energy consumed to reduce the temperature in the building, Zm represents lighting energy consumption, which refers to the electric energy used for lighting, and Sn represents equipment energy consumption, which refers to the energy consumed by various electrical equipment.

8. The building energy consumption analysis method based on multimodal sensing according to claim 7 is characterized in that: The formula for calculating energy intensity of S4 is as follows: In the formula, Nqdk represents energy consumption intensity, Lyzh represents total energy consumption of the building, and Jzmj represents the total area of ​​the building.

9. The building energy consumption analysis method based on multimodal sensing according to claim 8 is characterized in that: The formula for calculating the energy efficiency ratio of S4 is as follows: In the formula, COP represents energy efficiency ratio, Scnl represents output energy, and Srdl represents input energy.

10. A building energy consumption analysis system based on multimodal sensing, characterized in that: It includes data acquisition module, data preprocessing module, feature extraction and modeling module, energy consumption analysis and prediction module, data visualization module and optimization module; The data acquisition module collects building environment data, energy consumption data, personnel flow data and equipment status data from multiple data sources; The data preprocessing module cleans the collected data, removes outliers and noise information, integrates the data collected at different time points through a time alignment method, and standardizes the data from different sources and dimensions; The feature extraction and modeling module performs feature extraction and modeling on the preprocessed data, extracts time features, environmental features and personnel flow features, and establishes a mathematical model of energy consumption changes; The energy consumption analysis and prediction module performs energy consumption analysis on the extracted features, calculates the total energy consumption of the building for energy efficiency evaluation, obtains energy consumption intensity and energy efficiency ratio, uses historical data for time series analysis, and predicts future building energy consumption trends; The data visualization module visualizes the analysis results and uses intuitive forms such as charts and heat maps to display the changing trend and distribution of energy consumption; The optimization module proposes energy-saving and emission-reduction suggestions and optimization decisions based on the visual analysis results.

Images