A Big Data-Based Building Energy Consumption Management Method and System

By combining historical building data with current equipment data, a cause-response diagram is generated and the target is verified, which solves the problem of misjudgment of multiple causes in building energy consumption management and achieves accurate identification of abnormal causes and targeted rectification.

CN123117545APending Publication Date: 2026-11-17济南铭云工程咨询有限责任公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611351029.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-02
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing building energy management methods struggle to accurately distinguish multiple possible causes when the same continuous energy consumption anomaly occurs, resulting in non-targeted rectification tasks and susceptibility to changes in equipment operating conditions.

Method used

By acquiring historical building metering data and current equipment operation data, and combining them with the energy consumption baseline of similar reference equipment, a cause-response diagram is formed. Equipment status data is obtained using the target verification task, and response breakpoints are determined along the control response path and energy consumption response branch to generate targeted rectification tasks.

Benefits of technology

It improves the accuracy of identifying the causes of anomalies, reduces misjudgments caused by changes in operating conditions and response delays, ensures that rectification tasks correspond to actual abnormal processes, and reduces the occurrence of ineffective rectification.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123117545A_ABST
    Figure CN123117545A_ABST
Patent Text Reader

Abstract

This invention discloses a building energy consumption management method and system based on big data, including: acquiring historical metering data, current metering data, and equipment operation data of the building; identifying persistent energy consumption anomalies based on the historical metering data, current metering data, and equipment operation data, and determining multiple candidate causes corresponding to the persistent energy consumption anomalies; determining a target verification task from preset verification tasks, the target verification task being used to acquire equipment status data or energy consumption data that can distinguish at least two candidate causes; executing the target verification task, retaining matching candidate causes based on the judgment conditions corresponding to each candidate cause and the acquired equipment status data or energy consumption data; and generating rectification tasks for building equipment, control configuration, or metering configuration based on the candidate causes. This solves the problem of accurately determining the actual cause of an anomaly when multiple possible causes correspond to the same persistent energy consumption anomaly, achieving the technical effect of improving the accuracy of anomaly cause judgment and the targeted nature of rectification.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of building energy management technology, specifically to a building energy consumption management method and system based on big data. Background Technology

[0002] With the application of building sub-metering systems, building automation systems, and building energy management platforms, metering data such as electricity consumption and power consumption of various energy-consuming devices can be continuously collected during building operation, as well as operational data such as operating schedules, control commands, equipment status, and environmental conditions. By analyzing historical metering data and current operational data, abnormal situations such as continuous energy consumption during non-service periods, high equipment energy consumption, and inconsistencies between control status and actual energy consumption can be identified, and equipment maintenance or control parameter adjustments can be carried out accordingly.

[0003] Existing building energy management methods typically determine whether energy consumption is abnormal based on historical energy consumption baselines, preset thresholds, or equipment operating rules, and match possible causes of the anomaly based on its characteristics, thereby generating corresponding alarm information or maintenance tasks. However, the actual energy consumption of building equipment is affected by a variety of factors, including equipment operating modes, environmental conditions, control status, and service load, and the same energy consumption anomaly may be caused by different factors.

[0004] For example, if building equipment continues to consume energy during non-service hours, this could be due to several reasons. These reasons could include abnormal operating schedule configurations or forced operation causing the control system to keep the equipment running; changes in control commands preventing the actuators from responding; or even when the equipment is actually operating normally, abnormal energy consumption displays caused by metering point configurations or metering data. These different causes may exhibit similar abnormal characteristics at the energy consumption data level, making it difficult to accurately distinguish them based solely on current energy consumption deviations or the status of a single device.

[0005] In this situation, if the cause of the fault is directly determined based on the abnormal characteristics and a rectification task is generated, the rectification target may not be consistent with the actual abnormal link; while if multiple possible causes are checked one by one in turn, a lot of equipment status data needs to be obtained, and it is easily affected by changes in the current operating conditions of the equipment. Summary of the Invention

[0006] This application provides a building energy consumption management method and system based on big data, which solves the technical problem in the existing building energy consumption management process that it is difficult to accurately determine the actual cause of the anomaly when the same continuous energy consumption anomaly corresponds to multiple possible causes, and to generate targeted rectification tasks. This achieves the technical effect of improving the accuracy of anomaly cause judgment and the targeted nature of rectification.

[0007] In a first aspect, the present invention provides a building energy consumption management method based on big data, comprising: Acquire historical building metering data, current metering data, and equipment operation data; Based on historical building metering data, current metering data, and equipment operation data, identify continuous energy consumption anomalies and determine multiple candidate causes corresponding to the continuous energy consumption anomalies; A target verification task is determined from the preset verification tasks. The target verification task is used to obtain device status data or energy consumption data that can distinguish at least two candidate causes. Execute the target verification task, and retain the matching candidate causes based on the obtained device status data or energy consumption data and the judgment conditions corresponding to each candidate cause; Based on the candidate reasons, rectification tasks are generated for building equipment, control configurations, or metering configurations.

[0008] Furthermore, the identification of persistent energy consumption anomalies based on historical building metering data, current metering data, and equipment operation data includes: The building equipment to be analyzed is identified as the target equipment. Based on the equipment's operating mode, time type, and environmental conditions, comparable historical metering data of the target equipment is extracted from the building's historical metering data. The energy consumption baseline of the target equipment is then determined based on the comparable historical metering data. From building equipment with the same equipment type and operating mode as the target equipment, identify similar reference equipment that does not currently exhibit energy consumption anomalies, and determine a common energy consumption deviation based on the energy consumption deviation of the similar reference equipment relative to its own energy consumption baseline. The corrected energy consumption residual is determined based on the difference between the energy consumption deviation of the target device relative to its own energy consumption baseline and the common energy consumption deviation. When the corrected energy consumption residual continuously meets the abnormal conditions, the continuous energy consumption abnormality is determined.

[0009] Furthermore, the determination of multiple candidate causes corresponding to the continuous energy consumption anomaly includes: Based on the device identifier and unified time reference, the control configuration data, control commands, execution component status and service side status data in the device operation data, and the energy consumption data in the current metering data are associated to form a cause response diagram corresponding to the continuous energy consumption anomaly. The cause-response diagram includes a control response path consisting of the control configuration data, control commands, and execution component status in sequence, as well as an energy consumption response branch and a service response branch that are respectively pointed to by the execution component status from the energy consumption data and the service side status data. Each candidate cause is associated with a node in the cause-response graph or with the response propagation relationship between adjacent nodes, and the expected response direction and expected response time window of the corresponding node are configured for each candidate cause.

[0010] Furthermore, when the target device has a continuously adjustable control quantity, the target verification task includes: Save the original control values ​​of the target device; During the verification period that meets the equipment safety conditions and building service conditions, the original control quantity is adjusted to the first test control quantity, the first response data of each node in the cause-response diagram is obtained, and the target equipment is restored to the original control quantity. After the target device is restored to its original operating state, the original control quantity is adjusted to the second test control quantity, the second response data of each node in the cause response diagram is obtained, and the target device is restored to the original control quantity again. The first test control quantity and the second test control quantity are located on both sides of the original control quantity, and their absolute changes relative to the original control quantity are the same.

[0011] Furthermore, when performing the target verification task, multiple similar reference devices that match the target device in terms of equipment type, operating mode, environmental conditions, and service load are identified from the building equipment that has never performed the target verification task. During the response time period corresponding to the first test control quantity and the second test control quantity, the control commands, execution component status, energy consumption data and service side status data of the target device and each similar reference device are acquired synchronously. A common change is determined based on the data change of each similar reference device, and a differential response sequence corresponding to the target device is formed based on the difference between the data change of the target device and the common change.

[0012] Further, forming the differential response sequence corresponding to the target device includes: Extract historical anomaly handling data from the historical data on rectification results that have passed energy consumption verification, and extract historical confirmed cases that match the equipment type, operating mode and target verification task of the target equipment. Based on the execution time of the target verification task described in each historical confirmed case, as well as the response time of the control command, execution component status, energy consumption data, and service side status data, the historical response delay interval of each node in the cause response diagram is determined respectively. The response time window corresponding to the current target verification task is set according to the historical response delay interval of each node, and the differential response sequence is formed according to the response direction, response start time and recovery time of each node within the corresponding response time window.

[0013] Furthermore, the step of retaining matching candidate causes based on the acquired device status data or energy consumption data and the judgment conditions corresponding to each candidate cause includes: The actual response direction and actual response time of each node in the differential response sequence are compared with the expected response direction and expected response time window corresponding to each candidate cause, respectively. Along the control response path, energy consumption response branch and service response branch of the cause response diagram, determine the first response transmission relationship where the upstream node meets the expected response and the actual response direction of the downstream node deviates from the expected response direction or the actual response time exceeds the expected response time window, and determine the first response transmission relationship as the response breakpoint. Retain candidate reasons associated with the node corresponding to the response breakpoint or the response propagation relationship; When the data integrity of the node involved in the response breakpoint is lower than the preset integrity threshold, or the time synchronization deviation exceeds the preset synchronization threshold, the candidate cause associated with that node is marked as a cause to be verified, and a data verification task for that node is generated.

[0014] Furthermore, when the response breakpoint is located in the energy consumption response branch and the service response branch meets the corresponding expected response, a metering topology is established, including the upper-level metering point, the target metering point, and other metering points at the same metering level as the target metering point. Within the response time window corresponding to the target verification task, the energy consumption changes of the superior metering point, the target metering point, and other metering points are obtained, and the response metering point is determined according to the correspondence between the energy consumption change direction and response time of each metering point and the state of the target device execution component. The metering balance residual is determined by the difference between the energy consumption change at the higher-level metering point and the sum of the energy consumption changes at each lower-level metering point. When the response metering point is one of the other metering points, and the metering balance residual satisfies the metering balance condition, the candidate cause corresponding to the association configuration error between the target device and the metering point is retained. When the energy consumption changes of the target metering point and other metering points are both lower than the response judgment threshold, and the metering balance residual exceeds the metering balance threshold, candidate causes corresponding to missed metering data collection or metering equipment malfunction are retained.

[0015] Furthermore, when there are at least two candidate causes remaining, the candidate causes are combined in pairs to form undistinguished cause pairs; Based on the number of unidentified cause pairs that each preset verification task can distinguish, as well as the corresponding control variable change range, expected service-side state deviation, and expected recovery time, preset verification tasks that exceed the corresponding allowable conditions in terms of control variable change range, expected service-side state deviation, or expected recovery time are screened out, and the target verification task that can distinguish the most unidentified cause pairs is determined from the remaining preset verification tasks. When the retained candidate reasons converge to one and correspond to the response breakpoint, a rectification task is generated based on the candidate reason; After the rectification task is executed, the target verification task is executed again. Based on the rectified differential response sequence and metering data, it is determined whether the continuous energy consumption anomaly has been eliminated. When the rectified differential response sequence or metering data still meets the judgment conditions corresponding to the continuous energy consumption anomaly, the candidate cause screening results supported by data that meets the data integrity and time synchronization conditions are retained, the candidate causes lacking reliable data support are restored as causes to be verified, and the target verification task is redefined.

[0016] Secondly, a building energy consumption management system based on big data includes: The data acquisition module is used to acquire historical building metering data, current metering data, and equipment operation data. The candidate cause identification module is used to identify continuous energy consumption anomalies based on historical building metering data, current metering data, and equipment operation data, and to determine multiple candidate causes corresponding to the continuous energy consumption anomalies. The target verification task determination module is used to determine the target verification task from the preset verification tasks. The target verification task is used to obtain device status data or energy consumption data that can distinguish at least two candidate causes. The candidate cause verification module is used to perform the target verification task and retain the matching candidate causes based on the obtained device status data or energy consumption data and the judgment conditions corresponding to each candidate cause. The rectification task generation module is used to generate rectification tasks for building equipment, control configuration, or metering configuration based on the candidate reasons.

[0017] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. This invention combines historical building metering data, current metering data, and equipment operation data to identify continuous energy consumption anomalies. Furthermore, it can use the target equipment's own historical energy consumption baseline and the common energy consumption changes of similar reference equipment to correct the current energy consumption deviation. This can reduce the impact of changes in environmental conditions, service load, and other factors on the anomaly identification results, making the identified continuous energy consumption anomalies more reflective of the target equipment's own operating status.

[0018] 2. After identifying continuous energy consumption anomalies, this invention first determines multiple candidate causes, and then obtains equipment status data or energy consumption data that can distinguish the candidate causes through target verification tasks. Furthermore, it can apply recoverable test control quantities to the target equipment, form differential responses by combining similar reference equipment that has not performed verification tasks, and use historical confirmed cases to determine the response time windows of each response node. This allows for the determination of response breakpoints along the control response path, energy consumption response branch, and service response branch, enabling the anomaly cause to be further located from simple phenomenon matching to specific nodes or response transmission relationships, reducing misjudgments of causes caused by changes in operating conditions and response delays.

[0019] 3. This invention generates rectification tasks for building equipment, control configuration, or metering configuration based on the verification results of candidate causes. For anomalies located in the energy consumption response branch, it can further distinguish between metering association configuration errors and metering data anomalies by combining the actual response of the target equipment and the metering relationship between upper and lower level metering points and same level metering points. After rectification, it can again determine whether the continuous energy consumption anomaly has been eliminated based on the equipment response and metering data, thereby reducing the situation where the rectification object does not correspond to the actual abnormal link and the same invalid rectification is repeatedly executed. Attached Figure Description

[0020] Figure 1 This is a flowchart of a big data-based building energy consumption management method according to Embodiment 1 of this application; Figure 2 This is a block diagram for continuous energy consumption anomaly identification in Embodiment 1 of this application; Figure 3 This is a block diagram of candidate causes and cause responses for Embodiment 1 of this application; Figure 4 This is a closed-loop block diagram of the metrological topology verification and rectification in Embodiment 1 of this application; Figure 5 This is a system block diagram of Embodiment 2 of this application; Detailed Implementation The preset thresholds, preset time windows, permissible conditions, permissible ranges, and other preset parameters involved in this embodiment, unless otherwise specified, can be determined based on one or more of the following: the rated operating parameters of the target building equipment, the equipment manufacturer's technical data, the existing control limits and protection interlocking conditions of the building automation system, the measurement accuracy and sampling period of the metering equipment or sensor, building design and commissioning data, statistical results of normal historical operating data, and on-site trial operation calibration results. For thresholds compared with specific physical quantities, the same physical dimensions, statistical periods, and data processing methods as the corresponding data are used. For different equipment types, operating modes, or environmental conditions, corresponding parameters can be set separately. As historical confirmed cases accumulate, the corresponding parameters can be updated based on verified and valid historical operating data, but the updated parameters should meet the equipment safety operating conditions and building service conditions. In cases where historical samples are insufficient, initial parameters can be determined based on the equipment manufacturer's technical data, building design, or on-site commissioning results.

[0021] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0022] Example 1 Reference Figure 1 This embodiment provides a building energy consumption management method based on big data. The method can be deployed in a building energy management platform, a building automation system, or a data processing server that communicates with both, and is used to continuously identify, verify the causes of, and manage the rectification of energy consumption anomalies in building equipment such as air handling units, supply and exhaust fans, water pumps, and chillers that have metering data and operating status data.

[0023] In this embodiment, historical and current building metering data can be sourced from building sub-metering systems, electricity meters, heating and cooling meters, or energy management platforms; equipment operation data can be sourced from building automation systems, equipment controllers, frequency converters, and field sensors. Equipment operation data can include at least equipment identification, equipment operating mode, control configuration data, control commands, actuator status, environmental conditions, and service-side status data. Service-side status data characterizes the actual technical service status provided by building equipment, such as the air volume, static pressure, or temperature corresponding to air handling equipment, and the flow rate or pressure corresponding to water pumps.

[0024] For data from different sources, time alignment is performed according to a unified time base, and the data is converted to a unified statistical period. For instantaneous power data, the corresponding energy consumption can be obtained by integration or accumulation over the statistical period; for cumulative electrical energy data, the energy consumption for the statistical period can be obtained based on the difference in cumulative values ​​at adjacent statistical times. Metering data participating in the same energy consumption calculation use the same energy unit. Data with different dimensions, such as control commands, actuator status, temperature, pressure, and flow rate, are processed separately, and cross-dimensional addition and subtraction are not performed directly.

[0025] In one implementation, historical anomaly handling data is also acquired. This historical anomaly handling data includes at least the historical anomaly object, equipment type, operating mode, verification task, verification task execution time, response data of each response node, confirmed cause, rectification task, and energy consumption verification results after rectification. Only cases where the anomaly cause has been confirmed and the rectification effect has been verified are marked as historical confirmed cases in the historical anomaly handling data, for subsequent determination of the normal response characteristics of each response node.

[0026] S100. Obtain historical building metering data, current metering data, and equipment operation data.

[0027] Historical building metering data is stored according to equipment identification, time type, equipment operating mode, and corresponding environmental conditions. Current metering data consists of energy consumption data for each building device within the period to be analyzed. Equipment operating data and metering data are correlated using equipment identification and time using a unified time base, enabling the analysis of control commands, actuator status, energy consumption data, and service-side status data for the same device on the same timeline.

[0028] Reference Figure 2 S200. Identify continuous energy consumption anomalies based on historical building metering data, current metering data, and equipment operation data, and determine multiple candidate causes corresponding to the continuous energy consumption anomalies.

[0029] In this embodiment, the device to be analyzed can first be determined based on energy consumption deviation, operating period, or preset alarm status. Then, according to S211-S213, it can be determined whether the device constitutes a continuous energy consumption anomaly. For the confirmed continuous energy consumption anomaly, a cause-response diagram is formed and a candidate cause set is generated according to S221-S223.

[0030] S211. Identify the building equipment with abnormal energy consumption as the target equipment, extract comparable historical metering data of the target equipment from the building's historical metering data according to the equipment's operating mode, time type and environmental conditions, and determine the target equipment's own energy consumption baseline based on the comparable historical metering data.

[0031] Comparable historical metering data is preferentially extracted from historical periods that share the same operating mode, time type, and environmental conditions as the current equipment, and fall within the corresponding comparable interval. Time types can include weekdays, non-weekdays, service hours, and non-service hours; environmental conditions can be categorized based on outdoor temperature, outdoor humidity, or other existing environmental monitoring parameters related to the target equipment's energy consumption. For equipment with different seasonal operating strategies, comparable historical data can also be extracted under the same seasonal operating strategy.

[0032] Let the target device be device. At the current statistical moment The corresponding number of comparable historical samples is , No. The energy consumption of a comparable historical sample is Then, a self-consumption baseline can be expressed as: in, Indicates the target device At the statistical time The corresponding baseline of its own energy consumption; Indicates the first Energy consumption of the target equipment within a comparable historical period; This indicates the number of comparable historical samples used in the baseline calculation.

[0033] When historical samples contain records of downtime for maintenance, communication interruptions, or abnormal energy consumption, the corresponding samples can be removed from the comparable historical samples. The baseline energy consumption can also be determined using the median, the mean after removing abnormal samples, or historical statistical values ​​updated according to a time window, as long as it can characterize the normal energy consumption level of the target equipment under the current comparable operating conditions.

[0034] To reduce the differences in energy consumption caused by different rated capacities of different equipment, the relative energy consumption deviation of the target equipment relative to its own energy consumption baseline can be further calculated: in, This indicates the energy consumption of the target device within the current statistical period; This indicates the relative energy consumption deviation of the target device relative to its own energy consumption baseline; This indicates a pre-defined positive number, whose dimensions are the same as... Similarly, this is used to avoid numerical anomalies when the baseline value is close to zero. The value can be determined based on the minimum effective metering resolution of the corresponding metering equipment or the minimum baseline value that can be effectively used for relative deviation calculation under normal operating conditions. When the target equipment is in a normal shutdown state and its own energy consumption baseline is close to zero, the absolute deviation between the current energy consumption and the energy consumption corresponding to the normal shutdown state can be used for anomaly judgment to avoid distorted judgment results due to the denominator being too small in the relative deviation.

[0035] S212. From building equipment with the same equipment type and operating mode as the target equipment, identify similar reference equipment that does not currently exhibit energy consumption anomalies, and determine a common energy consumption deviation based on the energy consumption deviation of the similar reference equipment relative to its own energy consumption baseline.

[0036] Similar reference equipment is used to characterize the common environmental changes and common service load changes experienced by the building where the target equipment is located during the current time period. Reference equipment is preferentially selected based on its equipment type, operating mode, environmental conditions, and service load, which match the target equipment, and which exhibit stable operation and complete metering data during the current analysis period. Service load can be matched using data from the equipment's existing control system that reflects the load status, such as set air volume, flow rate requirements, or existing load indication status.

[0037] For any similar reference device Determine its own energy consumption baseline in the same way as the target device. and relative energy consumption deviation Let the set of similar reference devices be . including If there are several similar reference devices, the common energy consumption deviation can be expressed as: in, This represents the common energy consumption deviation among similar reference devices at the current moment. For situations with a large number of reference devices or significant fluctuations in the performance of individual reference devices, the median can be used, or outlier reference devices can be removed before determining the common energy consumption deviation.

[0038] S213. Determine the corrected energy consumption residual based on the difference between the energy consumption deviation of the target device relative to its own energy consumption baseline and the common energy consumption deviation. When the corrected energy consumption residual continuously meets the abnormal conditions, determine the continuous energy consumption abnormality.

[0039] The corrected energy consumption residual can be expressed as: in, This represents the corrected energy consumption residual of the target device itself after deducting common variations among similar devices.

[0040] In one implementation, when the following conditions are met: And the above state continues for a preset duration If the anomaly persists within a given timeframe, or occurs repeatedly in multiple analysis periods of the same type, it indicates that the target device exhibits a persistent energy consumption anomaly. To correct the abnormal threshold of energy consumption residual, it can be determined based on the statistical distribution of corrected energy consumption residual during the normal historical operation phase of the target equipment, or it can be configured according to the existing operating requirements of the building energy management system.

[0041] Among them, the abnormal threshold The same data processing method is used as for correcting energy consumption residuals; when the corrected energy consumption residual is a relative deviation, the abnormal threshold is... It is a dimensionless quantity. Continuous preset duration. It can be determined based on the sampling period of the current metering data, the transient duration during the normal start-up or load adjustment process of the target equipment, and the short-term fluctuation characteristics in the normal historical operating data, so as to distinguish between continuous energy consumption anomalies and transient energy consumption deviations caused by normal start-up, shutdown, and short-term load adjustment.

[0042] Through the above treatment, when changes in outdoor temperature cause multiple similar devices to simultaneously increase energy consumption, the commonly changing portion is handled by... Characterization, the additional energy consumption deviation of the target device compared to similar devices is caused by This characterization reduces the interference of changes in common operating conditions on anomaly identification.

[0043] S221. Based on the device identifier and unified time reference, associate the control configuration data, control commands, execution component status, energy consumption data and service side status data in the device operation data to form a cause response diagram corresponding to the continuous energy consumption anomaly.

[0044] The nodes in the cause-response diagram correspond to the actual technical processes of the target equipment. Control configuration data is used to characterize the equipment's operating schedule, setpoints, or control modes; control commands are used to characterize the control requirements output by the controller to the executed object; the status of the executed components is used to characterize the actual status of executed objects such as frequency converters, contactors, valves, fans, or water pumps; energy consumption data is used to characterize the electrical energy, heating and cooling energy generated by the equipment, or corresponding energy consumption changes; server-side status data is used to characterize the technical results such as air supply, water supply, temperature, pressure, or flow rate formed after the equipment operates.

[0045] Data from each node is associated with a device through device identifiers and sorted according to a unified timestamp. When different data sources have different sampling periods, resampling or time window aggregation can be performed according to a preset analysis period; when there is a fixed communication delay in the data source, the timestamp can be corrected based on the system's existing time synchronization information.

[0046] S222. The cause-response diagram includes a control response path consisting of the control configuration data, control commands, and execution component states in sequence, as well as an energy consumption response branch and a service response branch that are respectively directed by the execution component states to the energy consumption data and service-side status data.

[0047] Taking a variable frequency air handling unit as an example, the control configuration data can include the operating schedule and frequency setpoint; the control command can be the inverter frequency command; the status of the actuator can be the actual output frequency of the inverter and the fan operating status; the energy consumption data can be the fan power or the electrical energy of the statistical period; and the service-side status data can be the air volume or the air static pressure. Therefore, the cause-response diagram can reflect whether the control configuration generates correct control commands, whether the control commands are executed by the actuators, whether the actions of the actuators produce the expected energy consumption change, and whether the actions of the actuators produce the expected service-side change.

[0048] Reference Figure 3 S223. Associate each candidate cause with a node in the cause-response diagram or the response transmission relationship between adjacent nodes, and configure the expected response direction and expected response time window for each candidate cause.

[0049] A cause database is pre-established for different equipment types. Each cause entry includes at least a cause identifier, applicable equipment type, corresponding node or response propagation relationship, available verification task, and judgment conditions. Based on the target equipment type, the abnormal characteristics of continuous energy consumption anomalies, and current equipment operating data, the system extracts cause entries with corresponding occurrence conditions from the cause database to form the current candidate cause set.

[0050] For example, if an air handling unit continues to consume energy during non-service hours, candidate causes may include abnormal operating schedule configuration, failure to deactivate forced operation, abnormal control command output, failure of actuators to respond to shutdown commands, abnormal configuration of the association between the equipment and the metering point, and abnormal metering data acquisition. If the current data indicates that the equipment operating schedule is correct, the corresponding cause can be downgraded in the verification priority; if the existing data is insufficient to make a reliable judgment, the corresponding cause will remain in the candidate cause set.

[0051] Reasons for being a candidate and nodes in the cause-response diagram Its expected response can be expressed as: in, Indicate the reason for the candidate Node at the time of establishment The expected response direction to the target verification task, wherein the response direction can be expressed as increasing, decreasing or being within a preset stable range; Represents a node The expected response time window. The expected response direction and the expected response time window together constitute the judgment conditions corresponding to the candidate cause.

[0052] The expected response direction is determined based on the control logic of the target equipment, the action relationship of the execution components, and the relationship between the corresponding energy consumption or service-side physical quantities and the equipment state. For example, when the operating frequency of a wind turbine is increased, the actual operating frequency and corresponding power of the wind turbine should change in the corresponding direction under normal circumstances. For nodes from which response data can be obtained from historical confirmed cases, the expected response time window is determined primarily based on the historical response delay under the same equipment type, operating mode, and verification task. When the number of historical confirmed cases is insufficient, the initial response time window can be determined based on the equipment manufacturer's technical data, the existing control delay of the building automation system, on-site commissioning records, or verification results under normal operating conditions, and updated according to S3231-S3233 after the historical confirmed cases are accumulated.

[0053] S300. Determine a target verification task from the preset verification tasks, wherein the target verification task is used to obtain device status data or energy consumption data that can distinguish at least two candidate causes.

[0054] A pre-established verification task library is created, with each pre-defined verification task associated with at least the applicable equipment type, applicable candidate causes, controlled object or data verification object, response node to be acquired, expected response for each candidate cause, equipment safety execution conditions, building service execution conditions, and recovery actions. For data verification tasks that only need to read existing states, the corresponding node data can be directly obtained; for verification tasks that require actively changing the equipment's operating state, they are only executed during verification periods that meet the equipment safety conditions and building service conditions.

[0055] The ability to distinguish at least two candidate causes means that for the same verification task, there are at least two candidate causes with different expected response directions at one or more response nodes, or identifiable differences in the expected response time windows, so that the actual response after the verification task is executed can narrow down the range of candidate causes.

[0056] When the target device has continuously adjustable control quantities, recoverable bidirectional verification can be implemented in accordance with S311-S313.

[0057] S311. Save the original control quantity of the target device.

[0058] Save the original control values ​​of the target device before it performs the verification task. It also saves the baseline state of each response node before the verification task is executed. The original control quantity can be the inverter frequency setpoint, valve opening setpoint, water pump speed setpoint, or other continuously adjustable control parameters.

[0059] S312. During the verification period that meets the equipment safety conditions and building service conditions, the original control quantity is adjusted to the first test control quantity, the first response data of each node in the cause response diagram is obtained, and the target equipment is restored to the original control quantity.

[0060] The first test control variable can be expressed as: ,in, Indicates the first test control variable; This indicates the amount of change in test control. The size is determined based on the target device's allowable control range and the server-side status's allowable change range, so that the verification action can elicit a identifiable response, while avoiding exceeding the allowable range for server-side status such as temperature, pressure, and flow.

[0061] The value of the test control variation Δu is limited by both the identifiable response conditions and the equipment operating constraints. Specifically, the lower limit of Δu can be determined based on the fluctuation range of the response data under normal operating conditions of the target equipment, the control resolution of the actuators, and the measurement resolution of the metering equipment or sensors, ensuring that the expected response generated by the test control can be distinguished from normal data fluctuations. The upper limit of Δu can be determined based on the allowable control range specified by the equipment manufacturer, the protection or interlocking boundaries set by the building automation system, and the allowable variation range of the service-side status. When the above conditions are met, a smaller test control variation that can generate an identifiable response can be prioritized to reduce the impact of the verification process on the normal operating state of the target equipment.

[0062] After the first test control input is executed, control commands, execution component status, energy consumption data, and service-side status data are collected according to the response time window corresponding to each node. After the first response data collection is completed, the target device is restored to the original control input. And wait for the execution component status and service side status to recover to the original operating status or the corresponding allowed recovery range.

[0063] S313. After the target device is restored to its original operating state, the original control quantity is adjusted to the second test control quantity, the second response data of each node in the cause response diagram is obtained, and the target device is restored to the original control quantity again, wherein the first test control quantity and the second test control quantity are located on both sides of the original control quantity, and the absolute change amount relative to the original control quantity is the same.

[0064] The second test control variable can be expressed as: Therefore, the following is satisfied: By performing recoverable tests in opposite directions on both sides of the original operating point, the response direction, response time, and recovery process of each response node can be observed when the control quantity increases and decreases. For a normal response chain, control changes in both directions should result in directional changes at the corresponding nodes that are consistent with the equipment's technical mechanism; for faulty nodes or response transmission relationships, the response direction, response time, or recovery process may differ from the expected conditions.

[0065] For building equipment with only discrete control quantities such as start and stop states, verification data can be obtained by state verification tasks or one recoverable state switching tasks, and bidirectional testing methods for continuous control quantities are not required.

[0066] S321. When performing the target verification task, identify multiple similar reference devices from the building equipment that have never performed the target verification task, which match the target equipment in terms of equipment type, operating mode, environmental conditions and service load.

[0067] Similar reference equipment maintains its original operating mode during the target verification task, serving to characterize common data changes caused by variations in outdoor environment, building usage, or public load during the verification period. Reference equipment is preferentially selected based on its current metering and equipment status being complete, and which does not have a direct control linkage with the target equipment, to reduce the possibility that the target verification actions will simultaneously affect the reference equipment.

[0068] S322. During the response time period corresponding to the first test control quantity and the second test control quantity, synchronously acquire the control commands, execution component status, energy consumption data and service side status data of the target device and each similar reference device.

[0069] State variables can be recorded using their original values ​​or 0 / 1 state codes; continuous variables are recorded using their original physical units. All data retains the device identifier, data type, and timestamp for separate processing in subsequent differential response calculations.

[0070] S323. Determine a common change amount based on the data change amount of each similar reference device, and form a differential response sequence corresponding to the target device based on the difference between the data change amount of the target device and the common change amount.

[0071] For any data item The target device relative to the reference time before the verification action was performed. The data change is: in, Indicates the target device number Class data at time The value of .

[0072] For similar reference equipment The amount of data change is: for The first similar reference device, the The common change of data of the same type can be expressed as: Target device The differential response of the data type is: in, This represents the target device response after deducting the common variations of similar reference devices. Different data types form corresponding differential responses, and data with different physical dimensions are not directly summed or subtracted from each other.

[0073] For direction determination, a response threshold corresponding to the physical dimensions of each data type can be set. .when A response is defined as positive if it is greater than a positive threshold, negative if it is less than a negative threshold, and stable if it is within a preset stable range. The response thresholds can be determined based on the normal historical fluctuation range of the corresponding data item.

[0074] Record the differential response direction, response start time, response amplitude, and recovery time of each node in a unified time sequence to form the differential response sequence corresponding to the target device.

[0075] The forward response threshold, reverse response threshold, and stable response range for each data item are determined separately, and different physical quantities do not share the same numerical threshold. The response threshold can be determined based on the historical fluctuation range of the corresponding data item during the normal and stable operation phase of the target equipment, the measurement accuracy of the corresponding sensor or metering equipment, and the results of on-site normal response tests, using the same physical dimensions and sampling period as the data being judged. When a data item is judged using a relative change, its corresponding response threshold is converted to a relative threshold in the same way.

[0076] S3231. Extract historical confirmed cases that match the equipment type, operating mode and target verification task of the target equipment from the historical anomaly handling data from which the rectification results have passed energy consumption verification.

[0077] Historically confirmed cases must meet at least the following conditions: the cause of the anomaly has been identified, the corresponding rectification task has been completed, and the energy consumption status has returned to normal after rectification. By using only historical cases that have completed closed-loop verification to determine response time characteristics, the impact of error cause labels and invalid rectification records on current verification judgments can be reduced.

[0078] S3232. Based on the execution time of the target verification task described in each historical confirmed case, as well as the response time of the control command, execution component status, energy consumption data, and service side status data, determine the historical response delay interval of each node in the cause response diagram.

[0079] For the Nodes in historically confirmed cases Its response delay can be expressed as: in, Indicates the first The execution time of the target verification task in a historical confirmed case; Represents a node The moment when a valid response occurs; Represents a node Response delay to target verification tasks.

[0080] The moment when a node produces a valid response can be determined as the initial moment when the node's differential response first reaches the corresponding response threshold and maintains the same response direction within a preset series of sampling points. This reduces misjudgments of the response time caused by single-point measurement noise.

[0081] Based on multiple historical confirmed cases, nodes were obtained. Historical response delay samples were collected and historical response delay intervals were formed: in, and Representing nodes respectively The lower and upper bounds of the historical response delay are defined. These bounds can be determined using the minimum and maximum values ​​after removing outliers, or according to a preset quantile interval of the historical response delay distribution.

[0082] S3233. Set the response time window corresponding to the current target verification task according to the historical response delay interval of each node, and form the differential response sequence according to the response direction, response start time and recovery time of each node within the corresponding response time window.

[0083] Verify the execution time of the current target task. Starting point, node The corresponding current response time window can be based on The differential response direction, response start time, response amplitude, and recovery time of the nodes are extracted within the corresponding response time window and saved according to the node order of the cause-response diagram and a unified time order, thus forming a differential response sequence for candidate cause judgment.

[0084] If a node lacks sufficient historical confirmed cases, the response time range obtained from equipment technical data, existing control system settings, or verification under normal operating conditions can be used as the initial response time window, and updated after accumulating historical confirmed cases.

[0085] S400. Based on the device status data or energy consumption data obtained after the execution of the target verification task and the judgment conditions corresponding to each candidate cause, retain the matching candidate causes.

[0086] This step uses the actual response generated by the target verification task already executed in the S300 phase, without repeating the same verification task. The actual response includes the differential response direction, response start time, response amplitude, recovery time, and corresponding data quality status.

[0087] S411. The actual response direction and actual response time of each node in the differential response sequence are compared with the expected response direction and expected response time window corresponding to each candidate cause.

[0088] A node satisfies the judgment condition corresponding to a candidate cause when its actual response direction is consistent with the expected response direction corresponding to the candidate cause, and the start time of the actual response is within the corresponding expected response time window. For nodes that only require a stable state, they are considered to satisfy the stable response condition when their differential response remains within the stable response range of the corresponding data item.

[0089] S412. Along the control response path, energy consumption response branch and service response branch of the cause-response diagram, determine the first response transmission relationship where the upstream node meets the expected response and the actual response direction of the downstream node deviates from the expected response direction or the actual response time exceeds the expected response time window, and determine the first response transmission relationship as the response breakpoint.

[0090] Specifically, the response propagation relationship is checked sequentially from the control configuration node to the execution component status node along the control response path. Then, from the execution component status node, the response propagation relationship is checked along the energy consumption response branch and the service response branch. For a response path, if the upstream node meets the expected response, but the downstream node adjacent to it deviates from the expected response direction or times out, the response propagation relationship between the adjacent nodes is determined as a response breakpoint on the path. If both the energy consumption response branch and the service response branch are abnormal, the first response breakpoint of each branch is recorded, and the candidate causes associated with any valid response breakpoint are retained.

[0091] By finding the first response breakpoint, we can avoid mistaking downstream derived anomalies as independent causes when upstream failures cause multiple downstream nodes to malfunction simultaneously.

[0092] S413. Retain candidate reasons associated with the node corresponding to the response breakpoint or the response transit relationship.

[0093] The cause database pre-records the correspondence between candidate causes and nodes in the cause-response graph or the response propagation relationship. When filtering causes based on response breakpoints, the candidate causes corresponding to the current response breakpoint are retained as the current candidate causes. For causes with complete and reliable time-synchronized actual response evidence, and whose response paths are inconsistent with the expected conditions of the candidate causes, they can be recorded as causes that have been verified, thereby avoiding the need to repeatedly select the same verification task for that cause in the future.

[0094] S414. When the data integrity of the node involved in the response breakpoint is lower than the preset integrity threshold, or the time synchronization deviation exceeds the preset synchronization threshold, the candidate cause associated with the node is marked as the cause to be verified, and a data verification task for the node is generated.

[0095] node Data completeness can be expressed as: in, Indicates the number of valid data points actually obtained within the response time window; This indicates the number of data points that should be obtained according to the sampling period of this data item; Represents a node Data completeness.

[0096] The preset integrity threshold can be determined based on the sampling period of the corresponding data item, the minimum number of valid samples required to complete the response judgment within the response time window, and the statistical results of the data integrity of normal historical data. Data integrity is the ratio of the number of valid sampling points to the theoretical number of sampling points, and is a dimensionless quantity. The preset synchronization threshold can be determined based on the time synchronization accuracy, communication period, and response time scale between adjacent response nodes that need to be distinguished for each data acquisition system; its unit is consistent with the unit of time synchronization deviation. For different data sources with different sampling periods or time synchronization capabilities, corresponding data integrity thresholds and time synchronization thresholds can be set separately.

[0097] Time synchronization deviation is determined based on the difference between the node data timestamp and the unified time base. When the data integrity meets the preset integrity condition, the time synchronization deviation meets the preset synchronization condition, and the data quality indicator is normal, the node data is determined to be reliable data. For nodes whose data quality does not meet the judgment requirements, a final cause conclusion is not directly formed based on the node; instead, a data verification task is generated to obtain the complete status data of the corresponding node.

[0098] S421. When the response breakpoint is located in the energy consumption response branch and the service response branch meets the corresponding expected response, a metering topology is established, including the upper-level metering point, the target metering point, and other metering points at the same metering level as the target metering point.

[0099] A satisfactory service response branch indicates that the target equipment has generated actual technical output consistent with the verification actions; at this point, a response breakpoint appears in the energy consumption response branch, and the anomaly range can be further limited to the difference between the actual actions of the target equipment and the metering results. The metering topology can be established based on the existing master table, sub-table, and equipment metering point hierarchy in the building sub-metering system.

[0100] S422. Within the response time window corresponding to the target verification task, obtain the energy consumption changes of the superior metering point, the target metering point, and other metering points, and determine the response metering point according to the correspondence between the energy consumption change direction and response time of each metering point and the state of the target device execution component.

[0101] For any measurement point In a unified response time window The change in energy consumption within can be expressed as: in, and Representing measurement points The cumulative energy consumption at the start and end of the response time window. If periodic energy consumption is used, the energy consumption changes corresponding to the same statistical period before and after the verification action can be directly compared.

[0102] When determining the response measurement point, it is necessary not only to compare the energy consumption changes under a single test, but also to combine the verification results of two tests with opposite directions corresponding to the first and second test control variables. When the energy consumption change direction under the first test control, the energy consumption change direction under the second test control, and the recovery process of a certain measurement point all match the change direction and timing of the target equipment's actuator state, that measurement point is determined as the response measurement point. Utilizing consistent responses through bidirectional verification can reduce misjudgments caused by accidental changes in other loads on the same floor.

[0103] S423. Determine the metering balance residual based on the difference between the energy consumption change at the higher-level metering point and the sum of the energy consumption changes at each lower-level metering point.

[0104] Let the upper-level metering point be Its subordinate set of measurement points is The measurement balance residual can then be expressed as: in, This represents the measurement balance residual; This indicates the change in energy consumption at the higher-level metering point; Indicates the lower-level metering point The energy consumption changes are calculated using the same time window and the same energy unit.

[0105] The metering balance threshold can be determined based on the allowable error of the metering equipment, normal communication error, normal influence of non-independent metering branches, and allowable line loss, or it can be determined based on the statistical range of metering balance residuals corresponding to the metering topology in normal historical periods.

[0106] S424. When the response measurement point is one of the other measurement points, and the measurement balance residual satisfies the measurement balance condition, retain the candidate cause corresponding to the association configuration error between the target device and the measurement point.

[0107] For example, if the target device is configured with target metering point A, but during bidirectional verification, metering point A fails to generate a response matching the target device's actions, while metering point B at the same metering level generates energy consumption responses matching the target device's actions in two tests conducted in opposite directions, and both the upper and lower level metering still meet the normal balance range, this indicates that the energy consumption changes generated by the target device have been correctly recorded by the metering system, but there is a deviation in the configuration correspondence between the device and the metering point. In this case, the candidate cause corresponding to the incorrect association configuration should be retained, and metering point B can be used as the object for subsequent association verification.

[0108] S425. When the energy consumption changes of the target metering point and other metering points are both lower than the response judgment threshold, and the metering balance residual exceeds the metering balance threshold, retain the candidate reasons corresponding to the missed metering data or the abnormality of the metering equipment.

[0109] The response judgment threshold for each metering point can be determined based on the measurement resolution of the metering equipment, the energy consumption fluctuation range of the corresponding metering point during normal and stable operation, and the metering response amplitude generated by normal equipment under the same or similar verification tasks. The response judgment threshold and the energy consumption change of the corresponding metering point use the same metering time window and energy unit to distinguish between the effective metering response caused by the target verification task and metering noise or normal load fluctuations. For metering points with different rated capacities or metering ranges, response judgment thresholds matching their respective metering characteristics can be set separately.

[0110] Before making this judgment, first confirm that the data integrity and time synchronization status of the superior metering point meet the reliable data conditions. When the execution component status and service side status of the target device both indicate that the device has responded to the verification action, but none of the subordinate metering points have generated a corresponding energy consumption response, and the energy consumption change of the superior metering point causes the metering balance residual to exceed the normal range, it indicates that the corresponding energy consumption change of the target device is not fully reflected by the current subordinate metering data. Candidate reasons such as missing metering data, abnormal metering communication, or abnormal metering equipment can be further retained, and a corresponding data verification task can be generated.

[0111] S431. When there are at least two candidate causes remaining, the candidate causes are combined in pairs to form undistinguished cause pairs.

[0112] Let the current set of candidate reasons be: The set of undistinguished causes can then be represented as: Undistinguished cause pairs represent two candidate causes that still need to be distinguished by subsequent response data after the current verification.

[0113] S432. Based on the number of unidentified cause pairs that each preset verification task can distinguish, as well as the corresponding control quantity change range, expected service-side state deviation, and expected recovery time, filter out preset verification tasks whose control quantity change range, expected service-side state deviation, or expected recovery time exceeds the corresponding allowable conditions, and determine the target verification task that can distinguish the most unidentified cause pairs from the remaining preset verification tasks.

[0114] For each preset verification task The verification task compares the expected responses of two candidate causes within the same undistinguished cause pair under the given verification task. The verification task is considered capable of distinguishing the undistinguished cause pair if at least one reliable response node corresponds to a different expected response direction, or if the expected response time windows have identifiable differences.

[0115] Set up verification task The number of indistinguishable causes is First, verify the range of control quantity changes, expected service-side status deviation, and expected recovery time according to equipment safety conditions and building service conditions. Only add verification tasks where all conditions are within their respective allowable ranges to the set of executable verification tasks. Changes in control quantities with different dimensions, service-side state deviations, and recovery times are constrained and judged separately, and are not combined into a single evaluation quantity.

[0116] The permissible conditions for the range of change in control quantities can be determined based on the rated parameters of the target equipment, the manufacturer's allowable adjustment range, and the existing control limits of the building automation system. The permissible conditions for the expected service-side state deviation can be determined based on the normal service state range of the building equipment, the existing control setting range, or the building's commissioning and operation requirements. The permissible conditions for the expected recovery time can be determined based on the length of the verification period, the historical recovery time of the target equipment, and the start time of subsequent service tasks. These three types of conditions are judged separately, and different physical quantities are not directly combined numerically.

[0117] Determined from the set of executable verification tasks: in, This represents the target verification task that can currently distinguish the most pairs of unidentified causes and satisfies the execution constraints. If multiple verification tasks correspond to... If the data verification tasks are identical, prioritize those that do not change the equipment's operating status; if they still cannot be distinguished, then select the recoverable control verification task. After determining the new target verification task, return to S300 to continue verification until the candidate cause meets the conditions for generating a rectification task.

[0118] Reference Figure 4 S500. Generate rectification tasks for building equipment, control configuration, or metering configuration based on the candidate reasons.

[0119] In this embodiment, after one or more rounds of target verification tasks, when the candidate cause converges into a candidate cause with a clear rectification action, and the candidate cause corresponds to the response breakpoint determined in the cause response diagram, the candidate cause is determined as the target cause, and a rectification task is generated based on the target cause.

[0120] When the target cause corresponds to a control configuration node, rectification tasks may include modifying the operating schedule, forcibly deactivating the operating status, or correcting control parameters. When the target cause corresponds to the response transmission relationship between control commands and the status of execution components, rectification tasks may include verification and maintenance of execution components, frequency converters, contactors, or control output links. When the target cause corresponds to the configuration of equipment associated with metering points, rectification tasks may include correcting metering association relationships. When the target cause corresponds to abnormal metering data acquisition, rectification tasks may include verification of metering communication links or metering equipment. Rectification tasks must at least include the rectification object, rectification content, permissible execution conditions, and rectification result feedback requirements.

[0121] After the rectification task is completed, under the same or comparable equipment operating conditions as before the rectification, the target verification task to confirm the cause of the problem is executed again, and the differential response sequence and metering data after rectification are re-acquired. The corrected energy consumption residual after rectification can be expressed as: in, This indicates the corrected energy consumption residual of the target equipment after rectification; This indicates the energy consumption deviation of the target equipment relative to its own energy consumption baseline after rectification; This indicates the common energy consumption deviation of similar reference equipment after rectification.

[0122] When all relevant nodes in the cause-response diagram after rectification generate responses according to the expected response direction and corresponding response time window, and the corrected energy consumption residual after rectification meets the normal operating conditions, it is determined that the continuous energy consumption anomaly has been eliminated. The target verification task, actual response, target cause, rectification task, and rectification verification result are written into the historical anomaly handling data, and the record is marked as a historical confirmed case.

[0123] If persistent energy consumption anomalies persist after rectification, the cause screening results supported by reliable data are retained. Candidate causes with data quality statuses of pending verification or requiring further differentiation are re-included into the current candidate cause set, and the target verification task is redefined according to S431-S432. This allows subsequent verification to inherit reliable diagnostic evidence already established, reducing redundant verification and the re-execution of the same ineffective rectification measures.

[0124] In one implementation, judgment conditions corresponding to the identification of continuous energy consumption anomalies can be used to determine the anomaly elimination status after rectification. This is achieved when the corrected energy consumption residual remains below the anomaly threshold within a preset verification period. When the response node related to the target cause generates a response according to the normal expected response direction and response time window, it is determined that the current continuous energy consumption anomaly has been eliminated. The preset verification period can be determined based on the continuous anomaly determination duration in S213, the normal operating cycle of the target device, and the recovery process after the verification task is executed.

[0125] This embodiment identifies continuous energy consumption anomalies by combining the target device's own historical baseline with common changes in similar reference devices. It generates distinguishable device responses through recoverable verification tasks and uses historical confirmed cases to form response time windows, further locating response breakpoints along the cause-response diagram. For anomalies in the energy consumption response branch, it also performs cause verification by combining the actual service response of the device and the metering topology, so that the cause judgment of continuous energy consumption anomalies corresponds to the actual control, execution, energy consumption and service response process of building equipment, and forms rectification tasks for building equipment, control configuration or metering configuration accordingly.

[0126] Example 2 Reference Figure 5 A building energy consumption management system based on big data includes: The data acquisition module is used to acquire historical building metering data, current metering data, and equipment operation data. The candidate cause identification module is used to identify continuous energy consumption anomalies based on historical building metering data, current metering data, and equipment operation data, and to determine multiple candidate causes corresponding to the continuous energy consumption anomalies. The target verification task determination module is used to determine the target verification task from the preset verification tasks. The target verification task is used to obtain device status data or energy consumption data that can distinguish at least two candidate causes. The candidate cause verification module is used to perform the target verification task and retain the matching candidate causes based on the obtained device status data or energy consumption data and the judgment conditions corresponding to each candidate cause. The rectification task generation module is used to generate rectification tasks for building equipment, control configuration, or metering configuration based on the candidate reasons.

[0127] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0128] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0129] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0130] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0131] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0132] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from the scope of the invention. The spirit and scope of the invention are as follows: Thus, if these modifications and variations of the invention fall within the scope of the claims of the invention and their equivalents, the invention is also intended to include these modifications and variations.

Claims

1. A building energy consumption management method based on big data, characterized in that, include: Acquire historical building metering data, current metering data, and equipment operation data; Based on historical building metering data, current metering data, and equipment operation data, identify continuous energy consumption anomalies and determine multiple candidate causes corresponding to the continuous energy consumption anomalies; A target verification task is determined from the preset verification tasks. The target verification task is used to obtain device status data or energy consumption data that can distinguish at least two candidate causes. Execute the target verification task, and retain the matching candidate causes based on the obtained device status data or energy consumption data and the judgment conditions corresponding to each candidate cause; Based on the candidate reasons, rectification tasks are generated for building equipment, control configurations, or metering configurations.

2. The building energy consumption management method based on big data according to claim 1, characterized in that, The method of identifying persistent energy consumption anomalies based on historical building metering data, current metering data, and equipment operation data includes: The building equipment to be analyzed is identified as the target equipment. Based on the equipment's operating mode, time type, and environmental conditions, comparable historical metering data of the target equipment is extracted from the building's historical metering data. The energy consumption baseline of the target equipment is then determined based on the comparable historical metering data. From building equipment with the same equipment type and operating mode as the target equipment, identify similar reference equipment that does not currently exhibit energy consumption anomalies, and determine a common energy consumption deviation based on the energy consumption deviation of the similar reference equipment relative to its own energy consumption baseline. The corrected energy consumption residual is determined based on the difference between the energy consumption deviation of the target device relative to its own energy consumption baseline and the common energy consumption deviation. When the corrected energy consumption residual continuously meets the abnormal conditions, the continuous energy consumption abnormality is determined.

3. The building energy consumption management method based on big data according to claim 2, characterized in that, The determination of multiple candidate causes corresponding to the continuous energy consumption anomaly includes: Based on the device identifier and unified time reference, the control configuration data, control commands, execution component status and service side status data in the device operation data, and the energy consumption data in the current metering data are associated to form a cause response diagram corresponding to the continuous energy consumption anomaly. The cause-response diagram includes a control response path consisting of the control configuration data, control commands, and execution component status in sequence, as well as an energy consumption response branch and a service response branch that are respectively pointed to by the execution component status from the energy consumption data and the service side status data. Each candidate cause is associated with a node in the cause-response graph or with the response propagation relationship between adjacent nodes, and the expected response direction and expected response time window of the corresponding node are configured for each candidate cause.

4. The building energy consumption management method based on big data according to claim 3, characterized in that, When the target device has a continuously adjustable control quantity, the target verification task includes: Save the original control values ​​of the target device; During the verification period that meets the equipment safety conditions and building service conditions, the original control quantity is adjusted to the first test control quantity, the first response data of each node in the cause-response diagram is obtained, and the target equipment is restored to the original control quantity. After the target device is restored to its original operating state, the original control quantity is adjusted to the second test control quantity, the second response data of each node in the cause response diagram is obtained, and the target device is restored to the original control quantity again. The first test control quantity and the second test control quantity are located on both sides of the original control quantity, and their absolute changes relative to the original control quantity are the same.

5. The building energy consumption management method based on big data according to claim 4, characterized in that, When performing the target verification task, multiple similar reference devices that match the target device in terms of equipment type, operating mode, environmental conditions and service load are identified from the building equipment that has never performed the target verification task. During the response time period corresponding to the first test control quantity and the second test control quantity, the control commands, execution component status, energy consumption data and service side status data of the target device and each similar reference device are acquired synchronously. A common change is determined based on the data change of each similar reference device, and a differential response sequence corresponding to the target device is formed based on the difference between the data change of the target device and the common change.

6. The building energy consumption management method based on big data according to claim 5, characterized in that, The process of forming the differential response sequence corresponding to the target device includes: Extract historical anomaly handling data from the historical data on rectification results that have passed energy consumption verification, and extract historical confirmed cases that match the equipment type, operating mode and target verification task of the target equipment. Based on the execution time of the target verification task described in each historical confirmed case, as well as the response time of the control command, execution component status, energy consumption data, and service side status data, the historical response delay interval of each node in the cause response diagram is determined respectively. The response time window corresponding to the current target verification task is set according to the historical response delay interval of each node, and the differential response sequence is formed according to the response direction, response start time and recovery time of each node within the corresponding response time window.

7. The building energy consumption management method based on big data according to claim 6, characterized in that, The step of retaining matching candidate causes based on the acquired device status data or energy consumption data and the judgment conditions corresponding to each candidate cause includes: The actual response direction and actual response time of each node in the differential response sequence are compared with the expected response direction and expected response time window corresponding to each candidate cause, respectively. Along the control response path, energy consumption response branch and service response branch of the cause response diagram, determine the first response transmission relationship where the upstream node meets the expected response and the actual response direction of the downstream node deviates from the expected response direction or the actual response time exceeds the expected response time window, and determine the first response transmission relationship as the response breakpoint. Retain candidate reasons associated with the node corresponding to the response breakpoint or the response propagation relationship; When the data integrity of the node involved in the response breakpoint is lower than the preset integrity threshold, or the time synchronization deviation exceeds the preset synchronization threshold, the candidate cause associated with that node is marked as a cause to be verified, and a data verification task for that node is generated.

8. The building energy consumption management method based on big data according to claim 7, characterized in that, When the response breakpoint is located in the energy consumption response branch, and the service response branch meets the corresponding expected response, a metering topology is established, including the upper-level metering point, the target metering point, and other metering points at the same metering level as the target metering point. Within the response time window corresponding to the target verification task, the energy consumption changes of the superior metering point, the target metering point, and other metering points are obtained, and the response metering point is determined according to the correspondence between the energy consumption change direction and response time of each metering point and the state of the target device execution component. The metering balance residual is determined by the difference between the energy consumption change at the higher-level metering point and the sum of the energy consumption changes at each lower-level metering point. When the response metering point is one of the other metering points, and the metering balance residual satisfies the metering balance condition, the candidate cause corresponding to the association configuration error between the target device and the metering point is retained. When the energy consumption changes of the target metering point and other metering points are both lower than the response judgment threshold, and the metering balance residual exceeds the metering balance threshold, candidate causes corresponding to missed metering data collection or metering equipment malfunction are retained.

9. The building energy consumption management method based on big data according to claim 8, characterized in that, When there are at least two candidate causes retained, the candidate causes are combined in pairs to form undistinguished cause pairs; Based on the number of unidentified cause pairs that each preset verification task can distinguish, as well as the corresponding control variable change range, expected service-side state deviation, and expected recovery time, preset verification tasks that exceed the corresponding allowable conditions in terms of control variable change range, expected service-side state deviation, or expected recovery time are screened out, and the target verification task that can distinguish the most unidentified cause pairs is determined from the remaining preset verification tasks. When the retained candidate reasons converge to one and correspond to the response breakpoint, a rectification task is generated based on the candidate reason; After the rectification task is executed, the target verification task is executed again. Based on the rectified differential response sequence and metering data, it is determined whether the continuous energy consumption anomaly has been eliminated. When the rectified differential response sequence or metering data still meets the judgment conditions corresponding to the continuous energy consumption anomaly, the candidate cause screening results supported by data that meets the data integrity and time synchronization conditions are retained, the candidate causes lacking reliable data support are restored as causes to be verified, and the target verification task is redefined.

10. A building energy consumption management system based on big data, characterized in that, The building energy consumption management method based on big data according to any one of claims 1-8 includes: The data acquisition module is used to acquire historical building metering data, current metering data, and equipment operation data. The candidate cause identification module is used to identify continuous energy consumption anomalies based on historical building metering data, current metering data, and equipment operation data, and to determine multiple candidate causes corresponding to the continuous energy consumption anomalies. The target verification task determination module is used to determine the target verification task from the preset verification tasks. The target verification task is used to obtain device status data or energy consumption data that can distinguish at least two candidate causes. The candidate cause verification module is used to perform the target verification task and retain the matching candidate causes based on the obtained device status data or energy consumption data and the judgment conditions corresponding to each candidate cause. The rectification task generation module is used to generate rectification tasks for building equipment, control configuration, or metering configuration based on the candidate reasons.