Personalized energy-saving management and control method and device based on GRU model and medium

By adopting the personalized energy-saving control method of GRU model in smart buildings, combining real-time energy consumption and flow data, an energy consumption prediction model is built, which solves the problem of misjudgment of energy consumption under general analysis standards, and achieves refined management and efficient energy saving control.

CN118917972APending Publication Date: 2024-11-08山东浪潮智慧建筑科技有限公司
View PDF 0 Cites 6 Cited by

Patent Information

Application Number
CN202410942139.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-15
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

In smart building application scenarios with large scale and time-change cycles of people flow, the general energy consumption data analysis standards ignore the differences in energy consumption standards between different regions, resulting in misjudgment or misjudgment of equipment energy consumption, and energy-saving control poor targeted and accurate.

Method used

Using a personalized energy-saving control method based on the GRU model, the equipment energy consumption monitoring strategy of the monitoring sub-region is determined by collecting energy consumption data and flow data in real time, the GRU area energy consumption prediction model is constructed, the current energy-saving reference data is generated, and the abnormal reference is based on Information development of targeted energy-saving control strategies.

Benefits of technology

It has achieved refined management of energy consumption demands in different regions, timely detection of energy consumption abnormalities, formulate targeted and timely energy-saving control strategies, reduce energy waste, and improve energy saving efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118917972A_ABST
    Figure CN118917972A_ABST
Patent Text Reader

Abstract

The embodiment of the invention discloses a personalized energy-saving management and control method and device based on a GRU model and a medium, and relates to the technical field of smart buildings, and the method comprises the steps: collecting the real-time energy consumption data of a plurality of energy consumption devices of a smart building, determining a monitoring sub-region corresponding to each piece of real-time energy consumption data, obtaining the real-time people flow data of the monitoring sub-region, and obtaining the real-time people flow data of the monitoring sub-region; determining equipment energy consumption monitoring strategies corresponding to the monitoring sub-regions, wherein the equipment energy consumption monitoring strategies comprise a lowest energy consumption operation strategy, a comfort guarantee operation strategy and an emergency high energy consumption operation strategy; determining a GRU area energy consumption prediction model through an equipment energy consumption monitoring strategy so as to determine corresponding current energy-saving reference data, and monitoring a plurality of energy consumption equipment; on the basis of a GRU equipment energy consumption prediction model, performing trend prediction on energy consumption of each piece of equipment with abnormal energy consumption, and determining predicted energy consumption information of each piece of equipment with abnormal energy consumption; and according to the predicted energy consumption information and the current abnormal reference information, generating an energy-saving control strategy to realize energy-saving management and control.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of smart building technology, and in particular to a personalized energy-saving control method, device and medium based on a GRU model. Background Art

[0002] Smart Building is a form of building that realizes intelligent management and service of buildings by integrating advanced information technology, Internet of Things, big data and artificial intelligence. Smart buildings have a wide range of application scenarios, including smart hotels, smart office buildings, smart shopping malls, smart factories, smart hospitals, etc. When performing energy-saving control on smart buildings, it is usually necessary to monitor the energy consumption in the building in real time, understand the energy consumption of each device, analyze the collected energy consumption data, identify energy waste and unreasonable consumption points, and perform targeted intelligent control of equipment.

[0003] For large-scale buildings, there are many devices in the building. The energy consumption data analysis of each device generates a lot of energy consumption data analysis workload. In addition, the energy consumption judgment standards of equipment in different areas are different. The use of general judgment rules for the same type of equipment is prone to misjudgment or omission of abnormal energy consumption. In addition, in the application scenarios of smart buildings, most of them have the characteristics of human flow, and there is a certain time change cycle in the flow of human flow. If a general energy consumption identification method is used, it is also easy to misjudgment or omission of abnormal energy consumption when the flow of human flow is small or the flow of human flow increases suddenly. Due to the different functions of different areas inside the building, such as office areas, factories, computer rooms, reception areas and other functional areas, the working requirements of energy-consuming equipment in each area are different, and there are personalized differences in energy consumption standards between different areas inside the building. After the misjudgment or omission of energy consumption, it is impossible to obtain the abnormal energy consumption equipment in the building in a timely and accurate manner, resulting in inaccurate control objects and control schemes in the subsequent energy-saving control process, and hidden dangers in the energy-saving control process.

[0004] Therefore, in smart building application scenarios where the scale is large and the flow of people has a time-varying cycle, energy consumption data analysis is performed on each type of equipment according to the same standards, ignoring the differences in energy consumption standards between energy-consuming equipment and different areas within the building, which can easily lead to misjudgment or omission of equipment energy consumption, and poor targeting and accuracy of energy-saving management and control. Summary of the invention

[0005] One or more embodiments of the present specification provide a personalized energy-saving management and control method, device and medium based on the GRU model, which are used to solve the following technical problems: in a large-scale smart building application scenario with a time-varying cycle of human flow, energy consumption data of each type of equipment is analyzed according to the same standard, ignoring the differences in energy consumption standards between energy-consuming equipment and different areas within the building, which can easily lead to misjudgment or omission of equipment energy consumption, and poor pertinence and accuracy of energy-saving management and control.

[0006] One or more embodiments of this specification adopt the following technical solutions:

[0007] One or more embodiments of the present specification provide a personalized energy-saving management and control method based on a GRU model, the method comprising: collecting real-time energy consumption data of multiple energy-consuming devices in a smart building, and determining a monitoring sub-area corresponding to each of the real-time energy consumption data, wherein the real-time energy consumption data includes real-time energy consumption, a collection time identifier, and a data source identifier; obtaining real-time crowd flow data of the monitoring sub-area, and determining an equipment energy consumption monitoring strategy corresponding to the monitoring sub-area based on the real-time crowd flow data of the monitoring sub-area and pre-acquired regional attribute data corresponding to the monitoring sub-area, wherein the equipment energy consumption monitoring strategy includes a minimum energy consumption operation strategy, a comfort guarantee operation strategy, and an emergency high energy consumption operation strategy; determining the corresponding GRU regional energy consumption prediction model through the equipment energy consumption monitoring strategy. model to determine the current energy-saving reference data corresponding to each of the energy-consuming devices, and monitor the multiple energy-consuming devices based on the current energy-saving reference data and the real-time energy consumption; when at least one energy-consuming abnormal device is monitored, the current abnormal reference information of each energy-consuming abnormal device is generated, and based on the pre-trained GRU device energy consumption prediction model, the energy consumption trend of each energy-consuming abnormal device is predicted to determine the predicted energy consumption information of each energy-consuming abnormal device, wherein the abnormal reference information includes abnormal type information and abnormal device information; according to the predicted energy consumption information corresponding to each energy-consuming abnormal device and the current abnormal reference information, the energy-saving control strategy corresponding to the at least one energy-consuming abnormal device is generated to realize energy-saving management and control of the at least one energy-consuming abnormal device.

[0008] One or more embodiments of this specification provide a personalized energy-saving control device based on a GRU model, including:

[0009] at least one processor; and,

[0010] a memory communicatively connected to the at least one processor; wherein,

[0011] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the above method.

[0012] One or more embodiments of the present specification provide a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured to execute the above method.

[0013] At least one of the above technical solutions adopted in the embodiments of this specification can achieve the following beneficial effects: through the above technical solution, the energy consumption data of multiple energy-consuming devices are collected in real time and accurately to each monitoring sub-area, so that energy consumption management is more refined and helps to accurately identify energy consumption hotspots and abnormal points; through the combination of real-time human flow data and regional attribute data, the system can more accurately judge the energy consumption needs of different sub-areas, so as to formulate a more reasonable equipment energy consumption monitoring strategy; using the GRU regional energy consumption prediction model, combined with the regional predicted energy consumption information, the current energy-saving reference data of each energy-consuming device is generated, taking into account the correlation between the current energy-saving reference data and the region, strategy, and human flow, to meet the personalized differences in energy consumption standards in different regions in the smart building application scenario where there is a time-varying cycle in the flow of human flow. In addition, the current Energy-saving reference data can meet the differences in energy consumption standards between energy-consuming equipment and different areas within the building, which helps to detect energy consumption anomalies in a timely manner and provide a scientific basis for the subsequent formulation of energy-saving control strategies; when energy-consuming abnormal equipment is monitored, the system will immediately generate abnormal reference information and predict energy consumption trends, making the formulation of energy-saving control strategies more targeted and timely, and able to respond quickly to energy consumption anomalies and reduce energy waste; energy-saving control strategies can comprehensively consider multiple energy-consuming abnormal equipment and optimize global energy consumption by adjusting priorities. At the same time, the strategy contains specific control parameters and plans for each energy-consuming abnormal equipment, ensuring targeted and personalized control of each device, realizing an organic combination of the global and local, and helping to more accurately control energy consumption, improve energy-saving efficiency, and thus achieve better energy-saving effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art description. Obviously, the drawings described below are only some embodiments recorded in this specification. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative labor. In the drawings:

[0015] Figure 1 A flowchart of a personalized energy-saving control method based on a GRU model provided in an embodiment of this specification;

[0016] Figure 2 A structural diagram of a personalized energy-saving control device based on a GRU model provided in an embodiment of this specification. DETAILED DESCRIPTION

[0017] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0018] The embodiment of this specification provides a personalized energy-saving control method based on the GRU model. It should be noted that the execution subject in the embodiment of this specification can be a server or any device with data processing capabilities. Figure 1 A flowchart of a personalized energy-saving control method based on a GRU model provided in an embodiment of this specification is shown in FIG. Figure 1 As shown, it mainly includes the following steps:

[0019] Step S101, collect real-time energy consumption data of multiple energy-consuming devices in the smart building, and determine the monitoring sub-area corresponding to each real-time energy consumption data.

[0020] In one embodiment of the present specification, real-time energy consumption data of multiple energy-consuming devices in a smart building are collected through the Internet of Things technology, wherein the real-time energy consumption data includes real-time energy consumption, collection time identifier and data source identifier. It should be noted that the type of energy consumption data to be collected is clearly defined, such as electricity, water, gas, etc. According to the collected energy consumption data type, a suitable sensor is selected and deployed in the energy-consuming device, and the energy consumption data is collected in real time through the sensor. In addition, there are multiple types of energy consumption data types in a smart building, and energy-saving control can be performed for each type of energy consumption data. This embodiment takes a single energy consumption data type, electric energy, as an example for explanation. Real-time energy consumption data of multiple energy-consuming devices in a smart building are collected, and the real-time energy consumption is the real-time energy consumption information of the energy-consuming device. The collection time identifier is used to indicate the time corresponding to the collected real-time energy consumption. The data source identifier is used to identify the data collection device. Since the data collection device is deployed in the energy-consuming device, the data collection device can also be associated with the energy-consuming device, that is, the data source identifier can also refer to the identifier of the energy-consuming device.

[0021] In the application scenarios of smart buildings with large scale and time-varying cycles of human flow, such as office buildings, smart factories, large shopping malls, etc., the buildings are large in scale and their human flow has the characteristics of time-varying cycles. The higher the flow of people, the higher the needs of users in the building need to be met, and the higher the corresponding energy consumption. For example, the more users there are in the building, the more water and electricity consumption will show an upward trend. In this case, if a general judgment standard is used, it is easy to judge that water or electricity exceeds the standard, which will trigger energy-saving control and produce misjudgment. The current regional distribution map of smart buildings is mostly a fixed regional distribution. There may be multiple flows of people in each area, resulting in a lack of targeted energy-saving control in a single area. If the distribution of pedestrian flow is combined with the regional distribution, the flow of pedestrians in each area is relatively uniform. When performing energy-saving control in the future, a unified regional judgment rule can be adopted for this area. In addition, there are more energy-consuming equipment in large-scale smart buildings. In the case of a large number of energy-consuming equipment, if judgment rules are set for each energy-consuming equipment to compare the energy consumption of a single device, a large amount of calculation will be generated, resulting in low monitoring efficiency. Therefore, on the basis of achieving targeted monitoring and identification, the work of comparing the energy consumption of a single device is avoided, reducing the workload.

[0022] Determine the monitoring sub-area corresponding to each of the real-time energy consumption data, specifically including: obtaining an initial distribution map corresponding to the smart building, and obtaining a plurality of historical crowd flow distribution data of each preset location point corresponding to the smart building, wherein the historical crowd flow distribution data includes historical crowd flow data and historical crowd flow time data, and the historical crowd flow time data includes a date dimension and a time dimension; according to the plurality of historical crowd flow distribution data of each of the preset location points, analyze the time evolution process of the crowd flow at each of the preset location points to generate crowd flow evolution information corresponding to each of the preset location points; determine the location area corresponding to each of the preset location points in the initial distribution map, obtain the regional volume information of each of the location areas, and set a minimum reference crowd flow threshold and a maximum reference crowd flow threshold based on the regional volume information; and determine the location area corresponding to each of the preset location points according to the minimum reference crowd flow threshold and the maximum reference crowd flow threshold. The pedestrian flow evolution information corresponding to the location point is grouped according to time, and multiple groups of time reference intervals and the pedestrian flow level of the preset location point in each group of time reference intervals are determined, wherein each group of the time reference intervals includes multiple time periods, and the pedestrian flow level corresponding to each group of the time reference intervals is different; according to the pedestrian flow level of each preset location point in each group of time reference intervals, the multiple initial location areas in the initial distribution map are adjusted to determine the smart building area distribution map corresponding to each group of time reference intervals; according to the smart building area distribution map corresponding to each group of time reference intervals, a smart building area distribution map database is constructed; through the collection time identifier of the real-time energy consumption data, the real-time smart building area distribution map corresponding to the corresponding time interval is matched in the smart building area distribution map database; based on the data source identifier of the real-time energy consumption data and the real-time smart building area distribution map, the corresponding monitoring sub-area is determined.

[0023] In one embodiment of the present specification, the real-time energy consumption data is collected and the real-time smart building area distribution map corresponding to the time interval is matched in the pre-constructed smart building area distribution map database, so as to determine the corresponding monitoring sub-area based on the data source identifier of the real-time energy consumption data and the real-time smart building area distribution map. In order to implement the above scheme, a smart building area distribution map database corresponding to different time intervals should be constructed in advance according to the time variation characteristics of the historical flow of people. The following is an implementation method provided by the embodiment of the present specification.

[0024] Obtain the initial distribution map corresponding to the smart building. The initial distribution map here is a fixed area distribution map within the smart building. It can be divided into functional areas, such as computer room area, rest area, office area, visit area, etc. Since the functions inside the smart building are relatively fixed, the initial distribution map is a fixed area distribution map. The division of functional areas can be set according to the usage type and user needs of the smart building. Obtain multiple historical traffic flow distribution data for each preset location point corresponding to the smart building. The historical traffic flow distribution data includes historical traffic flow data and historical traffic flow time data. The historical traffic flow time data includes date dimension and time dimension. It should be noted that the preset location point can be set according to the location of the energy-consuming equipment. For example, when there are multiple energy-consuming equipment of the same type in an area, a preset location point can be set at each energy-consuming equipment. The traffic flow data collected at each preset location point represents the operating environment of each energy-consuming equipment.

[0025] The multiple historical crowd flow distribution data of each preset location point contain two characteristics: crowd flow and time. It can be found that there is a certain pattern in the change of crowd flow over time. The time evolution process of the crowd flow at each preset location point is analyzed to generate the crowd flow evolution information corresponding to each preset location point. The crowd flow evolution information here is used to represent the change process of the crowd flow over time. When performing evolution analysis, a fixed time period can be specified for analysis. For example, the daily crowd flow change pattern can be analyzed. For example, the hourly, weekly or monthly crowd flow change pattern can be analyzed. The analysis period can be set according to the building use requirements of the smart building body. For example, there is a periodic change pattern between the crowd flow on weekdays and the crowd flow on weekends in a shopping mall, so a daily time period can be set; there is a periodic change pattern between the crowd flow during working hours and rest time in a hospital, so an hourly time period can be set.

[0026] Determine the location area corresponding to each preset location point in the initial distribution map, obtain the regional volume information of each location area, and set the minimum reference flow threshold and the maximum reference flow threshold based on the regional volume information. The minimum reference flow threshold and the maximum reference flow threshold here can be set according to demand to represent the appropriate flow range in this location area, and are not specifically limited here. For example, the minimum reference flow threshold can be set to 40% of the number of people that can be accommodated and the maximum reference flow threshold can be set to 60% of the number of people that can be accommodated. Through the minimum reference pedestrian flow threshold and the maximum reference pedestrian flow threshold, the pedestrian flow evolution information corresponding to each preset location point is grouped according to time, and the time interval when the pedestrian flow is lower than the minimum reference pedestrian flow threshold is grouped into one group, the time interval when the pedestrian flow is higher than the maximum reference pedestrian flow threshold is grouped into one group, and the time interval when the pedestrian flow is within the range of the minimum reference pedestrian flow threshold and the maximum reference pedestrian flow threshold is grouped into one group, so as to determine multiple groups of time reference intervals, and determine the pedestrian flow level of the preset location point in each group of time reference intervals. Since the time area of ​​each group may be a discontinuous time period, each group of time reference intervals includes multiple time periods, and the pedestrian flow level between each group of time reference intervals is different. For example, the pedestrian flow level of the time reference interval below the minimum reference pedestrian flow threshold is set to the first level flow, the pedestrian flow level of the time interval within the range of the minimum reference pedestrian flow threshold and the maximum reference pedestrian flow threshold is set to the second level flow, and the pedestrian flow level of the time reference interval above the maximum reference pedestrian flow threshold is set to the third level flow. According to the flow level of each preset location point in each group of time reference intervals, multiple initial location areas in the initial distribution map are adjusted to determine the smart building area distribution map corresponding to each group of time reference intervals, and based on the smart building area distribution map corresponding to each group of time reference intervals, a smart building area distribution map database is constructed.

[0027] In one embodiment of the present specification, the initial area position information of multiple initial position areas in the initial distribution map is obtained, and the preset position point is matched with the multiple initial position areas according to the initial area position information and the current position information corresponding to each preset position point, and the corresponding relationship is determined, so as to determine the designated initial position area corresponding to each preset position point based on the corresponding relationship. In the current time reference interval, the flow level of each preset position point in the current time reference interval is counted, so as to determine the regional distribution information of the flow level of the current time reference interval according to the designated initial position area corresponding to each preset position point; it should be noted that the regional distribution information of the flow level of the people includes the position distribution of multiple initial position areas and at least one flow level information corresponding to each initial position area. According to the regional distribution information of the flow level of people, at least one initial position area that meets the preset requirements is processed to determine the current smart building regional distribution map corresponding to the current time reference interval, wherein the processing includes segmentation processing and merging processing. The initial position area that meets the preset requirements here includes the position requirements and the flow level requirements of people. In the obtained current smart building regional distribution map, the flow levels of the two connected monitoring areas are different, and each monitoring area corresponds to a flow level of people.

[0028] For example, assuming that the current time reference interval is the first time period from 8:00 to 12:00 and the second time period from 14:00 to 18:00, the initial continuously distributed office area corresponds to three preset position points, the first preset position point has a level of traffic flow of level 2 in the current time reference interval, the second preset position point has a level of traffic flow of level 3 in the current time reference interval, and the third preset position point has a level of traffic flow of level 3 in the current time reference interval. The traffic flow level at each preset position point in the initial distribution map is marked. In this continuous office area, there are multiple levels of traffic flow distribution, so the office area can be divided, and the area of ​​the first preset position point and the area of ​​other preset position points can be divided. Similarly, if the traffic flow levels of two adjacent initial position areas are the same, for example, the traffic flow levels of the first initial office area and the second initial office area are both level 2, and the two areas are adjacent areas, the first initial office area and the second initial office area can be merged to obtain the smart building area distribution map corresponding to the first time period from 8:00 to 12:00 and the second time period from 14:00 to 18:00.

[0029] Through the above technical solution, the regional distribution map within the smart building can be dynamically adjusted through historical pedestrian flow distribution data to make it more in line with actual usage, thereby improving the accuracy and practicality of regional division; based on real-time energy consumption data and dynamically adjusted regional distribution maps, the energy consumption of different areas can be more accurately monitored and managed, which helps to identify abnormal energy consumption areas and take targeted measures to reduce energy consumption and improve energy utilization efficiency; by analyzing the changes in pedestrian flow over time and adjusting the regional distribution accordingly, better pedestrian flow management and space planning can be carried out, and resource allocation can be optimized; by analyzing the characteristics of pedestrian flow in different time periods, customized services and management strategies can be provided for different areas within the smart building to meet the needs of different users and usage scenarios.

[0030] In one embodiment of the present specification, the real-time energy consumption data is collected and the real-time smart building area distribution map corresponding to the time interval is matched in the pre-built smart building area distribution map database, so as to determine the corresponding monitoring sub-area based on the data source identifier of the real-time energy consumption data and the real-time smart building area distribution map. The real-time matching of the smart building area distribution map corresponding to the time interval can ensure that the division of the monitoring sub-area is consistent with the actual situation of the current time period, thereby improving the accuracy of energy consumption monitoring; based on the rapid matching of real-time energy consumption data and time identifiers, the system can quickly locate the sub-area that needs to be monitored, and respond to energy consumption changes in real time, which helps to timely discover and handle energy consumption anomalies; by real-time monitoring of the energy consumption of different sub-areas, energy consumption management strategies can be formulated and adjusted more finely, and personalized energy-saving measures can be taken according to the characteristics and needs of different sub-areas; the combination of real-time energy consumption data and smart building area distribution maps makes a large amount of energy consumption data more effectively utilized, providing strong support for energy consumption analysis and optimization.

[0031] Step S102, acquiring real-time crowd flow data of the monitoring sub-area, and determining the equipment energy consumption monitoring strategy corresponding to the monitoring sub-area according to the real-time crowd flow data of the monitoring sub-area and pre-acquired regional attribute data corresponding to the monitoring sub-area.

[0032] Among them, the equipment energy consumption monitoring strategy includes minimum energy consumption operation strategy, comfort guarantee operation strategy and emergency high energy consumption operation strategy.

[0033] According to the real-time crowd flow data of the monitoring sub-area and the pre-acquired regional attribute data corresponding to the monitoring sub-area, the equipment energy consumption monitoring strategy corresponding to the monitoring sub-area is determined, specifically including: obtaining the regional attribute information corresponding to the monitoring sub-area in the real-time smart building regional distribution map, wherein the regional attribute information includes regional function attributes and regional crowd flow level attributes; determining the corresponding crowd flow threshold data through the regional crowd flow level attributes of the monitoring sub-area, so as to judge the current crowd flow state of the monitoring sub-area according to the real-time crowd flow data and the crowd flow threshold data, wherein the current crowd flow state includes a standard state and a non-standard state; determining the theoretical energy consumption monitoring strategy corresponding to the monitoring sub-area according to the regional function attributes of the monitoring sub-area; when the current crowd flow state of the monitoring sub-area is a non-standard state, determining the non-standard type according to the real-time crowd flow data and the crowd flow threshold data, wherein the non-standard type includes any one of an over-standard sub-state and a low-standard sub-state; according to the non-standard type, modifying the theoretical energy consumption monitoring strategy, and determining the equipment energy consumption monitoring strategy corresponding to the monitoring sub-area.

[0034] In one embodiment of the present specification, real-time crowd flow data of the monitoring sub-area is obtained through a pre-set collection device. The regional attribute information corresponding to the monitoring sub-area in the real-time smart building area distribution map is determined. The regional attribute information includes regional functional attributes and regional crowd flow level attributes. The regional functional attributes here are used to represent the functions of this area, such as office, rest and other functions. The regional crowd flow level attributes include level one, level two and level three. By monitoring the regional crowd flow level attributes of the sub-area, the corresponding crowd flow threshold data can be determined. The minimum crowd flow reference threshold and the maximum crowd flow reference threshold can be set with reference to the minimum crowd flow reference threshold. The minimum crowd flow reference threshold and the maximum crowd flow reference threshold are set to the crowd flow threshold range corresponding to the second level. The minimum crowd flow reference threshold is the maximum crowd flow threshold corresponding to the first level, and the maximum crowd flow reference threshold is the minimum crowd flow threshold corresponding to the third level. Determine the pedestrian flow threshold corresponding to the monitoring sub-area, judge the current pedestrian flow state of the monitoring sub-area, and judge whether the real-time pedestrian flow data meets the pedestrian flow threshold. If it meets, it is determined to be a standard state; if it does not meet, it is determined to be a non-standard state, and according to the specific situation, analyze the real-time pedestrian flow data and the pedestrian flow threshold data to determine the non-standard type. The non-standard type includes any one of the over-standard sub-state and the low-standard sub-state. For example, if the monitoring sub-area is level one and the real-time pedestrian flow data is higher than the highest pedestrian flow threshold, the current pedestrian flow state is a non-standard state and an over-standard sub-state. In general, a theoretical energy consumption monitoring strategy can be set according to the functional type of each area. For example, the theoretical energy consumption monitoring strategy corresponding to the area with office functions is a comfort guarantee operation strategy.

[0035] When the current flow state of the monitoring sub-area is the standard state, the corresponding theoretical energy consumption monitoring strategy is determined as the equipment energy consumption monitoring strategy corresponding to the monitoring sub-area. When the current flow state of the monitoring sub-area is the non-standard state, the energy consumption strategy is downgraded or upgraded based on the theoretical energy consumption monitoring strategy according to the non-standard type, and the equipment energy consumption monitoring strategy corresponding to the monitoring sub-area is obtained. In general, when the non-standard type is the super-standard sub-state, the energy consumption strategy is upgraded on the basis of the theoretical energy consumption monitoring strategy; when the non-standard type is the low-standard sub-state, the energy consumption strategy is downgraded on the basis of the theoretical energy consumption monitoring strategy. For example, when the current flow state of the monitoring sub-area is the low-standard sub-state of the secondary flow level, and the current monitoring sub-area is the comfort guarantee operation strategy, it is switched to the minimum energy consumption operation strategy. It should be noted that the levels of the minimum energy consumption operation strategy, the comfort guarantee operation strategy and the emergency high energy consumption operation strategy increase in sequence according to the required energy consumption.

[0036] Through the above technical solution, by real-time monitoring of pedestrian flow data, the actual usage and demand of each area can be judged more accurately, so as to determine the energy consumption strategy in each monitoring sub-area with reference to the real-time pedestrian flow, avoiding the situation where there is a large difference between the pedestrian flow and the historical change pattern, and providing a method for correcting the operation strategy to suit the actual situation in each monitoring sub-area, thereby ensuring the matching between the energy consumption monitoring strategy and each monitoring sub-area.

[0037] Step S103, through the equipment energy consumption monitoring strategy, determine the corresponding GRU area energy consumption prediction model to determine the current energy-saving reference data corresponding to each energy-consuming device, and monitor multiple energy-consuming devices based on the current energy-saving reference data and real-time energy consumption.

[0038] Through the equipment energy consumption monitoring strategy, the corresponding GRU regional energy consumption prediction model is determined to determine the current energy-saving reference data corresponding to each of the energy-consuming devices, specifically including: through the equipment energy consumption monitoring strategy, a plurality of historical energy consumption data and a plurality of historical traffic data corresponding to the monitoring sub-area under the equipment energy consumption monitoring strategy are obtained; the plurality of historical energy consumption data and the plurality of historical traffic data are integrated to construct a model training data set, and the pre-constructed initial GRU model is trained by using the training data set to determine the GRU regional energy consumption prediction model corresponding to the monitoring sub-area under the equipment energy consumption monitoring strategy; based on the GRU regional energy consumption prediction model, the regional energy consumption under the equipment energy consumption monitoring strategy is predicted to determine the predicted energy consumption data time series corresponding to the monitoring sub-area; through the predicted energy consumption data time series corresponding to the monitoring sub-area, the average value of the predicted energy consumption data is calculated to determine the regional energy-saving reference data within the monitoring sub-area; according to the regional energy-saving reference data within the monitoring sub-area, the current energy-saving reference data corresponding to each of the energy-consuming devices within the monitoring sub-area is determined.

[0039] In one embodiment of the present specification, in the process of executing the equipment energy consumption monitoring strategy, multiple historical energy consumption data and multiple historical traffic data are collected, and the data should cover different time periods (such as days, weeks, months, and years) to ensure the generalization ability of the model. The collected historical energy consumption data and historical traffic data are cleaned to remove outliers and missing values. The data is normalized for model training. The pre-processed multiple historical energy consumption data and multiple historical traffic data are fused to construct a model training data set. Data fusion needs to consider time synchronization and correlation analysis to ensure that the model can accurately capture the relationship between energy consumption and traffic. An initial GRU model is pre-built. GRU (Gated Recurrent Unit) is a variant of a recurrent neural network (RNN) suitable for processing sequence data, such as time series prediction. The initial GRU model is trained using the constructed training data set. During the training process, the model performance is optimized by adjusting the model parameters (such as learning rate, batch size, number of iterations, etc.), and a loss function (such as mean square error MSE) is usually used to quantify the difference between the model predicted value and the actual value. After training, the GRU regional energy consumption prediction model corresponding to the monitoring sub-area under the energy consumption monitoring strategy of the equipment is determined. The model can predict the energy consumption in the current monitoring sub-area within a certain period of time in the future based on historical data. Based on the GRU regional energy consumption prediction model, the regional energy consumption under the energy consumption monitoring strategy of the equipment is predicted, and the time series of predicted energy consumption data corresponding to the monitoring sub-area is determined. The average value of the predicted energy consumption data is calculated through the time series of predicted energy consumption data corresponding to the monitoring sub-area. This average value can be used as the regional energy-saving reference data in the monitoring sub-area, or it can be based on this average value. The reference range is set to obtain the regional energy-saving reference data in the monitoring sub-area. According to the regional energy-saving reference data in the monitoring sub-area, the current energy-saving reference data corresponding to each energy-consuming device located in the monitoring sub-area can also be determined in combination with the specific situation of each energy-consuming device (such as device type, power, operating time, etc.). The current energy-saving reference data obtained here is the normal energy consumption level of the energy-consuming equipment in this monitoring sub-area during the historical execution of the equipment energy consumption monitoring strategy. It can be a specific baseline value or a reference range.

[0040] In one embodiment of the present specification, multiple energy consuming devices are monitored based on current energy-saving reference data and real-time energy consumption to determine whether the real-time energy consumption conforms to the current energy-saving reference data. If it does not conform to the current energy-saving reference data, it is determined to be energy consumption abnormality; for example, energy consumption that is higher than a baseline threshold or not within a baseline range is determined to be energy consumption abnormality.

[0041] Through the above technical solution, by integrating historical energy consumption and passenger flow data and training the GRU model, the model takes into account the impact of the important factor of passenger flow on energy consumption, and can more accurately predict the energy consumption of the monitored sub-area; the current energy-saving reference data of each energy-consuming device is determined, so that energy consumption management can be refined to the level of a single device, thereby achieving more refined management; the current energy-saving reference data of each energy-consuming device provides a scientific basis for the formulation and implementation of energy-saving measures, which helps to ensure the effectiveness and pertinence of energy-saving measures; by predicting energy consumption and setting energy-saving reference data, equipment with abnormal or excessive energy consumption can be discovered in a timely manner, and corresponding energy-saving measures can be taken, thereby promoting energy conservation and emission reduction.

[0042] Step S104, when at least one abnormal energy consumption device is monitored, current abnormal reference information of each abnormal energy consumption device is generated, and based on the pre-trained GRU device energy consumption prediction model, the energy consumption trend of each abnormal energy consumption device is predicted to determine the predicted energy consumption information of each abnormal energy consumption device.

[0043] The abnormal reference information includes abnormal type information and abnormal device information.

[0044] In one embodiment of the present specification, when at least one abnormal energy consumption device is monitored, for each abnormal device, its device type information, such as air conditioning, lighting, etc., and the abnormal monitoring sub-area to which the device belongs are collected. The historical device energy consumption data of each energy-consuming device before the current data collection time is collected, including timestamps, energy consumption values, etc., and the collected data is cleaned to remove outliers, missing values, etc. The data is normalized or standardized so that the model can learn better. The processed data is constructed as a data set required for model training, including input features and labels. For time series prediction, continuous time periods are usually used as input features, and the energy consumption value at the next time point is used as a label. Use the GRU layer as the core of the model to build a neural network structure, and other layers, such as fully connected layers, Dropout layers, etc., can also be added according to actual needs. Initialize the weights, biases and other parameters of the GRU layer, and set hyperparameters such as optimizers and learning rates.

[0045] Divide the data set into training set, validation set and test set. Use the training set to train the model and update the model parameters through forward propagation and back propagation algorithms. After each training cycle (epoch), use the validation set to evaluate the performance of the model. According to the performance of the model on the validation set, adjust the hyperparameters, such as learning rate, number of hidden layer units, etc. Hyperparameter tuning can be performed using grid search, random search and other methods. Save the trained model according to the corresponding device ID so that the corresponding GRU device energy consumption prediction model can be obtained according to the device ID of the energy consumption abnormal device in the future. Through the GRU device energy consumption prediction model, the energy consumption trend of each energy consumption abnormal device is predicted to determine the predicted energy consumption information of each energy consumption abnormal device.

[0046] Through the above technical solution, through real-time monitoring, once an abnormal energy consumption device is found, abnormal reference information can be quickly generated, including device identification, abnormal type, abnormal degree, etc., to facilitate rapid response and problem location; using the pre-trained GRU device energy consumption prediction model, the energy consumption trend of each abnormal energy consumption device can be accurately predicted, which is helpful to analyze the cause of the abnormality, such as equipment aging, improper operation strategy, environmental factors, etc.; based on the predicted energy consumption information, targeted energy-saving measures or maintenance plans can be formulated. For example, for equipment with predicted high energy consumption, it may be necessary to repair or replace it; for equipment with abnormal energy consumption due to improper operation strategy, its operating parameters can be adjusted; by predicting energy consumption information, resource allocation can be optimized. For example, when it is predicted that the energy consumption of a certain device will increase significantly, resources can be arranged in advance for maintenance or upgrade to avoid production interruption or energy waste caused by equipment failure.

[0047] Step S105, generating an energy-saving control strategy corresponding to at least one energy-consuming abnormal device according to the predicted energy consumption information and current abnormal reference information corresponding to each energy-consuming abnormal device, so as to implement energy-saving control of the at least one energy-consuming abnormal device.

[0048] In one embodiment of the present specification, the abnormal type information and abnormal device information in the current abnormal reference information corresponding to each abnormal energy consumption device are obtained, and the abnormal device information includes the device type information and the abnormal monitoring sub-area to which the abnormal energy consumption device belongs. According to the predicted energy consumption information and the current energy-saving reference data corresponding to each abnormal energy consumption device, the predicted energy consumption exceeding standard data of each abnormal energy consumption device is determined. The predicted energy consumption information corresponding to each abnormal energy consumption device is compared with the current energy-saving reference data corresponding to this device, and the predicted energy consumption exceeding standard data in the future time is judged based on the predicted energy consumption, that is, the predicted exceeding standard situation. The energy-saving control priority of each abnormal energy consumption device is set according to the abnormal device information and the predicted energy consumption exceeding standard data of each abnormal energy consumption device.

[0049] According to the current energy consumption exceeding standard data and the predicted energy consumption exceeding standard data in the current abnormal reference information, the energy-saving control target energy consumption data of each energy consumption abnormal device is determined. It should be noted that the current energy consumption exceeding standard data here refers to the energy consumption that has exceeded the standard in the current state, and the predicted energy consumption exceeding standard data refers to the energy consumption predicted in the future. When performing energy-saving control, the control system needs response time, and the preset energy consumption exceeding standard amount corresponding to the response time can be considered when making predictions. If the current energy consumption exceeding standard data is simply used as the control target data, the energy consumption of the equipment changes in real time during the response time period, that is, the adjusted initial energy consumption changes, which will cause the energy consumption after energy-saving control to still not meet the benchmark, resulting in repeated control or invalid control. Therefore, when setting the energy-saving control target energy consumption data, the current energy consumption exceeding standard data and the predicted energy consumption exceeding standard data are comprehensively considered. If the difference between the current energy consumption exceeding standard data and the predicted energy consumption exceeding standard data exceeds the preset allowable threshold, the average value of the current energy consumption exceeding standard data and the predicted energy consumption exceeding standard data is taken as the energy-saving control target energy consumption data. Based on the energy consumption data of the energy-saving control target and the abnormal type information, the energy-saving control parameter information of each energy consumption abnormal device is determined. The energy-saving control parameter information here can be a control scheme, which can be determined in the form of an expert database or an experience database. Based on the energy-saving control priority of each energy consumption abnormal device and the energy-saving control parameter information of each energy consumption abnormal device, the energy-saving control strategy corresponding to the at least one energy consumption abnormal device is determined. The energy-saving control strategy here is a global energy-saving control strategy within the smart building, which includes the adjustment priority of multiple energy consumption abnormal devices to meet the global control needs within the smart building, and also includes the specific control parameters and schemes corresponding to each energy consumption abnormal device to meet the targeted and personalized control needs of each energy consumption abnormal device.

[0050] As a global strategy within smart buildings, the energy-saving control strategy can comprehensively consider multiple energy-consuming devices and optimize global energy consumption by adjusting priorities. At the same time, the strategy includes specific control parameters and plans for each energy-consuming device, ensuring targeted and personalized control of each device, and realizing an organic combination of the global and local. By predicting the analysis of energy consumption information and current abnormal reference information, it can more accurately identify energy-consuming devices and formulate corresponding control strategies, which helps to more accurately control energy consumption and improve energy-saving efficiency, thereby achieving better energy-saving effects.

[0051] The energy-saving control priority of each energy-consuming abnormal device is set through the abnormal device information of each energy-consuming abnormal device and the predicted energy consumption exceeding data, specifically including: when the predicted energy consumption exceeding data is not greater than the preset energy consumption exceeding threshold, determining the abnormal monitoring sub-area corresponding to each energy-consuming abnormal device and the device type information corresponding to each energy-consuming abnormal device to match the first device energy consumption monitoring strategy and the first area human flow level attribute corresponding to the abnormal monitoring sub-area; determining the monitoring sub-area priority of the abnormal monitoring sub-area corresponding to the energy-consuming abnormal device according to the device type information corresponding to each energy-consuming abnormal device, the first device energy consumption monitoring strategy and the first area human flow level attribute corresponding to the abnormal monitoring sub-area; generating the initial energy-saving control priority of the energy-consuming abnormal device through the abnormal monitoring sub-area corresponding to each energy-consuming abnormal device and the monitoring sub-area priority; adjusting the initial energy-saving control priority based on the predicted energy consumption exceeding data of each energy-consuming abnormal device, and determining the energy-saving control priority of the energy-consuming abnormal device.

[0052] In one embodiment of the present specification, when the predicted energy consumption exceeding standard data of each energy consumption abnormal device is not greater than the preset energy consumption exceeding standard threshold, it indicates that the energy consumption of each energy consumption abnormal device within the response time of the energy-saving control is an excessive energy consumption within the allowable range. In this case, the initial energy-saving control priority can be set according to the priority of the abnormal monitoring sub-area, and then the initial energy-saving control priority can be adjusted according to the predicted energy consumption exceeding standard data to determine the energy-saving control priority of the energy consumption abnormal device. When there is at least one energy consumption abnormal device whose predicted energy consumption exceeding standard data is greater than the preset energy consumption exceeding standard threshold, this energy consumption abnormal device is taken as the highest priority, and the energy-saving control priority of other energy consumption devices that are not greater than the preset energy consumption exceeding standard threshold is determined according to the situation that they are not greater than the preset energy consumption exceeding standard threshold.

[0053] When the predicted energy consumption exceeding standard data of each energy consumption abnormal device is not greater than the preset energy consumption exceeding standard threshold, the abnormal monitoring sub-area corresponding to each energy consumption abnormal device and the device type information corresponding to each energy consumption abnormal device are determined to match the first device energy consumption monitoring strategy corresponding to the abnormal monitoring sub-area and the first area human flow level attribute. According to the device type information corresponding to each energy consumption abnormal device, the first device energy consumption monitoring strategy corresponding to the abnormal monitoring sub-area and the first area human flow level attribute, the monitoring sub-area priority of the abnormal monitoring sub-area corresponding to the energy consumption abnormal device is determined.

[0054] Specifically, it includes: according to the pedestrian flow level attribute of the first area, setting the personnel flow simulation parameters in the abnormal monitoring sub-area, simulating the operation of multiple energy-consuming devices in the abnormal monitoring sub-area, and determining the number of designated devices that meet the preset conditions, wherein the number of designated devices is the minimum number of energy-consuming devices required to meet the energy consumption monitoring strategy of the first device. It should be noted that the personnel flow simulation parameter here is the median value of the pedestrian flow threshold range corresponding to the pedestrian flow level attribute of the first area, and the maximum reference pedestrian flow is used to simulate the operation of multiple energy-consuming devices belonging to this abnormal monitoring sub-area to determine the minimum number of energy-consuming devices required to meet the energy consumption monitoring strategy of the first device. The simulation method here can be set according to demand. For example, a fixed temperature value can be set through the first equipment energy consumption monitoring strategy, and the initial indoor environment parameters and pedestrian flow parameters are fixed. The air-conditioning equipment in the area is shut down one by one for simulation, and the minimum required equipment that can at least meet this fixed temperature value is determined.

[0055] According to the abnormal device type information corresponding to each of the energy consumption abnormal devices and the device type information of multiple energy consumption devices in the abnormal monitoring sub-area obtained in advance, determine whether there are at least the specified number of other energy consumption normal devices with the same abnormal device type information in each of the abnormal monitoring sub-areas; if there are at least the specified number of other energy consumption normal devices with the same abnormal device type information in the first abnormal monitoring sub-area, and there are not at least the specified number of other energy consumption normal devices with the same abnormal device type information in the second abnormal monitoring sub-area, the priority of the second abnormal monitoring sub-area is higher than that of the first abnormal monitoring sub-area; if there are at least the specified number of other energy consumption normal devices with the same abnormal device type information in multiple third abnormal monitoring sub-areas, or, there are not at least the specified number of other energy consumption normal devices with the same abnormal device type information in multiple third abnormal monitoring sub-areas, then according to the device backup number of other energy consumption normal devices with the same abnormal device type information, the monitoring sub-area priorities corresponding to the multiple third abnormal monitoring sub-areas are set, wherein the monitoring sub-area priority is negatively correlated with the device backup number.

[0056] The initial energy-saving control priority of the energy-abnormal device is generated by the abnormal monitoring sub-area corresponding to each energy-abnormal device and the monitoring sub-area priority corresponding to each monitoring sub-area. Based on the predicted energy consumption exceeding standard data of each energy-abnormal device, the initial energy-saving control priority is adjusted to determine the energy-saving control priority of the energy-abnormal device. In the initial energy-saving control priority, there are multiple energy-abnormal devices belonging to the same monitoring sub-area in parallel. In this case, according to the predicted energy consumption exceeding standard data of each energy-abnormal device, the priority of these multiple parallel energy-abnormal devices is set. The more predicted energy consumption exceeding standard data, the higher the energy-saving control priority.

[0057] Through the above technical solution, the energy consumption data of multiple energy-consuming devices are collected in real time and accurately to each monitoring sub-area, making energy consumption management more refined and helping to accurately identify energy consumption hotspots and abnormal points; through the combination of real-time pedestrian flow data and regional attribute data, the system can more accurately judge the energy consumption needs of different sub-areas, thereby formulating a more reasonable equipment energy consumption monitoring strategy; using the GRU regional energy consumption prediction model, combined with regional predicted energy consumption information, the current energy-saving reference data of each energy-consuming device is generated, taking into account the correlation between the current energy-saving reference data and the region, strategy, and pedestrian flow, to meet the personalized differences in energy consumption standards in different regions in the smart building application scenario where there is a time-varying cycle in the flow of pedestrians. In addition, it ensures that the current energy-saving reference data can meet the energy-consuming equipment and The difference in energy consumption standards between different areas within the building helps to detect energy consumption anomalies in a timely manner and provide a scientific basis for the subsequent formulation of energy-saving control strategies; when abnormal energy consumption equipment is monitored, the system will immediately generate abnormal reference information and predict energy consumption trends, making the formulation of energy-saving control strategies more targeted and timely, and able to respond quickly to energy consumption anomalies and reduce energy waste; energy-saving control strategies can comprehensively consider multiple energy consumption anomalies. By adjusting priorities, global energy consumption optimization is achieved. At the same time, the strategy contains specific control parameters and plans for each energy consumption anomaly device, ensuring targeted and personalized control of each device, realizing an organic combination of the global and local, and helping to more accurately control energy consumption, improve energy-saving efficiency, and thus achieve better energy-saving effects.

[0058] The embodiment of this specification also provides a personalized energy-saving control device based on the GRU model, such as Figure 2 As shown, the device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above method.

[0059] The embodiments of the present specification also provide a non-volatile computer storage medium storing computer executable instructions, wherein the computer executable instructions are configured to execute the above method.

[0060] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device, equipment, and non-volatile computer storage medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0061] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0062] The above description is only one or more embodiments of this specification and is not intended to limit this specification. For those skilled in the art, one or more embodiments of this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of one or more embodiments of this specification shall be included in the scope of the claims of this specification.

Claims

1. A personalized energy-saving control method based on a GRU model, characterized in that: The method comprises: Collecting real-time energy consumption data of multiple energy-consuming devices in a smart building, and determining a monitoring sub-area corresponding to each of the real-time energy consumption data, wherein the real-time energy consumption data includes real-time energy consumption, a collection time identifier, and a data source identifier; Acquire real-time crowd flow data of the monitoring sub-area, and determine the equipment energy consumption monitoring strategy corresponding to the monitoring sub-area according to the real-time crowd flow data of the monitoring sub-area and the pre-acquired regional attribute data corresponding to the monitoring sub-area, wherein the equipment energy consumption monitoring strategy includes a minimum energy consumption operation strategy, a comfort guarantee operation strategy and an emergency high energy consumption operation strategy; Determine the corresponding GRU area energy consumption prediction model through the equipment energy consumption monitoring strategy to determine the current energy-saving reference data corresponding to each of the energy-consuming equipment, and monitor the multiple energy-consuming equipment based on the current energy-saving reference data and the real-time energy consumption; When at least one abnormal energy consumption device is monitored, current abnormal reference information of each abnormal energy consumption device is generated, and based on the pre-trained GRU device energy consumption prediction model, the energy consumption trend of each abnormal energy consumption device is predicted to determine the predicted energy consumption information of each abnormal energy consumption device, wherein the abnormal reference information includes abnormal type information and abnormal device information; According to the predicted energy consumption information and the current abnormal reference information corresponding to each of the energy consumption abnormal devices, an energy-saving control strategy corresponding to the at least one energy consumption abnormal device is generated to achieve energy-saving management and control of the at least one energy consumption abnormal device.

2. According to claim 1, a personalized energy-saving control method based on a GRU model is characterized in that: Determining the monitoring sub-area corresponding to each of the real-time energy consumption data specifically includes: Obtain an initial distribution map corresponding to the smart building, and obtain a plurality of historical pedestrian flow distribution data of each preset location point corresponding to the smart building, wherein the historical pedestrian flow distribution data includes historical pedestrian flow data and historical pedestrian flow time data; Analyzing the time evolution of the flow of people at each of the preset locations according to a plurality of historical flow distribution data of people at each of the preset locations to generate flow evolution information corresponding to each of the preset locations; Determine a location area corresponding to each of the preset location points in the initial distribution map, obtain regional volume information of each of the location areas, and set a minimum reference human flow threshold and a maximum reference human flow threshold based on the regional volume information; By using the minimum reference human flow threshold and the maximum reference human flow threshold, the human flow evolution information corresponding to each preset location point is grouped according to time, and multiple groups of time reference intervals and the human flow level of the preset location point in each group of time reference intervals are determined, wherein each group of the time reference intervals includes multiple time periods, and each group of the time reference intervals corresponds to a different human flow level; According to the flow level of each preset location point in each time reference interval, multiple initial location areas in the initial distribution map are adjusted to determine the smart building area distribution map corresponding to each time reference interval; According to the smart building area distribution map corresponding to each group of time reference intervals, a smart building area distribution map database is constructed; Matching the real-time smart building area distribution map corresponding to the corresponding time interval in the smart building area distribution map database through the collection time mark of the real-time energy consumption data; Based on the data source identifier of the real-time energy consumption data and the real-time smart building area distribution map, the corresponding monitoring sub-area is determined.

3. According to claim 2, a personalized energy-saving control method based on a GRU model is characterized in that: According to the flow level of people at each preset location point in each time reference interval, multiple initial location areas in the initial distribution map are adjusted to determine the smart building area distribution map corresponding to each time reference interval, specifically including: Acquire initial area position information of a plurality of initial position areas in the initial distribution map; According to the initial area position information and the current position information corresponding to each of the preset position points, the preset position point is matched with the multiple initial position areas to determine a corresponding relationship, so as to determine the designated initial position area corresponding to each of the preset position points based on the corresponding relationship; In the current time reference interval, counting the level of pedestrian flow at each of the preset location points in the current time reference interval, so as to determine the regional distribution information of the level of pedestrian flow in the current time reference interval according to the designated initial location area corresponding to each of the preset location points; According to the regional distribution information of the pedestrian flow level, at least one initial position area that meets the preset requirements is processed to determine the current smart building area distribution map corresponding to the current time reference interval, wherein the processing includes segmentation processing and merging processing.

4. According to the personalized energy-saving control method based on the GRU model of claim 1, it is characterized in that: Determining the equipment energy consumption monitoring strategy corresponding to the monitoring sub-area according to the real-time crowd flow data of the monitoring sub-area and the pre-acquired regional attribute data corresponding to the monitoring sub-area, specifically includes: Acquire the regional attribute information corresponding to the monitoring sub-area in the real-time smart building regional distribution map, wherein the regional attribute information includes regional function attributes and regional human flow level attributes; Determine the corresponding human flow threshold data through the regional human flow level attribute of the monitoring sub-area, so as to judge the current human flow state of the monitoring sub-area according to the real-time human flow data and the human flow threshold data, wherein the current human flow state includes a standard state and a non-standard state; Determining a theoretical energy consumption monitoring strategy corresponding to the monitoring sub-area according to the regional functional attributes of the monitoring sub-area; When the current human flow state of the monitoring sub-area is a non-standard state, determining a non-standard type according to the real-time human flow data and the human flow threshold data, wherein the non-standard type includes any one of an over-standard sub-state and a low-standard sub-state; According to the non-standard type, the theoretical energy consumption monitoring strategy is modified to determine the equipment energy consumption monitoring strategy corresponding to the monitoring sub-area.

5. According to the personalized energy-saving control method based on the GRU model of claim 1, it is characterized in that: By using the equipment energy consumption monitoring strategy, the corresponding GRU area energy consumption prediction model is determined to determine the current energy-saving reference data corresponding to each of the energy-consuming equipment, specifically including: By using the equipment energy consumption monitoring strategy, a plurality of historical energy consumption data and a plurality of historical human flow data corresponding to the monitoring sub-area under the equipment energy consumption monitoring strategy are obtained; The multiple historical energy consumption data and the multiple historical traffic data are integrated to construct a model training data set, so as to use the training data set to train the pre-built initial GRU model and determine the GRU regional energy consumption prediction model corresponding to the monitoring sub-area under the equipment energy consumption monitoring strategy; Based on the GRU regional energy consumption prediction model, predict the regional energy consumption under the equipment energy consumption monitoring strategy, and determine the predicted energy consumption data time series corresponding to the monitoring sub-area; Calculate the average value of the predicted energy consumption data through the time series of the predicted energy consumption data corresponding to the monitoring sub-area to determine the regional energy-saving reference data within the monitoring sub-area; According to the regional energy-saving reference data in the monitoring sub-area, current energy-saving reference data corresponding to each of the energy-consuming devices located in the monitoring sub-area is determined.

6. According to the GRU model-based personalized energy-saving control method of claim 1, it is characterized in that: Generating an energy-saving control strategy corresponding to at least one energy consumption abnormality device according to the predicted energy consumption information corresponding to each energy consumption abnormality device and the current abnormality reference information, specifically comprising: Acquire the abnormal type information and abnormal device information in the current abnormal reference information corresponding to each of the abnormal energy consumption devices, wherein the abnormal device information includes the device type information and the abnormal monitoring sub-area to which the abnormal energy consumption device belongs; Determine the predicted energy consumption exceeding standard data of each of the energy consumption abnormal devices according to the predicted energy consumption information corresponding to each of the energy consumption abnormal devices and the current energy-saving reference data; Setting the energy-saving control priority of each of the energy-consuming abnormal devices according to the abnormal device information and the predicted energy consumption exceeding standard data of each of the energy-consuming abnormal devices; Determine the energy-saving control target energy consumption data of each of the energy consumption abnormal devices according to the current energy consumption exceeding data and the predicted energy consumption exceeding data in the current abnormal reference information, and determine the energy-saving control parameter information of each of the energy consumption abnormal devices based on the energy-saving control target energy consumption data and the abnormal type information; Based on the energy-saving control priority of each of the energy-consuming abnormal devices and the energy-saving regulation parameter information of each of the energy-consuming abnormal devices, an energy-saving control strategy corresponding to the at least one energy-consuming abnormal device is determined.

7. According to claim 6, a personalized energy-saving control method based on a GRU model is characterized in that: The energy-saving control priority of each of the energy-consuming abnormal devices is set according to the abnormal device information of each of the energy-consuming abnormal devices and the predicted energy consumption exceeding standard data, specifically including: When the predicted energy consumption exceeding standard data is not greater than the preset energy consumption exceeding standard threshold, the abnormal monitoring sub-area corresponding to each of the energy consumption abnormal devices and the device type information corresponding to each of the energy consumption abnormal devices are determined to match the first device energy consumption monitoring strategy corresponding to the abnormal monitoring sub-area and the first area human flow level attribute; Determine the monitoring sub-area priority of the abnormal monitoring sub-area corresponding to the abnormal energy consumption device according to the device type information corresponding to each of the abnormal energy consumption devices, the first device energy consumption monitoring strategy corresponding to the abnormal monitoring sub-area, and the first area human flow level attribute; Generate an initial energy-saving control priority of the energy consumption abnormal device through the abnormal monitoring sub-area corresponding to each of the energy consumption abnormal devices and the monitoring sub-area priority; Based on the predicted energy consumption exceeding standard data of each of the energy consumption abnormal devices, the initial energy-saving control priority is adjusted to determine the energy-saving control priority of the energy consumption abnormal devices.

8. According to claim 7, a personalized energy-saving control method based on a GRU model is characterized in that: Determining the monitoring sub-area priority of the abnormal monitoring sub-area corresponding to the abnormal energy consumption device according to the device type information corresponding to each of the abnormal energy consumption devices, the first device energy consumption monitoring strategy corresponding to the abnormal monitoring sub-area, and the first area human flow level attribute, specifically includes: According to the pedestrian flow level attribute of the first area, set the personnel flow simulation parameters in the abnormal monitoring sub-area, simulate the operation of multiple energy-consuming devices in the abnormal monitoring sub-area, and determine the number of designated devices that meet the preset conditions, wherein the number of designated devices is the minimum number of energy-consuming devices required to meet the energy consumption monitoring strategy of the first equipment; According to the abnormal device type information corresponding to each of the abnormal energy consumption devices and the device type information of multiple energy consumption devices in the abnormal monitoring sub-area obtained in advance, it is determined whether there are at least the specified number of other normal energy consumption devices with the same abnormal device type information in each of the abnormal monitoring sub-areas; If there are at least the specified number of other energy-normal devices with the same type information as the abnormal device in the first abnormal monitoring sub-area, and there are not at least the specified number of other energy-normal devices with the same type information as the abnormal device in the second abnormal monitoring sub-area, the priority of the second abnormal monitoring sub-area is higher than that of the first abnormal monitoring sub-area; If there are at least the specified number of other energy-normal devices with the same abnormal device type information in multiple third abnormal monitoring sub-areas, or there are not at least the specified number of other energy-normal devices with the same abnormal device type information in multiple third abnormal monitoring sub-areas, then the monitoring sub-area priorities corresponding to the multiple third abnormal monitoring sub-areas are set according to the number of backup devices of other energy-normal devices with the same abnormal device type information, wherein the monitoring sub-area priority is negatively correlated with the number of backup devices.

9. A personalized energy-saving control device based on a GRU model, characterized in that: The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can perform the method according to any one of claims 1 to 8.

10. A non-volatile computer storage medium storing computer executable instructions, characterized in that: The computer executable instructions are configured to execute the method according to any one of claims 1 to 8.

Citation Information

Cited By

  • Smart park energy consumption management method and platform based on multi-source data driving

    CN119904331A

  • Intelligent park energy efficiency management system and method based on IBMS

    CN120087805A

  • A smart park energy efficiency management system and method based on IBMS

    CN120087805B

  • Information acquisition control method and system for port energy consumption equipment

    CN120103768A

  • An information collection and control method and system for energy-consuming equipment in a port

    CN120103768B