A three-dimensional space-based visualized property management method and system

Abstract

The application discloses a three-dimensional space-based visual property management method and system, and belongs to the technical field of equipment operation and maintenance. The method comprises the following steps: collecting equipment historical operation logs, constructing statistical characteristic vectors, generating equipment behavior abnormal signal; taking the abnormal equipment as a starting node, searching associated equipment along the spatial connection relationship and giving a propagation weight to generate a spatial associated propagation path graph; calculating a joint abnormal risk value according to the cumulative propagation weight and the deviation degree index of the associated equipment to screen a set of attention equipment; connecting the attention equipment according to the propagation direction to generate a set of spatial associated path; weighting the historical deviation degree index according to the joint abnormal risk value to predict a future value, and when the predicted value exceeds a threshold value, marking the equipment and the path segment where the equipment is located as a proactive intervention item to generate a decision list and push the decision list to a management terminal. The application solves the problems that the existing property management system cannot make fault prediction based on equipment operation behavior variation, and cannot deduce the propagation path of the abnormality among the spatial adjacent equipment.

Description

Technical Field

[0001] This invention relates to the field of equipment operation and maintenance technology, and in particular to a property management method and system based on three-dimensional spatial visualization. Background Technology

[0002] In recent years, as the operation and maintenance management of smart parks and large public buildings has developed towards refinement and intelligence, 3D visualization technology has been gradually introduced into the property management field. Existing 3D visualization property management systems typically collect operational status data of controlled equipment through device gateways, map alarm events onto 3D spatial models for location display, and link with work order systems to dispatch maintenance tasks. Some systems also integrate historical data storage and query functions, supporting post-event traceability of equipment operation records. These systems have reached a certain level of maturity in equipment status monitoring and visualization, enabling location indication and work order flow after an alarm occurs.

[0003] However, the existing systems still have the following shortcomings: First, the system relies entirely on the physical threshold judgment of the equipment itself to trigger abnormal events, and can only respond passively after the equipment parameters exceed the limits or a fault alarm occurs. It lacks a means of proactively identifying abnormalities based on equipment operation behavior data, and cannot capture abnormal signals in advance during the fault latency period. Second, the system only independently locates and marks individual alarming devices in three-dimensional space, failing to establish a spatial adjacency model between devices. When a device malfunctions, the system cannot deduce the possible paths that the abnormality might propagate to surrounding devices through spatial connectivity, such as between rooms, adjacent rooms, or floors. Maintenance personnel find it difficult to quickly determine the potential impact range of the fault and the related devices requiring early intervention. These shortcomings cause the existing system to consistently lag behind the abnormal development process in maintenance decisions, making it difficult to achieve proactive maintenance and risk prevention. Summary of the Invention

[0004] The technical problem solved by this invention is that existing property management systems cannot predict faults based on variations in equipment operation behavior, nor can they deduce the propagation path of anomalies between adjacent devices in space.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: As a preferred embodiment of the three-dimensional spatial visualization-based property management method described in this invention, wherein: Collect historical operation event logs of each controlled device, construct an operation time series with device identifier as the primary key, and extract the operation frequency, operation interval variation coefficient and operation type entropy value within each time window as statistical feature vectors. The statistical feature vectors of each time window are compared with the group feature distribution of the same type of controlled equipment to calculate the deviation index of operation behavior. When the deviation index of operation behavior exceeds the preset threshold, an abnormal equipment behavior signal is generated. Based on the device identifier in the abnormal device behavior signal, the building space boundary data and the installation coordinates of the controlled device are read from the storage unit. Taking the abnormal device as the starting node and the spatial connectivity relationship as the edge, the set of associated devices that are spatially adjacent to the abnormal device is searched outward level by level. The propagation weight is assigned according to the search level to generate a spatial association propagation path diagram. Based on the cumulative propagation weight value of each associated device in the spatial correlation propagation path diagram and the deviation index of their respective operational behaviors, the joint anomaly risk value of each associated device is calculated, and associated devices whose joint anomaly risk value exceeds a preset risk threshold are selected as the set of devices of interest. Based on the three-dimensional spatial coordinates of each device in the set of devices of interest, and taking the three-dimensional coordinates of the abnormal device as the center, the three-dimensional coordinates of each device of interest are connected step by step according to the propagation direction in the spatial association propagation path diagram to generate a set of spatial association paths. Based on the joint anomaly risk value and the time-series change trend of the operational behavior deviation index of each concerned device in the spatial association path set, the operational behavior deviation index within the future time window is predicted to obtain the predicted value of the operational behavior deviation index. When the predicted value of the deviation index of the operation behavior exceeds the preset intervention threshold, the path segment where the device of interest is located in the spatial associated path set is marked as an item requiring active intervention, and an active operation and maintenance decision list is generated and pushed to the management terminal.

[0006] Furthermore, historical operation event logs of each controlled device are collected, and an operation time series is constructed using the device identifier as the primary key. The operation frequency, operation interval coefficient of variation, and operation type entropy value within each time window are extracted as statistical feature vectors, including: Collect historical operation event logs for each controlled device; the operation event logs include device identifier, operation type, and operation timestamp. Using the device identifier as the primary key, operation events belonging to the same controlled device are arranged in ascending order of time and resampled at equal intervals to a uniform time granularity to construct the operation time sequence of each controlled device. The operation time series of each controlled device is divided by using a sliding window to obtain the operation event subsequence within each time window; For each subsequence of operation events within a time window, the total number of times the controlled device is operated within that time window is counted based on the operation timestamp as the operation frequency. The ratio of the standard deviation to the average value of the time interval between two adjacent operations is calculated as the operation interval variation coefficient. The frequency of occurrence of each operation type is counted based on the operation type, and the entropy value is calculated as the operation type entropy value. A statistical feature vector is constructed based on the operation frequency, the operation interval variation coefficient, and the operation type entropy value.

[0007] Furthermore, the statistical feature vectors of each time window are compared with the group feature distribution of similar controlled equipment to calculate the operational behavior deviation index. When the operational behavior deviation index exceeds a preset threshold, an abnormal equipment behavior signal is generated, including: Using the type of controlled equipment as the classification label, the historical operation event logs of all controlled equipment are grouped by equipment type. The statistical feature vectors of each time window of all controlled equipment in each group are summarized to form the group feature distribution of each type of controlled equipment. For any controlled device at any given time window, the statistical feature vector of that time window is input into the Isolation Forest algorithm. The Isolation Forest algorithm uses the population feature distribution as a reference benchmark, constructs multiple binary trees by randomly selecting features and split values, calculates the outlier score of the statistical feature vector relative to the population feature distribution, and uses the outlier score as the deviation index of the controlled device's operational behavior in that time window. The operational behavior deviation index is used to characterize the degree to which the operational behavior of the controlled device deviates from the normal level of a group of similar devices. The deviation index of the operation behavior is compared with a preset threshold. When the deviation index of the operation behavior exceeds the preset threshold, it is determined that the controlled device has an abnormal operation behavior in the time window, and a device behavior abnormality signal containing the device identifier of the controlled device and the timestamp of the abnormality is generated.

[0008] Further, based on the device identifier in the device behavior anomaly signal, the building space boundary data and the installation coordinates of the controlled device are read from the storage unit. Taking the abnormal device as the starting node and spatial connectivity as the edges, a set of associated devices spatially adjacent to the abnormal device is searched outwards level by level. Propagation weights are assigned according to the search level to generate a spatial association propagation path graph, including: Based on the device identifier in the abnormal device behavior signal, the building space boundary data and the installation coordinates of the controlled device are read from the storage unit; the building space boundary data includes the hierarchical structure and outline of buildings, floors, and rooms; the spatial connectivity includes interconnection of devices in the same room, connection of adjacent rooms through wall holes or cable trays, and connection of upper and lower floors through pipe shafts or cable shafts. Based on the building space boundary data and the installation coordinates of the controlled equipment, the abnormal equipment is used as the starting node and the spatial connectivity is used as the edge. The search is carried outward level by level to find the associated equipment that is spatially adjacent to the abnormal equipment until the preset search depth limit is reached, so as to obtain each associated equipment and its corresponding search level. Based on each associated device and its corresponding search level, propagation weights are assigned in descending order from near to far, with higher propagation weights for lower search levels. A spatial association propagation path graph is generated using the abnormal device and each associated device as nodes, the spatial connectivity as edges, and the propagation weights as edge weights.

[0009] Furthermore, based on the building space boundary data and the installation coordinates of the controlled equipment, taking the abnormal device as the starting node and the spatial connectivity as the edges, the system searches outwards level by level for associated devices that are spatially adjacent to the abnormal device until a preset search depth limit is reached, thus obtaining each associated device and its corresponding search level, including: Based on the installation coordinates of the target controlled equipment and the building space boundary data, the target room where the target controlled equipment is located is determined; Starting with the target room as the initial search area, and using the spatial connectivity as edges, the search area is expanded outwards level by level: Search for all controlled devices belonging to the same room as the target room, and designate them as the first associated devices, with the search level being the first level; Using the target room as the center, search for controlled devices in all rooms adjacent to the target room, which are considered as second associated devices, and the search level is the second level. Search for controlled devices in rooms located on floors directly above and below the target room, and use them as the third associated devices, with the search level being the third level. The first associated device, the second associated device, and the third associated device found in the search are summarized to obtain each associated device and its corresponding search level.

[0010] Furthermore, based on the cumulative propagation weight values ​​of each associated device in the spatial correlation propagation path diagram and their respective operational behavior deviation indicators, the joint anomaly risk value of each associated device is calculated. Associated devices whose joint anomaly risk values ​​exceed a preset risk threshold are selected as a set of devices of interest, including: Traverse each associated device in the spatial association propagation path graph. For each associated device, accumulate the propagation weights of all incoming edges of that associated device to obtain the cumulative propagation weight value of that associated device. Obtain the deviation index of the operation behavior of each associated device; For each associated device, the cumulative propagation weight value is used as the risk propagation factor of the associated device in the spatial propagation path graph, and the operational behavior deviation index is used as the anomaly factor of the associated device itself. The cumulative propagation weight value and the operational behavior deviation index are multiplied and then divided by the number of incoming edges of the associated device in the spatial propagation path graph to obtain the joint anomaly risk value of the associated device. The joint anomaly risk value is used to characterize the comprehensive risk level of the associated device after being affected by spatial propagation and its own anomalies. The combined anomaly risk value of each associated device is compared with a preset risk threshold, and associated devices whose combined anomaly risk value exceeds the preset risk threshold are selected as the set of devices of interest.

[0011] Furthermore, based on the three-dimensional spatial coordinates of each device in the set of devices of interest, and taking the three-dimensional coordinates of the abnormal device as the center, the three-dimensional coordinates of each device of interest are connected step by step according to the propagation direction in the spatial association propagation path diagram to generate a spatial association path set, including: Obtain the three-dimensional spatial coordinates of each device of interest and the search level corresponding to each device of interest from the set of devices of interest; Using the three-dimensional spatial coordinates of the abnormal device as the starting point of the path, all devices of interest corresponding to the current level are determined step by step in order from low to high search level for each device of interest. For all the devices of interest corresponding to each search level, connect the three-dimensional spatial coordinates of every two adjacent devices of interest in the current level to generate the horizontal sub-path of the current level; Connect the two points with the closest spatial distance between the three-dimensional spatial coordinates of each interested device in the previous level and the three-dimensional spatial coordinates of each interested device in the current level to generate a vertical sub-path between adjacent levels; By alternately connecting horizontal and vertical sub-paths at all levels in order of increasing search level, and marking the direction of each sub-path in order of increasing search level, a set of spatially associated paths is obtained.

[0012] Furthermore, based on the joint anomaly risk value and the time-series change trend of the operational behavior deviation index of each interested device in the spatial association path set, the operational behavior deviation index within the future time window is predicted to obtain the predicted value of the operational behavior deviation index, including: The deviation index of the operation behavior of each device of interest in the spatial association path set is extracted in multiple consecutive historical time windows and arranged in chronological order to form a time sequence of the operation behavior deviation index of each device of interest. Extract the joint anomaly risk value of each device of interest from the spatially associated path set; For each device of interest, the joint anomaly risk value of that device of interest is used as a weighting coefficient to weight the deviation index of each time window in the time series of the operational behavior deviation index, resulting in a weighted deviation index time series; wherein, the higher the joint anomaly risk value, the greater the weighting impact on the weighted deviation index time series. The weighted rate of change is calculated based on the difference between adjacent time windows in the time series of the weighted deviation index. The weighted deviation index value of the current time window and the weighted rate of change are used as the predicted value of the deviation index of the operating behavior of the device under interest in the future time window.

[0013] Furthermore, when the predicted value of the operational behavior deviation index exceeds a preset intervention threshold, the path segment containing the device of interest in the spatially associated path set is marked as an item requiring proactive intervention, and a proactive operation and maintenance decision list is generated and pushed to the management terminal, including: The predicted values ​​of the operational behavior deviation index of each monitored device are compared with the preset intervention threshold. Devices whose predicted values ​​of the operational behavior deviation index exceed the preset intervention threshold are marked as devices that require active intervention. Traverse each path segment in the spatially associated path set, and mark the path segment containing the device requiring active intervention as the path segment requiring intervention; Starting from the abnormal device, along the propagation direction in the set of spatially associated paths, record all path segments from the abnormal device to each device requiring active intervention and all devices of interest on each path segment, and generate an active operation and maintenance decision list. The proactive maintenance decision list is pushed to the management terminal.

[0014] As a preferred embodiment of the three-dimensional spatial visualization-based property management system described in this invention, wherein: The operation log collection module is used to collect historical operation event logs of each controlled device, construct an operation time series with the device identifier as the primary key, and extract the operation frequency, operation interval variation coefficient and operation type entropy value within each time window as statistical feature vectors. The behavior deviation calculation module is used to compare the statistical feature vectors of each time window with the group feature distribution of the same type of controlled equipment, calculate the operation behavior deviation index, and generate an abnormal equipment behavior signal when the operation behavior deviation index exceeds the preset threshold. The spatial association propagation path generation module is used to read building space boundary data and controlled equipment installation coordinates from the storage unit based on the device identifier in the device behavior abnormal signal. Taking the abnormal device as the starting node and the spatial connectivity relationship as the edge, it searches outward level by level for a set of associated devices that are spatially adjacent to the abnormal device, and assigns propagation weights according to the search level to generate a spatial association propagation path diagram. The joint anomaly risk assessment module is used to calculate the joint anomaly risk value of each associated device based on the cumulative propagation weight value and the deviation index of each device's operation behavior in the spatial correlation propagation path diagram, and to screen the associated devices whose joint anomaly risk value exceeds a preset risk threshold as a set of devices of concern. The spatial association path generation module is used to generate a spatial association path set by connecting the three-dimensional spatial coordinates of each device of interest in the set of devices of interest, with the three-dimensional coordinates of the abnormal device as the center, and following the propagation direction in the spatial association propagation path diagram. The deviation prediction module is used to predict the deviation index of operation behavior within a future time window based on the joint abnormal risk value of each concerned device in the spatial association path set and the time-series change trend of the deviation index of operation behavior, and to obtain the predicted value of the deviation index of operation behavior. The proactive operation and maintenance decision module is used to mark the path segment where the device of interest is located in the spatially associated path set as an item requiring proactive intervention when the predicted value of the deviation index of the operation behavior exceeds the preset intervention threshold, and to generate a proactive operation and maintenance decision list and push it to the management terminal.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. This invention extracts operation frequency, operation interval variation coefficient, and operation type entropy value from equipment operation event logs as statistical feature vectors, and compares them with the group feature distribution of similar equipment. This allows for early identification of abnormal signals based solely on statistical deviations in operational behavior, even before equipment physical parameters exceed limits or alarms are triggered, thus shifting the anomaly detection point to the fault latency period. Furthermore, this invention uses the abnormal equipment as the starting node to search for related equipment level by level along spatial connectivity and assigns propagation weights. This transforms a single-point anomaly into a spatially related path with propagation direction and impact range, enabling maintenance personnel to directly know the related equipment that may be affected by the anomaly and its risk level. Based on the prediction results, proactive intervention can be implemented for equipment about to exceed the intervention threshold, thereby completing risk management before the actual fault occurs. This solves the problem that existing property management systems cannot predict faults based on variations in equipment operation behavior, nor can they deduce the propagation path of anomalies among spatially adjacent equipment.

[0016] 2. This invention directly extracts three statistical features from the existing operation event logs of the equipment: operation frequency, operation interval variation coefficient, and operation type entropy value. It can identify abnormalities in the health status of the equipment without the need to deploy additional physical sensors such as vibration and temperature sensors. It is also applicable to old equipment in existing buildings that do not have sensor interfaces.

[0017] 3. This invention takes the abnormal device as the starting node and searches for related devices level by level along three types of spatial relationships: interconnection within the same room, connection between adjacent rooms, and connection between upper and lower floors. The propagation weight is assigned decreasing from near to far, thus transforming the originally isolated single-point anomaly into a spatial association path with propagation direction and influence weight. Operation and maintenance personnel can directly know the related devices that may be affected by the anomaly and their risk level, without the need for manual investigation based on drawings.

[0018] 4. This invention uses the joint abnormal risk value as a weighting coefficient to weight the time series of deviation index before predicting future values. This gives higher weight to devices with greater spatial propagation impact in the prediction. When the predicted value exceeds the intervention threshold, the device and its path segment are marked as items requiring active intervention. The complete path segment from the abnormal device to the device is pushed to the management terminal, so that maintenance personnel can check and intervene step by step according to the path sequence before the fault actually occurs. Attached Figure Description

[0019] Figure 1 A basic flowchart of a three-dimensional spatial visualization-based property management method is provided as an embodiment of the present invention. Figure 2 This is a basic flowchart of a three-dimensional spatial visualization property management system provided as an embodiment of the present invention. Detailed Implementation

[0020] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0021] Example 1 like Figure 1 As shown in the figure, this embodiment introduces a property management method based on three-dimensional spatial visualization, including: Step 1: Collect historical operation event logs of each controlled device, construct an operation time series with the device identifier as the primary key, and extract the operation frequency, operation interval variation coefficient, and operation type entropy value within each time window as statistical feature vectors.

[0022] This embodiment directly extracts three statistical features—operation frequency, operation interval variation coefficient, and operation type entropy value—from the existing operation event logs of the equipment. It can provide a data foundation for subsequent anomaly identification without the need to deploy any additional physical sensors, and is also applicable to old equipment in existing buildings that do not have sensor interfaces.

[0023] Step 2: Compare the statistical feature vectors of each time window with the group feature distribution of similar controlled equipment, calculate the deviation index of operation behavior, and generate an abnormal equipment behavior signal when the deviation index of operation behavior exceeds the preset threshold.

[0024] This embodiment uses the group feature distribution of similar devices as a reference benchmark and employs the isolated forest algorithm to calculate the outlier score of the feature vector of each time window as an indicator of operational behavior deviation. It can identify abnormal signals in advance based solely on the statistical deviation of operational behavior when the physical parameters of the device have not exceeded the limits and the alarm has not been triggered, thus shifting the time of anomaly detection to the fault latency period.

[0025] Step 3: Based on the device identifier in the device behavior anomaly signal, read the building space boundary data and the installation coordinates of the controlled device from the storage unit. Taking the abnormal device as the starting node and the spatial connectivity relationship as the edge, search outward level by level for the set of associated devices that are spatially adjacent to the abnormal device. Assign propagation weights according to the search level to generate a spatial association propagation path map.

[0026] This embodiment takes the abnormal device as the starting node and searches for related devices level by level along three spatial relationships: interconnection within the same room, connection between adjacent rooms, and connection between upper and lower floors. The propagation weight is assigned decreasing from near to far, thus expanding the originally isolated single-point anomaly into a set of devices with spatial relationships. Maintenance personnel do not need to manually check the surrounding devices one by one according to the drawings.

[0027] Step 4: Based on the cumulative propagation weight value of each associated device in the spatial correlation propagation path diagram and their respective operational behavior deviation index, calculate the joint anomaly risk value of each associated device, and select associated devices whose joint anomaly risk value exceeds the preset risk threshold as the set of devices of interest.

[0028] In this embodiment, the cumulative propagation weight of each associated device is multiplied by its own deviation index and then divided by the number of incoming edges to obtain the joint anomaly risk value. This value reflects both the degree to which the device is affected by spatial propagation and the severity of its own anomaly, thus avoiding misjudgment or omission due to relying on only a single factor.

[0029] Step 5: Based on the three-dimensional spatial coordinates of each device in the set of devices of interest, and taking the three-dimensional coordinates of the abnormal device as the center, connect the three-dimensional coordinates of each device of interest level by level according to the propagation direction in the spatial association propagation path diagram to generate a set of spatial association paths.

[0030] In this embodiment, the three-dimensional coordinates of each device of interest are connected in ascending order of search level to generate a set of spatially related paths that alternate between horizontal and vertical. Maintenance personnel can directly view the complete path of an anomaly spreading outward from its source in three-dimensional space without having to manually refer to planar drawings to understand the spatial relationships.

[0031] Step Six: Based on the joint anomaly risk value and the time-series change trend of the operational behavior deviation index of each interested device in the spatial association path set, predict the operational behavior deviation index within the future time window to obtain the predicted value of the operational behavior deviation index.

[0032] In this embodiment, the joint anomaly risk value is used as a weighting coefficient to weight the time series of historical deviation indicators before predicting future values. Equipment with greater spatial propagation impact has a higher weight in the prediction, so that the prediction results simultaneously reflect the equipment's own change trend and the additional risks brought about by spatial propagation.

[0033] Step 7: When the predicted value of the deviation index of the operation behavior exceeds the preset intervention threshold, the path segment where the device of interest is located in the spatial associated path set is marked as an item requiring active intervention, and an active operation and maintenance decision list is generated and pushed to the management terminal.

[0034] In this embodiment, when the predicted value exceeds the intervention threshold, the device and its path segment are marked as items requiring active intervention. A decision list is generated along with the complete path segment from the abnormal device to the device and pushed to the management terminal. The operation and maintenance personnel can know the devices that need to be intervened in advance and the propagation path before the fault actually occurs, and check them step by step according to the path order.

[0035] Example 2 This embodiment describes the implementation steps of a property management method based on three-dimensional spatial visualization, including: Step 1: Collect historical operation event logs of each controlled device, construct an operation time series with device identifier as the primary key, and extract the operation frequency, operation interval variation coefficient and operation type entropy value within each time window as statistical feature vectors.

[0036] Step 1.1: Collect historical operation event logs for each controlled device; the operation event logs include device identifier, operation type, and operation timestamp.

[0037] The operation event logs collected in this embodiment include device identifier, operation type and operation timestamp. Subsequent anomaly identification can be completed using only the device's existing operation records, without the need for any additional sensors or monitoring devices.

[0038] Step 1.2: Using the device identifier as the primary key, sort the operation events belonging to the same controlled device in ascending order of time, resample them at equal intervals to a uniform time granularity, and construct the operation time sequence of each controlled device.

[0039] In this embodiment, the operation events are arranged in ascending order of time using the device identifier as the primary key and then resampled at equal intervals to a uniform time granularity. This transforms the originally irregular operation events into time-series data with equal intervals, which facilitates subsequent time window segmentation and feature extraction.

[0040] Step 1.3: Use a sliding window to segment the operation time series of each controlled device to obtain the operation event subsequence within each time window.

[0041] This embodiment uses a sliding window to segment the time series. There is an overlapping area between adjacent windows, which can smoothly capture the gradual changes in device operation behavior and avoid missing key variation signals due to the truncation caused by window boundary division.

[0042] Step 1.4: For each subsequence of operation events within a time window, the total number of times the controlled device is operated within that time window is calculated based on the operation timestamp as the operation frequency. The ratio of the standard deviation to the average value of the time interval between two adjacent operations is calculated as the operation interval variation coefficient. The frequency of occurrence of each operation type is calculated based on the operation type and the entropy value is calculated as the operation type entropy value.

[0043] This embodiment extracts operation frequency to reflect the overall operating load of the equipment, operation interval variation coefficient to reflect the stability of the operation rhythm, and operation type entropy value to reflect the diversity of operation behavior. These three features depict the operation behavior profile of the equipment from different dimensions. A single feature cannot fully reflect the abnormal operation behavior of the equipment.

[0044] Step 1.5: Construct a statistical feature vector based on the operation frequency, the operation interval variation coefficient, and the operation type entropy value.

[0045] This embodiment combines three features into a statistical feature vector, and aggregates the multi-dimensional operational behavior of the device within each time window into a computable vectorized representation, providing a unified data format for subsequent horizontal comparisons and outlier detection among similar devices.

[0046] Step 2: Compare the statistical feature vectors of each time window with the group feature distribution of the same type of controlled equipment, calculate the deviation index of operation behavior, and generate an abnormal equipment behavior signal when the deviation index of operation behavior exceeds the preset threshold.

[0047] Step 2.1: Using the controlled device type as the classification label, group all historical operation event logs of all controlled devices according to device type, and summarize the statistical feature vectors of each time window of all controlled devices in each group to form the group feature distribution of each type of controlled device.

[0048] This embodiment groups all controlled devices by device type as the classification label, and summarizes the statistical feature vectors of each time window of the same type of device to form a group feature distribution. This allows subsequent anomaly judgment to be based on the results of horizontal comparison between similar devices rather than absolute thresholds, which is suitable for scenarios where different types of devices have different operating characteristics.

[0049] Step 2.2: For any time window of any controlled device, input the statistical feature vector of that time window into the Isolation Forest algorithm. The Isolation Forest algorithm uses the population feature distribution as a reference benchmark, constructs multiple binary trees by randomly selecting features and split values, calculates the outlier score of the statistical feature vector relative to the population feature distribution, and uses the outlier score as the deviation index of the controlled device's operating behavior in that time window. The deviation index is used to characterize the degree to which the operating behavior of the controlled device deviates from the normal level of the same type of device group.

[0050] This embodiment uses the isolated forest algorithm to calculate the outlier score of the statistical feature vector relative to the population feature distribution as an indicator of the deviation of the operation behavior. This algorithm does not rely on the normal distribution assumption, can handle complex distribution patterns in high-dimensional feature space, and can achieve outlier identification without pre-labeling normal and abnormal samples.

[0051] Step 2.3: Compare the operation behavior deviation index with a preset threshold. When the operation behavior deviation index exceeds the preset threshold, it is determined that the controlled device has an abnormal operation behavior in the time window, and a device behavior abnormality signal containing the device identifier of the controlled device and the timestamp of the abnormality occurrence is generated.

[0052] In this embodiment, the deviation index of the operation behavior is compared with a preset threshold. When the deviation exceeds the threshold, it is determined that the device has an abnormal operation behavior and an abnormal signal is generated. This abnormal signal serves as the trigger condition for subsequent spatial propagation and deduction. Without waiting for the physical parameters of the device to exceed the limit or for an alarm to occur, the time point of abnormality discovery can be moved forward to the fault latency period.

[0053] Step 3: Based on the device identifier in the device behavior anomaly signal, read the building space boundary data and the installation coordinates of the controlled device from the storage unit. Taking the abnormal device as the starting node and the spatial connectivity relationship as the edge, search outward level by level for the set of associated devices that are spatially adjacent to the abnormal device. Assign propagation weights according to the search level to generate a spatial association propagation path diagram.

[0054] Step 3.1: Based on the device identifier in the device behavior anomaly signal, read the building space boundary data and the installation coordinates of the controlled device from the storage unit; the building space boundary data includes the hierarchical structure and outline of buildings, floors, and rooms; the spatial connectivity includes interconnection of devices in the same room, connection of adjacent rooms through wall holes or cable trays, and connection of upper and lower floors through pipe shafts or cable shafts.

[0055] The building space boundary data read from the storage unit in this embodiment includes the hierarchical structure and outline of buildings, floors, and rooms. The installation coordinates of the controlled equipment provide the precise positioning of the equipment in three-dimensional space. Based on this, three types of spatial connectivity relationships are defined: interconnection of the same room, connection of adjacent rooms, and connection between upper and lower floors. These cover the main physical proximity and connectivity methods between equipment in the building space, enabling subsequent searches to be expanded step by step according to the actual spatial topology.

[0056] Step 3.2: Based on the building space boundary data and the installation coordinates of the controlled equipment, starting with the abnormal equipment as the starting node and the spatial connectivity as the edge, search outwards level by level for related equipment that is spatially adjacent to the abnormal equipment until the preset search depth limit is reached, and obtain each related equipment and its corresponding search level.

[0057] This embodiment starts with the abnormal device as the starting node and searches outward step by step using three types of spatial connectivity as edges until the preset search depth limit is reached. This expands the originally isolated single-point anomaly into a set of related devices classified according to spatial proximity. Maintenance personnel no longer need to manually check the surrounding devices that may be affected by the anomaly according to the drawings. The system automatically completes the search and classification according to spatial relationships.

[0058] Step 3.2.1: Determine the target room where the target controlled device is located based on the installation coordinates of the target controlled device and the building space boundary data; In this embodiment, the target room where the device is located is accurately located in the hierarchical structure of building, floor and room based on the installation coordinates of the target controlled device and the building space boundary data, providing an accurate starting area for subsequent spatial adjacency search.

[0059] Step 3.2.2: Starting with the target room as the initial search area, and using the spatial connectivity as edges, gradually expand the search area outwards: Search for all controlled devices belonging to the same room as the target room, and designate them as the first associated devices, with the search level being the first level; Using the target room as the center, search for controlled devices in all rooms adjacent to the target room, which are considered as second associated devices, and the search level is the second level. The search is conducted on controlled devices located in rooms on adjacent floors above and below the target room, which are then used as the third associated devices, with the search level being the third level.

[0060] This embodiment takes the target room as the starting search area and expands outward step by step in three levels: the same room, adjacent rooms, and corresponding rooms on the upper and lower floors. The first level covers all devices in the same room, the second level covers horizontally adjacent rooms, and the third level covers vertically upper and lower floors. The three levels correspond to the tightness of spatial adjacency, so that the search range is naturally layered according to spatial distance.

[0061] Step 3.2.3: Summarize the first associated device, the second associated device, and the third associated device found to obtain each associated device and its corresponding search level.

[0062] In this embodiment, the results of the three search levels are summarized into each associated device and its corresponding search level. Each associated device is labeled with its level, which provides a hierarchical basis for assigning propagation weights in descending order of proximity.

[0063] Step 3.3: Based on each associated device and its corresponding search level, propagation weights are assigned in descending order from near to far. The lower the search level, the higher the propagation weight. Using the abnormal device and each associated device as nodes, the spatial connectivity relationship as edges, and the propagation weight as edge weights, a spatial association propagation path graph is generated.

[0064] In this embodiment, propagation weights are assigned in descending order of proximity, with lower search levels resulting in higher propagation weights. The first-level associated devices are closest to the anomalous devices and have the highest propagation weight, followed by the second-level devices, and the third-level devices have the lowest. This weight allocation method conforms to the actual physical law that the impact of anomalies decreases with increasing spatial distance. At the same time, a spatial association propagation path graph is generated using the anomalous devices and each associated device as nodes, spatial connectivity as edges, and propagation weights as edge weights. This integrates the originally scattered spatial location information and search levels into graph structure data that can be used for subsequent risk calculations.

[0065] Step 4: Based on the cumulative propagation weight value of each associated device in the spatial correlation propagation path diagram and their respective operational behavior deviation index, calculate the joint anomaly risk value of each associated device, and select associated devices whose joint anomaly risk value exceeds the preset risk threshold as the set of devices of interest.

[0066] Step 4.1: Traverse each associated device in the spatial association propagation path graph. For each associated device, sum the propagation weights of all incoming edges of that associated device to obtain the cumulative propagation weight value of that associated device.

[0067] In this embodiment, the cumulative propagation weight value is obtained by summing the propagation weights of all incoming edges of each associated device. This value reflects the total influence intensity of the abnormal device propagating to the associated device through multiple spatial connectivity paths, thus avoiding underestimating the actual propagation impact based solely on a single path.

[0068] Step 4.2: Obtain the deviation index of the operation behavior of each associated device.

[0069] This embodiment obtains the deviation index of the operation behavior corresponding to each associated device as the basis for assessing the severity of the device's own abnormality, so that the joint risk calculation can simultaneously consider the propagation effect from the abnormal device and the abnormal state of the device itself.

[0070] Step 4.3: For each associated device, the cumulative propagation weight value is used as the risk propagation factor of the associated device in the spatial propagation path graph, and the operational behavior deviation index is used as the anomalous factor of the associated device itself. The cumulative propagation weight value and the operational behavior deviation index are multiplied and then divided by the number of incoming edges of the associated device in the spatial propagation path graph to obtain the joint anomalous risk value of the associated device. The joint anomalous risk value is used to characterize the comprehensive risk level of the associated device after being affected by spatial propagation and its own anomalousness.

[0071] In this embodiment, the cumulative propagation weight value is multiplied by the deviation index of the operation behavior and then divided by the number of incoming edges to obtain the joint anomaly risk value. Multiplying by the cumulative propagation weight value reflects that the greater the propagation impact, the higher the risk. Multiplying by the deviation index reflects that the more severe the anomaly, the higher the risk. Dividing by the number of incoming edges avoids double calculation caused by multi-path convergence, so that the joint anomaly risk value comprehensively reflects the superposition effect of spatial propagation impact and the degree of anomaly.

[0072] Step 4.4: Compare the combined anomaly risk value of each associated device with the preset risk threshold, and select the associated devices whose combined anomaly risk value exceeds the preset risk threshold as the set of devices of interest.

[0073] This embodiment filters all associated devices by setting a preset risk threshold, retaining only associated devices whose combined abnormal risk value exceeds the threshold as the set of devices of interest, filtering out devices with lower risk, and reducing the amount of data to be processed in subsequent steps.

[0074] Step 5: Based on the three-dimensional spatial coordinates of each device in the set of devices of interest, and taking the three-dimensional coordinates of the abnormal device as the center, connect the three-dimensional coordinates of each device of interest level by level according to the propagation direction in the spatial association propagation path diagram to generate a set of spatial association paths.

[0075] Step 5.1: Obtain the three-dimensional spatial coordinates of each device of interest and the corresponding search level of each device of interest from the set of devices of interest.

[0076] This embodiment obtains the three-dimensional spatial coordinates of each interested device and its corresponding search level from the set of interested devices, providing two necessary data for subsequent path generation: spatial location and hierarchical relationship.

[0077] Step 5.2: Using the three-dimensional spatial coordinates of the abnormal device as the starting point of the path, determine all the devices of interest corresponding to the current level in ascending order of the search level corresponding to each device of interest.

[0078] In this embodiment, the three-dimensional spatial coordinates of the abnormal device are used as the starting point of the path. All devices of interest corresponding to the current level are determined in order of increasing search level, so that the path extends outward from the source. The starting point of each level is the device of the previous level and the ending point is the device of the current level, ensuring that the path direction is consistent with the direction of abnormal spatial propagation.

[0079] Step 5.3: For all the devices of interest corresponding to each search level, connect the three-dimensional spatial coordinates of every two adjacent devices of interest in the current level to generate the horizontal sub-path of the current level.

[0080] In this embodiment, all devices of interest within the same level are connected to the three-dimensional spatial coordinates of every two adjacent devices to generate a horizontal sub-path, thus linking devices in the same spatial adjacent level together and reflecting the path of anomaly propagation within the same level.

[0081] Step 5.4: Connect the two points with the closest spatial distance between the three-dimensional spatial coordinates of each interested device in the previous level and the three-dimensional spatial coordinates of each interested device in the current level to generate a vertical sub-path between adjacent levels.

[0082] In this embodiment, a vertical sub-path is generated by connecting the two points with the shortest spatial distance between the previous level and the current level. The connection point across levels is determined based on the principle of shortest distance, so that the path transition between different levels is the shortest and unreasonable connections that cross unrelated devices are avoided.

[0083] Step 5.5: Alternately connect the horizontal sub-paths and vertical sub-paths at all levels in order of increasing search level, and mark the direction of each sub-path in order of increasing search level to obtain a set of spatially associated paths.

[0084] In this embodiment, horizontal and vertical sub-paths at all levels are alternately connected in order of search level from low to high, and the direction identifiers of each sub-path are marked in the same order to obtain a complete set of spatially associated paths. Operation and maintenance personnel can directly obtain the complete path from the source of the anomaly outward and the direction of each segment of the path from this set.

[0085] Step 6: Based on the joint anomaly risk value and the time-series change trend of the operational behavior deviation index of each concerned device in the spatial association path set, predict the operational behavior deviation index within the future time window to obtain the predicted value of the operational behavior deviation index.

[0086] Step 6.1: Extract the deviation index of the operation behavior of each device of interest in multiple consecutive historical time windows from the set of spatially associated paths, arrange them in chronological order to form a time sequence of the deviation index of the operation behavior of each device of interest.

[0087] This embodiment extracts the deviation index of the operation behavior of each device of interest in a set of spatially associated paths within multiple consecutive historical time windows, arranges them in chronological order to form a time series, and provides historical data on the evolution of the deviation of each device of interest over time for subsequent prediction.

[0088] Step 6.2: Extract the joint anomaly risk value of each device of interest from the spatial association path set.

[0089] In this embodiment, the joint anomaly risk value of each device of interest is extracted from the set of spatially associated paths. This value has been comprehensively reflected in step four as the combined risk level of each device of interest due to the influence of spatial propagation and its own anomalies, and is used as the weighting coefficient for subsequent prediction.

[0090] Step 6.3: For each device of interest, the deviation index of each time window in the time series of the deviation index of the operation behavior is weighted using the joint anomaly risk value of the device of interest as a weighting coefficient to obtain the weighted deviation index time series; wherein, the higher the joint anomaly risk value, the greater the weighting influence on the weighted deviation index time series.

[0091] In this embodiment, the joint anomaly risk value is used as a weighting coefficient to weight the deviation index of each historical time window. The higher the joint anomaly risk value of the equipment, the higher the weight of its historical deviation data in the prediction, so that the historical change trend of the equipment with greater spatial propagation influence has a greater impact on the prediction result.

[0092] Step 6.4: Calculate the weighted rate of change based on the difference between adjacent time windows in the time series of the weighted deviation index.

[0093] In this embodiment, the weighted rate of change is calculated based on the difference between adjacent time windows after weighting. This rate reflects the overall direction and magnitude of the change in the equipment deviation index after weighting. When the rate of change is positive, the deviation shows an upward trend, and when it is negative, it shows a downward trend.

[0094] Step 6.5: The sum of the weighted deviation index value of the current time window and the weighted rate of change is used as the predicted value of the deviation index of the operating behavior of the device of interest in the future time window.

[0095] In this embodiment, the weighted deviation index value and the weighted rate of change of the current time window are used as the predicted value. The current actual level is superimposed with the historical trend of change. The predicted value reflects both the current deviation status of the equipment and the recent trend of change. If the predicted value shows a continuous upward trend, it indicates that the risk of the equipment is intensifying.

[0096] Step 7: When the predicted value of the deviation index of the operation behavior exceeds the preset intervention threshold, the path segment where the device of interest is located in the spatial associated path set is marked as an item requiring active intervention, and an active operation and maintenance decision list is generated and pushed to the management terminal.

[0097] Step 7.1: Compare the predicted values ​​of the operational behavior deviation index of each monitored device with the preset intervention threshold, and mark the monitored devices whose predicted values ​​of the operational behavior deviation index exceed the preset intervention threshold as devices that need active intervention.

[0098] In this embodiment, the predicted deviation index of each device of interest is compared with a preset intervention threshold. Devices whose predicted values ​​exceed the threshold are marked as devices that require active intervention, thus locking in the target devices that need to be dealt with in advance before the actual occurrence of the fault.

[0099] Step 7.2: Traverse each path segment in the spatially associated path set and mark the path segment containing the device requiring active intervention as the path segment requiring intervention.

[0100] In this embodiment, each path segment in the spatially associated path set is traversed, and the path segment containing the device that needs active intervention is marked as the path segment that needs intervention. This ensures that the entire path segment containing the device that needs active intervention is marked, rather than just the device itself, so that the operation and maintenance personnel can know the complete spatially associated range where the device is located.

[0101] Step 7.3: Starting from the abnormal device, along the propagation direction in the set of spatially associated paths, record all path segments from the abnormal device to each device requiring active intervention, as well as all devices of interest on each path segment, to generate an active operation and maintenance decision list.

[0102] This embodiment starts with the abnormal device and records all path segments from the abnormal device to each device requiring active intervention along the propagation direction, as well as all devices of interest on each path segment, generating a decision list containing the complete propagation path. Operation and maintenance personnel can directly know all intermediate nodes from the source of the anomaly to the device requiring intervention without having to trace back level by level themselves.

[0103] Step 7.4: Push the proactive maintenance decision list to the management terminal.

[0104] This embodiment pushes the proactive maintenance decision list to the management terminal, enabling maintenance personnel to obtain the list of devices that need proactive intervention and the complete propagation path from the source of the anomaly to the device before the actual occurrence of the fault, and to investigate and intervene step by step according to the path sequence.

[0105] Example 3 like Figure 2 As shown in the figure, this embodiment introduces a property management system based on three-dimensional spatial visualization, including: The operation log collection module is used to collect historical operation event logs of each controlled device, construct an operation time series with the device identifier as the primary key, and extract the operation frequency, operation interval variation coefficient and operation type entropy value within each time window as statistical feature vectors. The behavior deviation calculation module is used to compare the statistical feature vectors of each time window with the group feature distribution of the same type of controlled equipment, calculate the operation behavior deviation index, and generate an abnormal equipment behavior signal when the operation behavior deviation index exceeds the preset threshold. The spatial association propagation path generation module is used to read building space boundary data and controlled equipment installation coordinates from the storage unit based on the device identifier in the device behavior abnormal signal. Taking the abnormal device as the starting node and the spatial connectivity relationship as the edge, it searches outward level by level for a set of associated devices that are spatially adjacent to the abnormal device, and assigns propagation weights according to the search level to generate a spatial association propagation path diagram. The joint anomaly risk assessment module is used to calculate the joint anomaly risk value of each associated device based on the cumulative propagation weight value and the deviation index of each device's operation behavior in the spatial correlation propagation path diagram, and to screen the associated devices whose joint anomaly risk value exceeds a preset risk threshold as a set of devices of concern. The spatial association path generation module is used to generate a spatial association path set by connecting the three-dimensional spatial coordinates of each device of interest in the set of devices of interest, with the three-dimensional coordinates of the abnormal device as the center, and following the propagation direction in the spatial association propagation path diagram. The deviation prediction module is used to predict the deviation index of operation behavior within a future time window based on the joint abnormal risk value of each concerned device in the spatial association path set and the time-series change trend of the deviation index of operation behavior, and to obtain the predicted value of the deviation index of operation behavior. The proactive operation and maintenance decision module is used to mark the path segment where the device of interest is located in the spatially associated path set as an item requiring proactive intervention when the predicted value of the deviation index of the operation behavior exceeds the preset intervention threshold, and to generate a proactive operation and maintenance decision list and push it to the management terminal.

[0106] The specific steps for each of the above modules are as described in Embodiment 1 or 2, and will not be repeated here.

[0107] 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 containing computer-usable program code. It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A property management method based on three-dimensional spatial visualization, characterized in that, include: Collect historical operation event logs of each controlled device, construct an operation time series with device identifier as the primary key, and extract the operation frequency, operation interval variation coefficient and operation type entropy value within each time window as statistical feature vectors. The statistical feature vectors of each time window are compared with the group feature distribution of the same type of controlled equipment to calculate the deviation index of operation behavior. When the deviation index of operation behavior exceeds the preset threshold, an abnormal equipment behavior signal is generated. Based on the device identifier in the abnormal device behavior signal, the building space boundary data and the installation coordinates of the controlled device are read from the storage unit. Taking the abnormal device as the starting node and the spatial connectivity relationship as the edge, the set of associated devices that are spatially adjacent to the abnormal device is searched outward level by level. The propagation weight is assigned according to the search level to generate a spatial association propagation path diagram. Based on the cumulative propagation weight value of each associated device in the spatial correlation propagation path diagram and the deviation index of their respective operational behaviors, the joint anomaly risk value of each associated device is calculated, and associated devices whose joint anomaly risk value exceeds a preset risk threshold are selected as the set of devices of interest. Based on the three-dimensional spatial coordinates of each device in the set of devices of interest, and taking the three-dimensional coordinates of the abnormal device as the center, the three-dimensional coordinates of each device of interest are connected step by step according to the propagation direction in the spatial association propagation path diagram to generate a set of spatial association paths. Based on the joint anomaly risk value and the time-series change trend of the operational behavior deviation index of each concerned device in the spatial association path set, the operational behavior deviation index within the future time window is predicted to obtain the predicted value of the operational behavior deviation index. When the predicted value of the deviation index of the operation behavior exceeds the preset intervention threshold, the path segment where the device of interest is located in the spatial associated path set is marked as an item requiring active intervention, and an active operation and maintenance decision list is generated and pushed to the management terminal.

2. The property management method based on three-dimensional spatial visualization as described in claim 1, characterized in that, Historical operation event logs of each controlled device are collected. An operation time series is constructed using the device identifier as the primary key. The operation frequency, operation interval coefficient of variation, and operation type entropy value within each time window are extracted as statistical feature vectors, including: Collect historical operation event logs for each controlled device; the operation event logs include device identifier, operation type, and operation timestamp. Using the device identifier as the primary key, operation events belonging to the same controlled device are arranged in ascending order of time and resampled at equal intervals to a uniform time granularity to construct the operation time sequence of each controlled device. The operation time series of each controlled device is divided by using a sliding window to obtain the operation event subsequence within each time window; For each subsequence of operation events within a time window, the total number of times the controlled device is operated within that time window is counted based on the operation timestamp as the operation frequency. The ratio of the standard deviation to the average value of the time interval between two adjacent operations is calculated as the operation interval variation coefficient. The frequency of occurrence of each operation type is counted based on the operation type, and the entropy value is calculated as the operation type entropy value. A statistical feature vector is constructed based on the operation frequency, the operation interval variation coefficient, and the operation type entropy value.

3. The property management method based on three-dimensional spatial visualization as described in claim 2, characterized in that, The statistical feature vectors of each time window are compared with the group feature distribution of similar controlled equipment to calculate the operational behavior deviation index. When the operational behavior deviation index exceeds a preset threshold, an abnormal equipment behavior signal is generated, including: Using the type of controlled equipment as the classification label, the historical operation event logs of all controlled equipment are grouped by equipment type. The statistical feature vectors of each time window of all controlled equipment in each group are summarized to form the group feature distribution of each type of controlled equipment. For any controlled device at any given time window, the statistical feature vector of that time window is input into the Isolation Forest algorithm. The Isolation Forest algorithm uses the population feature distribution as a reference benchmark, constructs multiple binary trees by randomly selecting features and split values, calculates the outlier score of the statistical feature vector relative to the population feature distribution, and uses the outlier score as the deviation index of the controlled device's operational behavior in that time window. The operational behavior deviation index is used to characterize the degree to which the operational behavior of the controlled device deviates from the normal level of a group of similar devices. The deviation index of the operation behavior is compared with a preset threshold. When the deviation index of the operation behavior exceeds the preset threshold, it is determined that the controlled device has an abnormal operation behavior in the time window, and a device behavior abnormality signal containing the device identifier of the controlled device and the timestamp of the abnormality is generated.

4. The property management method based on three-dimensional spatial visualization as described in claim 3, characterized in that, Based on the device identifier in the abnormal device behavior signal, the building space boundary data and the installation coordinates of the controlled equipment are read from the storage unit. Taking the abnormal device as the starting node and spatial connectivity as the edges, a set of associated devices that are spatially adjacent to the abnormal device is searched outward level by level. Propagation weights are assigned according to the search level to generate a spatial association propagation path graph, including: Based on the device identifier in the abnormal device behavior signal, the building space boundary data and the installation coordinates of the controlled device are read from the storage unit; the building space boundary data includes the hierarchical structure and outline of buildings, floors, and rooms; the spatial connectivity includes interconnection of devices in the same room, connection of adjacent rooms through wall holes or cable trays, and connection of upper and lower floors through pipe shafts or cable shafts. Based on the building space boundary data and the installation coordinates of the controlled equipment, the abnormal equipment is used as the starting node and the spatial connectivity is used as the edge. The search is carried outward level by level to find the associated equipment that is spatially adjacent to the abnormal equipment until the preset search depth limit is reached, so as to obtain each associated equipment and its corresponding search level. Based on each associated device and its corresponding search level, propagation weights are assigned in descending order from near to far, with higher propagation weights for lower search levels. A spatial association propagation path graph is generated using the abnormal device and each associated device as nodes, the spatial connectivity as edges, and the propagation weights as edge weights.

5. The property management method based on three-dimensional spatial visualization as described in claim 4, characterized in that, Based on the building space boundary data and the installation coordinates of the controlled equipment, taking the abnormal equipment as the starting node and the spatial connectivity as the edges, the system searches outwards level by level for associated equipment that is spatially adjacent to the abnormal equipment until a preset search depth limit is reached. This yields each associated equipment and its corresponding search level, including: Based on the installation coordinates of the target controlled equipment and the building space boundary data, the target room where the target controlled equipment is located is determined; Starting with the target room as the initial search area, and using the spatial connectivity as edges, the search area is expanded outwards level by level: Search for all controlled devices belonging to the same room as the target room, and designate them as the first associated devices, with the search level being the first level; Centered on the target room, search for controlled devices in all rooms adjacent to the target room, which are considered as second associated devices, and the search level is the second level. Search for controlled devices in rooms located on adjacent floors above and below the target room, and use them as the third associated devices, with the search level being the third level. The first associated device, the second associated device, and the third associated device found in the search are summarized to obtain each associated device and its corresponding search level.

6. The property management method based on three-dimensional spatial visualization as described in claim 5, characterized in that, Based on the cumulative propagation weight values ​​and respective operational behavior deviation indices of each associated device in the spatial correlation propagation path diagram, the joint anomaly risk value of each associated device is calculated. Associated devices whose joint anomaly risk values ​​exceed a preset risk threshold are selected as a set of devices of interest, including: Traverse each associated device in the spatial association propagation path graph. For each associated device, accumulate the propagation weights of all incoming edges of that associated device to obtain the cumulative propagation weight value of that associated device. Obtain the deviation index of the operation behavior of each associated device; For each associated device, the cumulative propagation weight value is used as the risk propagation factor of the associated device in the spatial propagation path graph, and the operational behavior deviation index is used as the anomaly factor of the associated device itself. The cumulative propagation weight value and the operational behavior deviation index are multiplied and then divided by the number of incoming edges of the associated device in the spatial propagation path graph to obtain the joint anomaly risk value of the associated device. The joint anomaly risk value is used to characterize the comprehensive risk level of the associated device after being affected by spatial propagation and its own anomalies. The combined anomaly risk value of each associated device is compared with a preset risk threshold, and associated devices whose combined anomaly risk value exceeds the preset risk threshold are selected as the set of devices of interest.

7. The property management method based on three-dimensional spatial visualization as described in claim 6, characterized in that, Based on the three-dimensional spatial coordinates of each device in the set of devices of interest, and taking the three-dimensional coordinates of the abnormal device as the center, the three-dimensional coordinates of each device of interest are connected step by step according to the propagation direction in the spatial association propagation path diagram to generate a set of spatial association paths, including: Obtain the three-dimensional spatial coordinates of each device of interest and the search level corresponding to each device of interest from the set of devices of interest; Using the three-dimensional spatial coordinates of the abnormal device as the starting point of the path, all devices of interest corresponding to the current level are determined step by step in order from low to high search level for each device of interest. For all the devices of interest corresponding to each search level, connect the three-dimensional spatial coordinates of every two adjacent devices of interest in the current level to generate the horizontal sub-path of the current level; Connect the two points with the closest spatial distance between the three-dimensional spatial coordinates of each interested device in the previous level and the three-dimensional spatial coordinates of each interested device in the current level to generate a vertical sub-path between adjacent levels; By alternately connecting horizontal and vertical sub-paths at all levels in order of increasing search level, and marking the direction of each sub-path in order of increasing search level, a set of spatially associated paths is obtained.

8. The property management method based on three-dimensional spatial visualization as described in claim 7, characterized in that, Based on the joint anomaly risk value and the time-series change trend of the operational behavior deviation index of each monitored device in the spatial association path set, the operational behavior deviation index within the future time window is predicted to obtain the predicted value of the operational behavior deviation index, including: The deviation index of the operation behavior of each device of interest in the spatial association path set is extracted in multiple consecutive historical time windows and arranged in chronological order to form a time sequence of the operation behavior deviation index of each device of interest. Extract the joint anomaly risk value of each device of interest from the spatially associated path set; For each device of interest, the joint anomaly risk value of that device of interest is used as a weighting coefficient to weight the deviation index of each time window in the time series of the operational behavior deviation index, resulting in a weighted deviation index time series; wherein, the higher the joint anomaly risk value, the greater the weighting impact on the weighted deviation index time series. The weighted rate of change is calculated based on the difference between adjacent time windows in the time series of the weighted deviation index. The weighted deviation index value of the current time window and the weighted rate of change are used as the predicted value of the deviation index of the operating behavior of the device under interest in the future time window.

9. The property management method based on three-dimensional spatial visualization as described in claim 8, characterized in that, When the predicted value of the operational behavior deviation index exceeds the preset intervention threshold, the path segment containing the device of interest in the spatially associated path set is marked as an item requiring active intervention, and an active operation and maintenance decision list is generated and pushed to the management terminal, including: The predicted values ​​of the operational behavior deviation index of each monitored device are compared with the preset intervention threshold. Devices whose predicted values ​​of the operational behavior deviation index exceed the preset intervention threshold are marked as devices that require active intervention. Traverse each path segment in the spatially associated path set, and mark the path segment containing the device requiring active intervention as the path segment requiring intervention; Starting from the abnormal device, along the propagation direction in the set of spatially associated paths, record all path segments from the abnormal device to each device requiring active intervention and all devices of interest on each path segment, and generate an active operation and maintenance decision list. The proactive maintenance decision list is pushed to the management terminal.

10. A property management system based on three-dimensional spatial visualization, characterized in that, include: The operation log collection module is used to collect historical operation event logs of each controlled device, construct an operation time series with the device identifier as the primary key, and extract the operation frequency, operation interval variation coefficient and operation type entropy value within each time window as statistical feature vectors. The behavior deviation calculation module is used to compare the statistical feature vectors of each time window with the group feature distribution of the same type of controlled equipment, calculate the operation behavior deviation index, and generate an abnormal equipment behavior signal when the operation behavior deviation index exceeds the preset threshold. The spatial association propagation path generation module is used to read building space boundary data and controlled equipment installation coordinates from the storage unit based on the device identifier in the device behavior abnormal signal, and search outwards level by level for a set of associated devices that are spatially adjacent to the abnormal device, with the abnormal device as the starting node and the spatial connectivity relationship as the edge, and assign propagation weights according to the search level to generate a spatial association propagation path diagram. The joint anomaly risk assessment module is used to calculate the joint anomaly risk value of each associated device based on the cumulative propagation weight value and the deviation index of each device's operation behavior in the spatial correlation propagation path diagram, and to screen the associated devices whose joint anomaly risk value exceeds a preset risk threshold as a set of devices of concern. The spatial association path generation module is used to generate a spatial association path set by connecting the three-dimensional spatial coordinates of each device of interest in the set of devices of interest, with the three-dimensional coordinates of the abnormal device as the center, and following the propagation direction in the spatial association propagation path diagram. The deviation prediction module is used to predict the deviation index of operation behavior within a future time window based on the joint abnormal risk value of each concerned device in the spatial association path set and the time-series change trend of the deviation index of operation behavior, and to obtain the predicted value of the deviation index of operation behavior. The proactive operation and maintenance decision module is used to mark the path segment where the device of interest is located in the spatially associated path set as an item requiring proactive intervention when the predicted value of the deviation index of the operation behavior exceeds the preset intervention threshold, and to generate a proactive operation and maintenance decision list and push it to the management terminal.

Images