Building energy-saving intelligent management and control method and system based on Internet of Things

Abstract

The invention discloses a building energy-saving intelligent management and control method and system based on the Internet of Things, particularly relates to the technical field of intelligent building energy management, and is used for solving the problem that in the prior art, when building internal and external environment data are collected, resolution ratios of input data in the space-time dimension are not matched. The method comprises the following steps: acquiring Internet of Things sensor space-time distribution data of each area in a building; identifying space-time heterogeneity characteristics of data acquisition density and sampling frequency, and dynamically adjusting the sampling frequency of a low-frequency region; establishing a mapping relationship between the environmental parameters and the energy consumption parameters; traversing an instruction path and eliminating an inefficient instruction based on a conflict path weight; generating a regional differentiation energy-saving strategy and analyzing an electromagnetic interference distribution entropy value to insert an execution time difference; the optimized energy-saving strategy is executed; data acquisition optimization, energy consumption modeling precision and equipment instruction conflict resolution are realized, the reliability of an energy-saving strategy and the energy efficiency of the system are effectively improved, and the influence of electromagnetic interference on a building cable network is inhibited at the same time.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent building energy management. More specifically, the present invention relates to a building energy-saving intelligent control method and system based on the Internet of Things. Background Art

[0002] In an Internet-of-Things-based building energy-saving system, environmental data inside and outside the building is usually collected in real time by deploying various sensors (such as temperature and humidity, light, and people flow density sensors), and a machine learning model is used to predict the energy consumption trend to generate optimization strategies. However, due to the differences in building functional areas, sensor deployment is often concentrated in high-traffic areas (such as corridors and halls), while it is sparse in low-frequency activity areas (such as warehouses and spare meeting rooms). At the same time, there are significant differences in the data sampling frequencies of sensors in different areas due to hardware resource limitations or uneven communication bandwidth allocation, resulting in non-uniform distribution characteristics of data in terms of spatial coverage and temporal continuity.

[0003] When collecting environmental data inside and outside the building in the prior art, there is a problem of mismatched resolution of the input data of the machine learning model in the spatio-temporal dimension. Specifically, after the low-density data in the sparse area is mixed with the high-frequency data in the dense area, it is difficult for the model to accurately capture the local energy consumption characteristics, resulting in significant deviations in the prediction results in the low-density area. This regional distortion will further lead to over-execution or under-execution of the energy-saving strategy in some areas. For example, in low-density areas such as warehouses, reducing the lighting power excessively due to prediction errors may cause safety hazards, or in high-density areas, computational resources are wasted due to data redundancy, ultimately reducing the overall energy-saving effect. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present invention provide a building energy-saving intelligent control method and system based on the Internet of Things to solve the problems mentioned in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions: A building energy-saving intelligent control method based on the Internet of Things, comprising the following steps: S1. Obtain the spatio-temporal distribution data collected by the Internet of Things sensors in each area of the building; S2. Identify the spatio-temporal heterogeneity characteristics of the data collection density and sampling frequency in each area of the spatio-temporal distribution data; S3. Dynamically adjust the sampling frequency of the low-frequency area according to the spatio-temporal heterogeneity characteristics to synchronize the sampling frequency of the low-frequency area with that of the high-frequency area; S4. Establish a mapping relationship between the environmental parameters and energy consumption parameters of each area based on the synchronized spatio-temporal distribution data; S5. Traverse all the combined paths of device control instructions in the mapping relationship, identify the instruction conflict topology caused by physical space overlap, and eliminate inefficient instructions based on the conflict path weights; S6. Generate region-differentiated energy-saving strategies according to the mapping relationship after eliminating inefficient instructions, analyze the distribution entropy value of the electromagnetic pulse interference caused by the simultaneous start and stop of multi-region devices in the building cable network. If the distribution entropy value exceeds the preset entropy value, insert a time difference for strategy execution; S7. Execute the region-differentiated energy-saving strategies after inserting the time difference for strategy execution.

[0006] In a preferred embodiment, obtain the spatio-temporal distribution data collected by the Internet of Things sensors in each region of the building, including: Deploy temperature and humidity sensors and light sensors in the high-traffic areas of the building, and deploy people flow density sensors in the low-frequency activity areas; Collect the temperature and humidity data, light data in the high-traffic areas and the people flow density data in the low-frequency activity areas in real time based on the Internet of Things communication protocol; Bind the collected temperature and humidity data, light data and people flow density data according to the time stamp and location coordinates to generate spatio-temporal distribution data; Store the spatio-temporal distribution data in the distributed database of the edge computing node.

[0007] In a preferred embodiment, identify the spatio-temporal heterogeneity characteristics of the data collection density and sampling frequency in each region of the spatio-temporal distribution data, including: Divide the spatio-temporal distribution data into grid cells according to the region numbers, count the number of Internet of Things sensors in each grid cell, and generate a data collection density distribution map; Based on the time window, count the number of sensor data points in each region to generate a sampling frequency distribution map; Compare the data collection density distribution map with the sampling frequency distribution map to identify the spatio-temporal heterogeneity characteristics of density and frequency; Mark the spatio-temporal heterogeneity characteristics as high-frequency high-density regions, high-frequency low-density regions, low-frequency high-density regions and low-frequency low-density regions.

[0008] In a preferred embodiment, dynamically adjust the sampling frequency of the low-frequency regions according to the spatio-temporal heterogeneity characteristics to synchronize the sampling frequency of the low-frequency regions with that of the high-frequency regions, including: Obtain the sampling frequency distribution maps of the high-frequency high-density regions and high-frequency low-density regions, and extract the sampling frequency reference values of the high-frequency regions; According to the characteristic labels of the low-frequency low-density regions and low-frequency high-density regions, calculate the sampling frequency synchronization multiple between the low-frequency regions and the high-frequency regions; Adjust the sampling frequency of IoT sensors in the low-frequency area based on the sampling frequency synchronization multiple, so that the sampling frequency in the low-frequency area is consistent with the sampling frequency reference value in the high-frequency area; The adjusted sampling frequency instruction is sent to the IoT sensors in the low-frequency area, overwriting the original sampling configuration parameters.

[0009] In a preferred embodiment, a mapping relationship between environmental parameters and energy consumption parameters of each region is established based on the synchronized spatiotemporal distribution data, including: Fill missing values ​​and remove outliers in the synchronized spatiotemporal distribution data to generate a standardized spatiotemporal dataset; Extract environmental parameters and energy consumption parameters from the standardized spatiotemporal data set and correlate parameter pairs using the Pearson correlation coefficient; A multivariate linear regression model is constructed based on parameter pairs, and the model is trained to fit the mapping relationship between environmental parameters and energy consumption parameters; Verify the model fit. If the determination coefficient is greater than the preset threshold, the mapping relationship is stored in the relational database of the edge computing node.

[0010] In a preferred embodiment, traversing the combined paths of all device control instructions in the mapping relationship, identifying the instruction conflict topology caused by physical space overlap, and eliminating inefficient instructions based on the conflict path weights, including: Traverse the combined paths of all device control instructions in the mapping relationship to generate a device instruction path topology graph, in which each node represents a device instruction, and the connecting edges between nodes represent the time and space overlap relationship of instruction execution; Identify conflicting nodes in the device instruction path topology graph caused by physical space overlap. Specifically, if the device physical location coordinates corresponding to two nodes are in the same grid unit, and the ratio of the intersection duration of the instruction execution time window to the total duration of the time window exceeds a preset ratio, it is determined to be a conflicting node. The conflict path weight is generated based on the device type and energy consumption parameter of the conflicting node, and the conflict path weight is the product of the device type weight and the energy consumption parameter weight; If the conflict path weight is lower than the dynamic adjustment threshold, the corresponding inefficient instructions are eliminated and a conflict-free instruction set is generated.

[0011] In a preferred embodiment, the device type weight is preset according to the importance of the device function, and the energy consumption parameter weight is set based on the energy consumption influencing factor in the mapping relationship.

[0012] In a preferred embodiment, a regional differentiated energy-saving strategy is generated according to the mapping relationship after removing inefficient instructions, and the distribution entropy value of electromagnetic pulse interference caused by simultaneous start and stop of multi-region equipment in the building cable network is analyzed. If the distribution entropy value exceeds the preset entropy value, the strategy execution time difference is inserted, including: Based on a conflict-free instruction set, a region-differentiated energy-saving strategy is generated according to the device type weight and the cable node hierarchy, and the cable node hierarchy is divided according to the physical connection relationship between the grid unit and the building cable network topology; Analyze the electromagnetic pulse interference signals caused by the simultaneous start and stop of multi-region devices. Based on the signal propagation path of the cable node hierarchy, statistically analyze the distribution characteristics of electromagnetic pulse signals at the cable nodes, and calculate the distribution entropy value, which is quantified by the randomness dispersion degree of the signal intensity; If the distribution entropy value exceeds the preset entropy value, insert a policy execution time difference according to the cable node hierarchy and the device type weight: for the device instructions at the same cable node hierarchy, stagger the execution time windows from high to low according to the device type weight; the duration of the policy execution time difference is dynamically adjusted according to the electromagnetic load attenuation period of the cable node; Align the policy after inserting the policy execution time difference with the sampling frequency, update the region-differentiated energy-saving strategy and store it in the policy execution queue of the edge computing node.

[0013] In a preferred embodiment, executing the region-differentiated energy-saving strategy after inserting the policy execution time difference includes: Obtain the region-differentiated energy-saving strategy after inserting the policy execution time difference from the policy execution queue of the edge computing node and parse the device instruction, cable node hierarchy, and execution time difference fields; According to the cable node hierarchy and the device type weight, perform timing control on the device instructions at the same node hierarchy according to the policy execution time difference; Send instructions to the target device through the Internet of Things gateway; Record the instruction execution result and send it back to the edge computing node, and update the execution status field in the policy execution queue for optimizing the dynamic adjustment threshold.

[0014] On the other hand, the present invention provides an intelligent building energy-saving management and control system based on the Internet of Things, including the following modules: Data acquisition module: used to obtain the spatio-temporal distribution data collected by the Internet of Things sensors in each region of the building; Feature recognition module: used to identify the spatio-temporal heterogeneity features of the data acquisition density and sampling frequency in each region of the spatio-temporal distribution data; Frequency synchronization module: used to dynamically adjust the sampling frequency of the low-frequency region according to the spatio-temporal heterogeneity features, so that the sampling frequency of the low-frequency region is synchronized with the high-frequency region; Parameter mapping module: used to establish the mapping relationship between the environmental parameters and energy consumption parameters of each region based on the synchronized spatio-temporal distribution data; Conflict optimization module: used to traverse all the combined paths of the device control instructions in the mapping relationship, identify the instruction conflict topology caused by the physical space overlap, and eliminate the inefficient instructions based on the conflict path weight; Strategy generation module: used to generate region-differentiated energy-saving strategies based on the mapping relationship after eliminating inefficient instructions, analyze the distribution entropy value of electromagnetic pulse interference caused by the simultaneous start and stop of multi-region devices in the building cable network, and insert a time difference for strategy execution if the distribution entropy value exceeds the preset entropy value; Strategy execution module: used to execute the region-differentiated energy-saving strategy after inserting the time difference for strategy execution.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. Through the spatio-temporal heterogeneity feature recognition and dynamic adjustment mechanism, the regional imbalance problem of data collection density and sampling frequency in the building energy-saving scenario is effectively solved; through the heterogeneity analysis of data collection density and sampling frequency in each region, the sampling frequency of the low-frequency region is dynamically adjusted to be synchronized with the high-frequency region, ensuring that the spatio-temporal distribution data tends to be uniform in terms of spatial coverage and time continuity; enabling the subsequent established mapping relationship of energy consumption parameters to accurately reflect the true energy consumption characteristics of each region, avoiding model prediction deviations caused by data resolution differences, thereby improving the global reliability of the energy-saving strategy; at the same time, the parameter mapping model based on synchronous data quantifies the response law of energy consumption in different regions to environmental factors by coupling environmental parameters (such as temperature, humidity, and light) and energy consumption parameters (such as air-conditioning power and lighting energy consumption), providing a data basis for the generation of differentiated strategies; 2. By traversing the device instruction paths in the mapping relationship and introducing a conflict path weight optimization mechanism, the physical space conflict and superposition of electromagnetic interference problems during the execution of multi-region device instructions are solved; the rule for inserting the time difference for strategy execution based on the cable node hierarchy and device type weight can suppress the propagation of electromagnetic pulse interference while ensuring the reliable execution of high-priority device instructions; the dynamically adjusted time difference duration is adapted to the electromagnetic load attenuation period, avoiding the strategy rigidity caused by the traditional fixed time difference, thereby optimizing the overall energy efficiency on the premise of ensuring the operational safety of the device; through the synergy effect, the whole process management from data collection optimization to strategy execution control is realized, significantly improving the stability and resource utilization efficiency of the building energy-saving system. Brief Description of the Drawings

[0016] Figure 1 It is a flowchart of an intelligent building energy-saving control method based on the Internet of Things according to the present invention; Figure 2 It is a structural schematic diagram of an intelligent building energy-saving control system based on the Internet of Things according to the present invention. Detailed Embodiments

[0017] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0018] Embodiment 1: Figure 1 A building energy-saving intelligent control method based on the Internet of Things according to the present invention is provided, including the following steps: S1. Obtain the spatio-temporal distribution data collected by the Internet of Things sensors in each area of the building; S2. Identify the spatio-temporal heterogeneity characteristics of the data collection density and sampling frequency in each area in the spatio-temporal distribution data; S3. Dynamically adjust the sampling frequency of the low-frequency area according to the spatio-temporal heterogeneity characteristics to synchronize the sampling frequency of the low-frequency area with that of the high-frequency area; S4. Establish a mapping relationship between the environmental parameters and energy consumption parameters of each area based on the synchronized spatio-temporal distribution data; S5. Traverse all the combined paths of the device control instructions in the mapping relationship, identify the instruction conflict topology caused by physical space overlap, and eliminate inefficient instructions based on the conflict path weights; S6. Generate area-differentiated energy-saving strategies according to the mapping relationship after eliminating inefficient instructions, analyze the distribution entropy value of the electromagnetic pulse interference caused by the simultaneous start and stop of multi-area devices in the building cable network, and insert a strategy execution time difference if the distribution entropy value exceeds the preset entropy value; S7. Execute the area-differentiated energy-saving strategy after inserting the strategy execution time difference.

[0019] Deploy temperature and humidity sensors and light sensors in high-traffic areas of the building. High-traffic areas include, for example, corridors, lobbies, and elevator shafts. The temperature and humidity sensors and light sensors are evenly distributed, for example, at intervals of 10 meters, and the installation height is, for example, 2.5 meters above the ground. The temperature and humidity sensors are used to detect air temperature and relative humidity, and the light sensors are used to detect light intensity. Deploy people flow density sensors in low-frequency activity areas. Low-frequency activity areas include, for example, warehouses, spare meeting rooms, and equipment rooms. The people flow density sensors are installed above the door frames at the entrances of each area and use, for example, infrared sensing technology to detect the frequency of people entering and leaving. When people enter or leave the area, the people flow density sensors generate count signals. All sensors are connected to the Internet of Things gateway inside the building via, for example, wired or wireless means. The Internet of Things gateway supports, for example, the MQTT protocol or the CoAP protocol. The MQTT protocol is used for efficient data transmission in low-bandwidth environments, and the CoAP protocol is used for resource-constrained sensor devices. The Internet of Things gateway transmits the raw data collected by the sensors to the edge computing node. The edge computing node is, for example, an industrial-grade computer deployed on each floor of the building and has real-time data processing capabilities.

[0020] Collect the temperature and humidity data, light data in high-traffic areas, and people flow density data in low-frequency activity areas in real time based on the Internet of Things communication protocol. Specifically: The temperature and humidity sensors upload the current ambient temperature and humidity values, for example, once every 30 seconds, and the value range is, for example, temperature -20°C to 50°C, humidity 0% to 100%; the light sensors collect the light intensity data, for example, once every 10 seconds, and the data unit is, for example, lux; the people flow density sensors upload the count signals in real time when people enter and leave. The count signals include, for example, the number of entries and the number of exits. All sensor data is transmitted to the edge computing node through the Internet of Things gateway. The edge computing node verifies the data, and the verification rules include, for example, value range checks and timestamp continuity checks. For example, when the temperature data exceeds the range of -20°C to 50°C, it is marked as abnormal data.

[0021] Bind the collected temperature and humidity data, light data, and pedestrian flow density data to time stamps and location coordinates to generate spatio-temporal distribution data, specifically including: adding a time stamp to each sensor data, where the time stamp is based on accurate time information synchronized by, for example, the Network Time Protocol, and the time stamp format is, for example, "year-month-day hour:minute:second.millisecond", with a time error less than, for example, 1 millisecond; at the same time, binding the location coordinates to each data, where the location coordinates are determined according to the geographical location code preset during sensor deployment, and the geographical location code consists of, for example, the floor number, area number, and sensor type, and the coding rule is, for example, "F + floor number - area number - sensor type abbreviation", for example, "F1-A01-TH" represents the temperature and humidity sensor in area A01 on the first floor, and "F2-B02-LI" represents the light sensor in area B02 on the second floor. The data structure of the spatio-temporal distribution data is a triple containing a time stamp, location coordinates, and sensor value, and the triple is stored in, for example, JSON format, and the JSON fields include, for example, a time stamp, location coordinates, and a value, for example, the time stamp field is "timestamp", the location coordinate field is "location", and the value field is "value".

[0022] Store the spatio-temporal distribution data in the distributed database of the edge computing node, specifically: after the edge computing node receives the spatio-temporal distribution data, through, for example, the sharding storage mechanism of the distributed database, the data is sharded and stored in different physical storage nodes according to the floor number, and each storage node corresponds to, for example, one floor, for example, the data on the first floor is stored in node 1, and the data on the second floor is stored in node 2; each storage node adopts, for example, a time series database structure, and the time series database establishes an index with the time stamp as the primary key and supports fast retrieval of data according to the time range; the distributed database is, for example, the open source database InfluxDB or TimescaleDB, and is configured to perform data compression once every, for example, 5 minutes, and the data compression adopts, for example, a lossless compression algorithm, and the volume of the compressed data is reduced by, for example, 50% to 70%.

[0023] Divide the spatio-temporal distribution data into grid cells according to the regional numbers. The regional numbers are based on the preset geographical location codes in step S1. For example, when the geographical location code is "F1-A01-TH", the corresponding floor number is F1 and the regional number is A01. The rule for dividing grid cells is to divide the physical space corresponding to each regional number into square grids with a side length of, for example, 10 meters. Each grid cell corresponds to an independent data acquisition density calculation unit. Count the number of Internet of Things sensors in each grid cell. The number of Internet of Things sensors is determined by traversing the location coordinate fields in the spatio-temporal distribution data. For example, if there are 3 temperature and humidity sensors and 2 light sensors in grid cell F1-A01, the data acquisition density is 5 sensors. Generate a data acquisition density distribution map. The data acquisition density distribution map uses grid cells as the basic unit, and each unit is marked with the number of sensors and stored in, for example, a two-dimensional matrix format. The row number and column number of the matrix correspond to the longitude and latitude offsets of the grid cell.

[0024] Based on a time window, count the number of sensor data points in each region. The duration of the time window is, for example, 5 minutes. The start time of the time window is aligned with the nearest five-minute mark of the timestamp in step S1. For example, when the timestamp is "2023-10-01 08:30:00.500", the time window is from 08:30:00 to 08:35:00. When counting the number of sensor data points in each region, aggregate the total number of data points within the time window according to the regional numbers. For example, if region F1-A01 collects 30 temperature and humidity data and 50 light data within the time window from 08:30:00 to 08:35:00, the sampling frequency is 80 data points. Generate a sampling frequency distribution map. The sampling frequency distribution map uses the regional numbers as indexes and is stored in, for example, a key-value pair structure, where the key is the regional number and the value is the total number of data points within the corresponding time window.

[0025] Compare the data acquisition density distribution map with the sampling frequency distribution map to identify the spatio-temporal heterogeneity characteristics of density and frequency. Specifically: Associate the number of sensors in the data acquisition density distribution map with the number of data points in the sampling frequency distribution map according to the regional numbers, and calculate the deviation degree of density and frequency. The deviation degree calculation formula is, for example, (number of data points / number of sensors). For example, if the number of sensors in a certain region is 5 and the number of data points is 80, the deviation degree is 16. If the deviation degree is greater than, for example, a preset threshold of 10, it is determined as a high-frequency low-density region. If the deviation degree is less than, for example, a preset threshold of 5, it is determined as a low-frequency high-density region. The preset threshold is set according to, for example, the statistical results of historical data. For example, the average deviation degree of each region in the past 30 days is used as the threshold benchmark.

[0026] Mark the spatio-temporal heterogeneity features as high-frequency high-density regions, high-frequency low-density regions, low-frequency high-density regions, and low-frequency low-density regions, specifically: The high-frequency high-density region is the region where the deviation degree is greater than the threshold and the number of sensors is greater than, for example, the density threshold 5, and is marked as, for example, red; The high-frequency low-density region is the region where the deviation degree is greater than the threshold but the number of sensors is less than the density threshold, and is marked as, for example, yellow; The low-frequency high-density region is the region where the deviation degree is less than the threshold but the number of sensors is greater than the density threshold, and is marked as, for example, blue; The low-frequency low-density region is the region where both the deviation degree and the number of sensors are less than the threshold, and is marked as, for example, green. The marking result is stored as a feature label table, and the feature label table contains region numbers, density values, frequency values, and color labels, which are used for dynamic adjustment in subsequent steps.

[0027] Obtain the sampling frequency distribution maps of the high-frequency high-density region and the high-frequency low-density region. The marking results of the high-frequency high-density region and the high-frequency low-density region are based on the feature label table generated in step S2. The sampling frequency distribution maps are stored in a key-value pair structure, where the key is the region number and the value is the total number of data points within the time window; Extract the sampling frequency reference value of the high-frequency region. The sampling frequency reference value is the average sampling frequency of the high-frequency high-density region and the high-frequency low-density region. For example, if the sampling frequency of the high-frequency high-density region is 10 data points per second and the high-frequency low-density region is 5 data points per second, then the reference value is 7.5 data points per second.

[0028] According to the feature labels of the low-frequency low-density region and the low-frequency high-density region, calculate the sampling frequency synchronization multiple between the low-frequency region and the high-frequency region. The marking results of the low-frequency low-density region and the low-frequency high-density region are based on the feature label table generated in step S2. The sampling frequency synchronization multiple is the ratio of the sampling frequency reference value of the high-frequency region to the current sampling frequency of the low-frequency region. For example, if the current sampling frequency of the low-frequency region is 2 data points per second and the high-frequency reference value is 7.5 data points per second, then the synchronization multiple is 3.75 times; If the synchronization multiple contains a decimal, use, for example, the upward rounding rule. For example, 3.75 times is rounded up to 4 times.

[0029] Adjust the sampling frequency of the Internet of Things sensors in the low-frequency region based on the sampling frequency synchronization multiple to make the sampling frequency of the low-frequency region consistent with the sampling frequency reference value of the high-frequency region. Specifically: Multiply the original sampling frequency of the Internet of Things sensors in the low-frequency region by the synchronization multiple. For example, if the original sampling frequency is to collect once every 30 seconds and the synchronization multiple is 4 times, then the adjusted sampling frequency is to collect once every 7.5 seconds; If there is a deviation between the adjusted sampling frequency and the high-frequency region reference value, align it through, for example, a preset synchronization strategy. For example, align 7.5 seconds to the closest whole second number 8 seconds.

[0030] Send the adjusted sampling frequency instruction to the IoT sensors in the low-frequency region to overwrite the original sampling configuration parameters. Specifically: The edge computing node generates a configuration instruction containing the adjusted sampling frequency. The data structure of the configuration instruction is in JSON format, and the fields include region number, sensor type, new sampling frequency, and effective time. For example, the region number field is "region", the sensor type field is "sensor_type", the new sampling frequency field is "frequency", and the effective time field is "effective_time". The configuration instruction is transmitted to the target sensor through the IoT gateway. After receiving the instruction, the sensor updates the local configuration parameters and switches to the new sampling frequency at the effective time point.

[0031] Fill in the missing values and remove the outliers from the synchronized spatio-temporal distribution data to generate a standardized spatio-temporal dataset. Specifically: For missing value filling, the average value of the data in adjacent time windows within the same grid cell is used for completion. For example, if a temperature and humidity sensor has missing data in the time window from 08:30:00 to 08:35:00, then the average value of the data of this sensor in the time windows from 08:25:00 to 08:30:00 and from 08:35:00 to 08:40:00 is taken as the filling value. For outlier removal, the standard deviation method is used. If the deviation of a data point from the mean value of the data in the grid cell exceeds three standard deviations, it is determined as an outlier and removed. For example, if the average temperature of a certain region is 25°C and the standard deviation is 2°C, then the data points exceeding 31°C or lower than 19°C are removed.

[0032] Extract the environmental parameters and energy consumption parameters from the standardized spatio-temporal dataset. The environmental parameters include temperature, humidity, and light intensity, and the energy consumption parameters include air-conditioning energy consumption, lighting energy consumption, and equipment operating power. Correlate parameter pairs through the Pearson correlation coefficient. Specifically: Calculate the ratio of the covariance to the standard deviation of the environmental parameters and energy consumption parameters. The covariance is calculated as the average value of the product of the deviation of the parameter value from its mean within the same time window, and the standard deviation is the square root of the variance of the parameter value. If the absolute value of the Pearson coefficient is greater than 0.5, it is determined as a strongly correlated parameter pair. The threshold of 0.5 is set based on the statistical results of historical data. For example, in the energy consumption prediction, the accuracy rate of the parameter pairs with a correlation coefficient greater than 0.5 in the past 90 days exceeds 80%.

[0033] Build a multiple linear regression model based on the parameter pairs and train the model to fit the mapping relationship between the environmental parameters and energy consumption parameters. Specifically: Use the strongly correlated environmental parameters as independent variables and the energy consumption parameters as dependent variables to build a multiple linear equation. For example, air-conditioning energy consumption = α × temperature + β × humidity + γ × light intensity + intercept term. Solve the coefficients α, β, γ, and the intercept term through the least squares method. The least squares method determines the optimal parameters by minimizing the square difference between the predicted value and the actual value.

[0034] Verify the model fitting degree. If the coefficient of determination is greater than the preset threshold, store the mapping relationship in the relational database of the edge computing node. Specifically: The coefficient of determination R² is obtained by calculating the variance ratio of the predicted energy consumption to the actual energy consumption. For example, R² = 1 - (variance of prediction error / variance of actual energy consumption); the preset threshold is 0.7, and this threshold is set according to the historical model verification results. For example, when R² ≥ 0.7, the model prediction error is less than 15%; if R² ≥ 0.7, it is determined that the model fitting is effective, and store the coefficients and intercept terms of the multiple linear equation in the relational database; if R² < 0.7, reselect the parameter pairs or adjust the model parameters.

[0035] Traverse all the combined paths of device control instructions in the mapping relationship established in step S4 to generate a device instruction path topology graph. Each node in the device instruction path topology graph represents a device control instruction, and the connecting edges between nodes represent the time and space overlap relationships of instruction execution. The time overlap relationship is determined by the intersection of the instruction execution time windows, and the space overlap relationship is determined by the grid cell to which the device physical location coordinates belong. The definition of the grid cell is the same as the grid cell divided in step S2, and the definition of the time window is the same as the time window for statistical sampling frequency in step S2. The device control instructions include air conditioner start / stop instructions, lighting power adjustment instructions, and standby device switch instructions. The execution time and location coordinates of each instruction are extracted from the spatio-temporal distribution data in step S1.

[0036] Identify the conflict nodes in the device instruction path topology graph caused by physical space overlap. Specifically: If the device physical location coordinates corresponding to two nodes belong to the same grid cell, and the proportion of the intersection duration of the instruction execution time windows in the total duration of the time window exceeds the preset proportion, it is determined as a conflict node. For example, when the proportion of the intersection duration of the instruction execution time windows of two air conditioner devices exceeds the preset proportion, it is determined as a conflict. The overlap determination of the grid cell is based on the grid cell boundary divided in step S2, and the proportion of the intersection duration of the time window is determined by calculating the ratio of the intersection duration to the total duration of the time window. The device types of the conflict nodes include air conditioners, lighting, and standby devices, and the device types correspond to the types of IoT sensors deployed in step S1.

[0037] Generate conflict path weights based on the device type and energy consumption parameters of conflict nodes. The device type weight is preset according to the importance of device functions, and the importance of device functions is determined by the criticality of the device in building operation and maintenance. For example, the weight of air-conditioning equipment is higher than that of lighting equipment, and the weight of lighting equipment is higher than that of standby equipment. The energy consumption parameter weight is set based on the energy consumption impact factor in the mapping relationship of step S4, and the energy consumption impact factor is the contribution degree of environmental parameters to energy consumption. For example, the contribution degree of temperature to air-conditioning energy consumption is determined by the coefficient in the multiple linear regression model of step S4. The conflict path weight is the product of the device type weight and the energy consumption parameter weight. For example, when the device type weight of air-conditioning equipment is high and the contribution degree of temperature to air-conditioning energy consumption is large, the conflict path weight increases significantly.

[0038] If the conflict path weight is lower than the dynamic adjustment threshold, eliminate the corresponding inefficient instructions. The dynamic adjustment threshold is updated in real time according to the historical instruction execution success rate, and the historical instruction execution success rate is calculated by statistically analyzing the matching ratio of the actual energy consumption and the predicted energy consumption after the instruction execution in the past period. For example, when the historical instruction execution success rate is lower than the preset level, the dynamic adjustment threshold is increased to strengthen the elimination condition. The updated device instruction path topology diagram removes the nodes and connection edges with conflict path weights lower than the dynamic adjustment threshold to generate a conflict-free instruction set, which is used for generating the regional differential energy-saving strategy in step S6.

[0039] In step S5, by traversing the combined paths of device control instructions, identify the instruction conflict topology caused by physical space overlap, and eliminate inefficient instructions based on the conflict path weights. Compared with the prior art that only relies on a single spatial or temporal dimension to determine conflicts, this step accurately identifies the spatio-temporal coupling conflicts of device instructions by combining the grid cell (space) and the time window overlap rate (time), avoiding resource waste or equipment overload caused by instruction conflicts. At the same time, the conflict path weight is dynamically generated by the importance of device type functions (preset rules) and the energy consumption impact factor (output of the dynamic model), replacing the rigid strategies of traditional fixed priorities or energy consumption thresholds, and realizing the adaptive optimization of the strategy. Through cross-step data coupling and closed-loop feedback mechanisms, adapt to the building operation state in real time, solve the energy efficiency loss problem caused by strategy lag and static rules in traditional solutions, and significantly improve the conflict identification accuracy and strategy execution reliability through spatio-temporal collaborative determination and dynamic weight calculation, ultimately optimizing the overall energy-saving effect.

[0040] Based on the conflict-free instruction set generated in step S5, generate a region-differentiated energy-saving strategy according to the device type weight and the cable node level. The device type weight is the weight coefficient preset according to the importance of device functions in step S5. The importance of device functions is determined by the criticality of the device in building operation and maintenance. For example, the air conditioning device is given the highest weight because it directly affects indoor temperature control, the lighting device has the second highest weight due to safety requirements, and the standby device has the lowest weight. The cable node level is divided according to the physical connection relationship between the grid unit and the building cable network topology. For example, the core cable node connecting multiple grid units is marked as a first-level node, and the cable node connecting only a single grid unit is marked as a second-level node. The generation rule of the region-differentiated energy-saving strategy is: the instructions with high device type weights corresponding to the first-level nodes are executed first. For example, the air conditioning start / stop instruction of the first-level node takes precedence over the lighting adjustment instruction of the second-level node. When generating the strategy, the device position coordinates in the spatio-temporal distribution data of step S1 and the energy consumption parameters in the mapping relationship of step S4 are called to ensure the data consistency between the device instructions and the energy consumption model.

[0041] Analyze the electromagnetic pulse interference signals caused by the simultaneous start / stop of multi-region devices, and statistically analyze the distribution characteristics of the electromagnetic pulse signals at the cable nodes based on the signal propagation paths at the cable node levels. The signal propagation paths are determined according to the cable node levels and the grid unit positions. For example, the signal propagation path of the first-level node covers all the grid units it connects, and the signal propagation path of the second-level node is limited to its own grid unit. The electromagnetic pulse signal intensity is collected by the current sensors deployed at the cable nodes, and the sensor data is transmitted to the edge computing nodes through the Internet of Things gateway in step S1. The distribution entropy value is quantified by the randomness dispersion degree of the signal intensity. Specifically, it is to statistically analyze the fluctuation range of the signal intensities of each frequency band at the cable nodes within the same time window. The fluctuation range is calculated by the difference between the maximum value and the minimum value of the signal intensity. For example, when the signal intensity at a certain cable node fluctuates greatly within the time window, it is determined as a high distribution entropy value. The preset entropy value is set according to the statistical data of historical electromagnetic interference. For example, the average value of the distribution entropy values under normal operating conditions in the past is statistically analyzed as the benchmark.

[0042] If the distribution entropy value exceeds the preset entropy value, a time difference is inserted according to the cable node level and the device type weight insertion strategy. Specifically, the device instructions at the same cable node level are staggered in the execution time window according to the device type weight from high to low. The sorting rule of the device type weight from high to low is that the air-conditioning device takes precedence over the lighting device, and the lighting device takes precedence over the standby device. For example, the air-conditioning instruction of the first-level node is executed in the first half of the time window, and the lighting instruction of the same node is executed in the second half. The duration of the time difference for the strategy execution is dynamically adjusted according to the electromagnetic load attenuation period of the cable node. The electromagnetic load attenuation period is calculated through the cable specification parameters and historical electromagnetic interference data. For example, the attenuation period of the high-frequency interference signal is shorter, and the corresponding time difference is shortened. The strategy after inserting the time difference is aligned with the sampling frequency adjusted in step S3 to ensure that the device instruction execution time window is synchronized with the data acquisition frequency.

[0043] Update and store the region-differentiated energy-saving strategy after inserting the time difference into the strategy execution queue of the edge computing node. The data structure of the strategy execution queue is compatible with the spatio-temporal distribution data format in step S1, including fields such as timestamp, cable node level, device type weight, and execution time difference. The timestamp field is in the same format as the timestamp in step S1. The cable node level field is associated with the grid cell encoding. The device type weight field references the conflict path weight result in step S5. The updated strategy execution queue is used for the execution and feedback of the region-differentiated energy-saving strategy in step S7 to form a technical closed-loop from strategy generation to execution.

[0044] It should be noted that the importance of the device function is preset according to the building operation and maintenance requirements. For example, the air conditioner is given the highest weight because it directly affects the comfort of personnel and has the largest energy consumption ratio. The lighting device has the second highest weight due to the safety lighting requirements, and the standby device has the lowest weight because of its lowest usage frequency. The weight coefficient is determined through expert experience or the analysis of the impact of device failures in historical operation and maintenance logs. For example, the weight of the air conditioner is 0.8, the lighting is 0.5, and the standby device is 0.3.

[0045] The first-level node is the core cable node connecting multiple grid cells, such as the main building distribution cabinet or the intersection of the main cables. Its signal propagation range covers multiple areas. The second-level node is the end cable node that only connects a single grid cell, such as the power supply node of an independent meeting room or a warehouse. The basis for the level division is the mapping relationship between the building electrical wiring diagram and the grid cell to ensure that the physical connection relationship is traceable.

[0046] The distribution entropy value is quantified by the signal intensity fluctuation range. The larger the fluctuation range, the higher the randomness of the electromagnetic interference. For example, when the signal intensity suddenly increases from 100 mV to 500 mV within a short time window, the fluctuation range of 400 mV corresponds to a high entropy value. The preset fluctuation threshold is set according to the maximum fluctuation value under normal working conditions in historical data. For example, after statistically analyzing the data of the past 30 days, the maximum fluctuation value of 300 mV is taken as the threshold.

[0047] The electromagnetic load attenuation period is calculated through the impedance characteristics of the cable. The impedance characteristics include cable resistance, inductance, and capacitance parameters. For example, high-frequency interference attenuates faster due to the higher inductive reactance of the cable, while low-frequency interference attenuates slower due to the influence of capacitive reactance. The time difference duration is dynamically mapped according to the attenuation period. For example, if the attenuation period of high-frequency interference is 5 seconds, the time difference is set to 5 seconds; if the attenuation period of low-frequency interference is 20 seconds, the time difference is set to 20 seconds.

[0048] The timestamp field of the policy execution queue is strictly consistent with the timestamp format of step S1, such as "year-month-day hour:minute:second.millisecond"; the cable node hierarchy field references the geographical location encoding of the grid cell. For example, the first-level node encoding is "F1-MAIN", and the second-level node is "F2-B01". Data structure compatibility ensures that historical spatio-temporal data can be called for anomaly detection and feedback optimization during policy execution.

[0049] Step S6 generates a region-differentiated energy-saving policy by combining the device type weight and the cable node hierarchy, analyzes the distribution entropy value of the electromagnetic pulse interference caused by the simultaneous start and stop of multi-region devices, and inserts a dynamic time difference when the entropy value exceeds the limit, so as to solve the problems of electromagnetic interference accumulation and energy efficiency loss caused by device instruction conflicts in the traditional solution. Compared with the existing technology, first, based on the coupling strategy of the cable node hierarchy and the device type weight, it accurately identifies high-interference risk areas and preferentially schedules key devices to suppress the propagation of electromagnetic pulse signals in the cable network; second, the dynamic time difference adjustment mechanism is based on the electromagnetic load attenuation period of the cable node, adapts to the physical attenuation characteristics of different frequency band interferences, and avoids the policy rigidity caused by a fixed time difference; third, the distribution entropy value quantization method evaluates the interference risk through the randomness dispersion degree of the signal intensity, replaces the traditional threshold determination, and improves the detection accuracy of hidden interference. Through the synergistic effect, the synchronous optimization of electromagnetic compatibility and energy-saving efficiency is achieved, and the dynamic control logic of cross-step data coupling and physical rule driving is realized.

[0050] Obtain the region-differentiated energy-saving policy with the inserted policy execution time difference from the policy execution queue of the edge computing node and parse the device instruction, cable node hierarchy, and execution time difference fields. The data structure of the policy execution queue is compatible with the spatio-temporal distribution data format of step S1. The device instruction field contains the device type, start / stop operation, and target parameters. The cable node hierarchy field references the first-level and second-level node encodings divided. The execution time difference field is the time interval dynamically adjusted in step S6. When parsing, match the physical location of the device according to the cable node hierarchy field. For example, the first-level node encoding "F1-MAIN" corresponds to the devices in the main power distribution cabinet area, and the second-level node encoding "F2-B01" corresponds to the devices in the B01 area of the second floor.

[0051] According to the cable node level and device type weight, perform timing control on device instructions at the same node level according to the time difference of policy execution. The cable node level is divided into primary nodes and secondary nodes. The device type weight is the weight coefficient preset according to the importance of device functions in step S5. For example, the weight of air conditioning equipment is higher than that of lighting equipment. The timing control rule is: device instructions at the same cable node level are sorted from high to low according to the device type weight. High-weight devices are executed in the first half of the time window, and low-weight devices are executed in the second half. For example, the air conditioning equipment instruction at the primary node is executed in the time window 08:00 - 08:02, and the lighting equipment instruction at the same node is executed at 08:03 - 08:05. The execution time difference field defines the interval duration between adjacent device instructions. For example, the interval between the air conditioning and lighting instructions is 3 seconds to ensure that the electromagnetic pulse interference decays to a safe threshold.

[0052] Send instructions to the target device through the Internet of Things gateway, and the instruction execution time window is aligned with the synchronous sampling frequency adjusted in step S3. The synchronous sampling frequency is the device data acquisition frequency dynamically adjusted in step S3. For example, the sampling frequency of air conditioning equipment is once every 8 seconds, then the air conditioning instruction execution time window is allocated at 8-second intervals. The Internet of Things gateway matches the communication protocol of the target device according to the device type field. For example, air conditioning equipment uses the Modbus protocol, and lighting equipment uses the ZigBee protocol to ensure that the instruction format is compatible with the device interface. When sending instructions, verify the consistency of the cable node level and the physical location of the device. For example, verify whether the device corresponding to the primary node code "F1-MAIN" is located in the main distribution cabinet area to avoid mis-sending instructions.

[0053] Record the instruction execution result and send it back to the edge computing node, and update the execution status field in the policy execution queue for optimizing the dynamic adjustment threshold. The execution result includes the actual instruction execution time, device response status, and electromagnetic pulse interference intensity. The electromagnetic pulse interference intensity is collected by the current sensor of the cable node and transmitted to the edge computing node. The update rule of the execution status field is: if the electromagnetic pulse interference intensity is lower than the preset safety threshold after the instruction is executed, mark it as "execution successful"; if it exceeds the threshold, mark it as "execution abnormal". The optimization of the dynamic adjustment threshold is based on the statistical results of historical execution. For example, count the proportion of "execution successful" instructions in the past 24 hours. If the proportion is lower than the preset level, increase the dynamic adjustment threshold in step S6 to enhance the interference suppression condition. The updated policy execution queue is used for the next round of policy generation in step S6, forming a closed-loop control from policy execution to feedback optimization.

[0054] Embodiment 2: Figure 2 A structural schematic diagram of an Internet of Things-based intelligent building energy management and control system of the present invention is given. An Internet of Things-based intelligent building energy management and control system includes the following modules: Data acquisition module: used to obtain the spatio-temporal distribution data collected by Internet of Things sensors in each area of the building; Feature recognition module: used to identify the spatio-temporal heterogeneity features of data acquisition density and sampling frequency in each area of the spatio-temporal distribution data; Frequency synchronization module: used to dynamically adjust the sampling frequency of low-frequency areas according to the spatio-temporal heterogeneity features, so that the sampling frequency of low-frequency areas is synchronized with that of high-frequency areas; Parameter mapping module: used to establish the mapping relationship between the environmental parameters and energy consumption parameters of each area based on the synchronized spatio-temporal distribution data; Conflict optimization module: used to traverse all the combined paths of device control instructions in the mapping relationship, identify the instruction conflict topology caused by physical space overlap, and eliminate inefficient instructions based on the conflict path weight; Strategy generation module: used to generate area-differentiated energy-saving strategies according to the mapping relationship after eliminating inefficient instructions, analyze the distribution entropy value of electromagnetic pulse interference caused by the simultaneous start and stop of multi-area devices in the building cable network, and insert a strategy execution time difference if the distribution entropy value exceeds the preset entropy value; Strategy execution module: used to execute the area-differentiated energy-saving strategies after inserting the strategy execution time difference.

[0055] The calculations involved in the embodiments are all dimensionless numerical calculations, and the preset parameters and threshold selections in the calculations are set by those skilled in the art according to the actual situation.

[0056] It should be noted that the present invention can be deployed on the device itself to achieve embedded applications, or can also run on a PC or other terminal with a user interface, so as to meet various hardware environments and usage requirements.

[0057] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that contains one or more collections of available media. The available media can be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0058] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, devices, and modules described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0059] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there can be other division methods in actual implementation. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of the devices or modules can be in electrical, mechanical, or other forms.

[0060] The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules. They can be located in one place or distributed to multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0061] In addition, in each embodiment of the present application, each functional module can be integrated into a processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0062] If the above-mentioned function is implemented in the form of a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present application. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0063] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0064] Finally: The above is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An intelligent control method for building energy conservation based on the Internet of Things, characterized in that, It includes the following steps: S1. Obtain the spatio-temporal distribution data collected by the Internet of Things sensors in each area of the building; S2. Identify the spatio-temporal heterogeneity characteristics of the data collection density and sampling frequency in each area in the spatio-temporal distribution data; S3. Dynamically adjust the sampling frequency of the low-frequency area according to the spatio-temporal heterogeneity characteristics to synchronize the sampling frequency of the low-frequency area with that of the high-frequency area; S4. Establish the mapping relationship between the environmental parameters and energy consumption parameters of each area based on the synchronized spatio-temporal distribution data; S5. Traverse all the combined paths of the device control instructions in the mapping relationship, identify the instruction conflict topology caused by physical space overlap, and eliminate inefficient instructions based on the conflict path weights; S6. Generate area-differentiated energy-saving strategies according to the mapping relationship after eliminating inefficient instructions, analyze the distribution entropy value of the electromagnetic pulse interference caused by the simultaneous start and stop of multi-area devices in the building cable network. If the distribution entropy value exceeds the preset entropy value, insert a time difference for strategy execution; S7. Execute the area-differentiated energy-saving strategy after inserting the time difference for strategy execution.

2. The method for intelligent control of building energy conservation based on the Internet of Things according to claim 1, characterized in that, Obtain the spatio-temporal distribution data collected by the Internet of Things sensors in each area of the building, including: Deploy temperature and humidity sensors and light sensors in the high-traffic areas of the building, and deploy people flow density sensors in the low-frequency activity areas; Real-time collect the temperature and humidity data, light data in the high-traffic areas and the people flow density data in the low-frequency activity areas based on the Internet of Things communication protocol; Bind the collected temperature and humidity data, light data and people flow density data according to the time stamp and location coordinates to generate spatio-temporal distribution data; Store the spatio-temporal distribution data in the distributed database of the edge computing node.

3. The intelligent control method for building energy conservation based on the Internet of Things according to claim 1, characterized in that, Identify the spatio-temporal heterogeneity characteristics of the data collection density and sampling frequency in each area in the spatio-temporal distribution data, including: Divide the spatio-temporal distribution data into grid cells according to the area number, count the number of Internet of Things sensors in each grid cell, and generate a data collection density distribution map; Based on the time window, count the number of sensor data points in each area to generate a sampling frequency distribution map; Compare the data collection density distribution map with the sampling frequency distribution map to identify the spatio-temporal heterogeneity characteristics of density and frequency; Mark the spatio-temporal heterogeneity characteristics as high-frequency high-density areas, high-frequency low-density areas, low-frequency high-density areas and low-frequency low-density areas.

4. The intelligent control method for building energy conservation based on the Internet of Things according to claim 1, characterized in that Dynamically adjust the sampling frequency of the low-frequency area according to the spatio-temporal heterogeneity characteristics to synchronize the sampling frequency of the low-frequency area with that of the high-frequency area, including: Obtain the sampling frequency distribution maps of the high-frequency high-density areas and high-frequency low-density areas, and extract the sampling frequency reference value of the high-frequency area; According to the characteristic labels of the low-frequency low-density areas and low-frequency high-density areas, calculate the sampling frequency synchronization multiple between the low-frequency area and the high-frequency area; Based on the sampling frequency synchronization multiple, adjust the sampling frequency of the Internet of Things sensors in the low-frequency area to make the sampling frequency of the low-frequency area consistent with the sampling frequency reference value of the high-frequency area; Send the adjusted sampling frequency instruction to the Internet of Things sensors in the low-frequency area to overwrite the original sampling configuration parameters.

5. A building energy-saving intelligent control method based on the Internet of Things according to claim 1, characterized in that, Establish the mapping relationship between the environmental parameters and energy consumption parameters of each area based on the synchronized spatio-temporal distribution data, including: Fill in the missing values and eliminate the outliers in the synchronized spatio-temporal distribution data to generate a standardized spatio-temporal data set; Extract the environmental parameters and energy consumption parameters from the standardized spatio-temporal dataset, and correlate parameter pairs through the Pearson correlation coefficient; Construct a multiple linear regression model based on the parameter pairs, and train the model to fit the mapping relationship between the environmental parameters and the energy consumption parameters; Verify the model fitting degree. If the coefficient of determination is greater than the preset threshold, store the mapping relationship in the relational database of the edge computing node.

6. The intelligent control method for building energy conservation based on the Internet of Things according to claim 1, characterized in that, Traverse all the combined paths of the device control instructions in the mapping relationship, identify the instruction conflict topology caused by physical space overlap, and eliminate inefficient instructions based on the conflict path weight, including: Traverse all the combined paths of the device control instructions in the mapping relationship to generate a device instruction path topology diagram. Each node in the device instruction path topology diagram represents a device instruction, and the connection edges between the nodes represent the time and space overlap relationships of the instruction execution; Identify the conflict nodes in the device instruction path topology diagram caused by physical space overlap. Specifically: if the physical location coordinates of the devices corresponding to two nodes are within the same grid cell, and the proportion of the intersection duration of the instruction execution time windows in the total duration of the time windows exceeds the preset proportion, it is determined as a conflict node; Generate the conflict path weight based on the device type and energy consumption parameters of the conflict nodes. The conflict path weight is the product of the device type weight and the energy consumption parameter weight; If the conflict path weight is lower than the dynamically adjusted threshold, eliminate the corresponding inefficient instructions and generate a conflict-free instruction set.

7. A building energy-saving intelligent control method based on the Internet of Things according to claim 6, characterized in that, The device type weight is preset according to the importance of the device function, and the energy consumption parameter weight is set based on the energy consumption impact factor in the mapping relationship.

8. A building energy-saving intelligent control method based on the Internet of Things according to claim 1, characterized in that, Generate a region-differentiated energy-saving strategy according to the mapping relationship after eliminating inefficient instructions. Analyze the distribution entropy value of the electromagnetic pulse interference caused by the simultaneous start and stop of multi-region devices in the building cable network. If the distribution entropy value exceeds the preset entropy value, insert a strategy execution time difference, including: Based on the conflict-free instruction set, generate a region-differentiated energy-saving strategy according to the device type weight and the cable node level. The cable node level is divided according to the physical connection relationship between the grid cell and the building cable network topology; Analyze the electromagnetic pulse interference signal caused by the simultaneous start and stop of multi-region devices. Based on the signal propagation path of the cable node level, statistically analyze the distribution characteristics of the electromagnetic pulse signal at the cable nodes, and calculate the distribution entropy value. The distribution entropy value is quantified by the randomness dispersion degree of the signal intensity; If the distribution entropy value exceeds the preset entropy value, insert a strategy execution time difference according to the cable node level and the device type weight: for the device instructions at the same cable node level, stagger the execution time windows from high to low according to the device type weight; the duration of the strategy execution time difference is dynamically adjusted according to the electromagnetic load attenuation period of the cable node; Align the strategy after inserting the strategy execution time difference with the sampling frequency, update the region-differentiated energy-saving strategy and store it in the strategy execution queue of the edge computing node.

9. The method for intelligent control of building energy conservation based on the Internet of Things according to claim 1, characterized in that, Execute the region-differentiated energy-saving strategy after inserting the strategy execution time difference, including: Obtain the region-differentiated energy-saving strategy after inserting the strategy execution time difference from the strategy execution queue of the edge computing node and parse the device instruction, cable node level, and execution time difference fields; According to the cable node level and the device type weight, perform timing control on the device instructions at the same node level according to the strategy execution time difference; Send instructions to the target device through the Internet of Things gateway; Record the execution results of the instructions and send them back to the edge computing node, and update the execution status field in the policy execution queue for dynamic adjustment of threshold optimization.

10. An Internet of Things-based intelligent control system for building energy conservation, which is used to implement an Internet of Things-based intelligent control method for building energy conservation described in any one of claims 1-9, characterized in that, It includes the following modules: Data acquisition module: used to obtain the spatio-temporal distribution data collected by the Internet of Things sensors in each area of the building; Feature recognition module: used to identify the spatio-temporal heterogeneity features of data acquisition density and sampling frequency in each area of the spatio-temporal distribution data; Frequency synchronization module: used to dynamically adjust the sampling frequency of the low-frequency area according to the spatio-temporal heterogeneity features to synchronize the sampling frequency of the low-frequency area with that of the high-frequency area; Parameter mapping module: used to establish the mapping relationship between the environmental parameters and energy consumption parameters of each area based on the synchronized spatio-temporal distribution data; Conflict optimization module: used to traverse all the combined paths of device control instructions in the mapping relationship, identify the instruction conflict topology caused by physical space overlap, and eliminate inefficient instructions based on the conflict path weights; Policy generation module: used to generate area-differentiated energy-saving policies according to the mapping relationship after eliminating inefficient instructions, analyze the distribution entropy value of the electromagnetic pulse interference caused by the simultaneous start and stop of multi-area devices in the building cable network, and insert the policy execution time difference if the distribution entropy value exceeds the preset entropy value; Policy execution module: used to execute the area-differentiated energy-saving policy after inserting the policy execution time difference.

Images