A multivariable logic monitoring and fault diagnosis system for an active operation and maintenance control loop of a building structure

Abstract

The application discloses a kind of multivariate logic monitoring and fault diagnosis systems of building structure active operation and maintenance control loop, it is related to intelligent building structure operation and maintenance technical field, including constructing unified multivariate input data stream;Dynamic anomaly propagation path between state variables is identified;Self-adapting control instruction is generated;Trigger hierarchical early warning signal before the fault that has not occurred diagnosis.This application realizes the fundamental change of building structure operation and maintenance from passive response to active intervention, significantly improves the logical completeness and mechanism explainability of fault identification;Realize risk-driven self-adapting instruction generation and real-time feedback verification;The safety, reliability and intelligent operation and maintenance level of building structure under complex service environment are improved, and key technical support is provided for the whole life cycle health management of wisdom city infrastructure.

Description

Technical Field

[0001] This invention relates to the field of intelligent building structure operation and maintenance technology, and in particular to a multivariable logic monitoring and fault diagnosis system for an active operation and maintenance control loop of a building structure. Background Technology

[0002] With the accelerating pace of urbanization and the continuous expansion of infrastructure, the demands for the safety, reliability, and intelligent operation and maintenance of building structures are becoming increasingly prominent. Society is paying increasing attention to the entire lifecycle management of buildings, and building structures are evolving towards proactive sensing, intelligent early warning, and autonomous decision-making. At the same time, frequent extreme weather events, accelerated material aging, and complex load environments pose unprecedented safety challenges to building structures, necessitating an intelligent monitoring device capable of multi-source information fusion, real-time status assessment, and early fault identification to support the fundamental shift in building structure operation and maintenance from passive response to proactive management.

[0003] However, existing building structure monitoring technologies generally suffer from problems such as isolated variables, fragmented logic, and delayed diagnosis. Traditional systems often rely on single sensor data or simple threshold judgments, making it difficult to effectively capture structural behavior characteristics under the coupling of multiple physical fields. Simultaneously, the lack of a closed-loop linkage mechanism between the control loop and the monitoring system leads to low fault identification accuracy, slow response speed, and an inability to proactively intervene in potential risks. Furthermore, existing diagnostic methods often neglect the dynamic logical relationships between variables, making it difficult to adapt to nonlinear degradation processes under complex operating conditions, severely restricting the level of intelligent operation and maintenance of building structures. Summary of the Invention

[0004] In view of the problems existing in the multivariable logic monitoring and fault diagnosis system of the active operation and maintenance control loop of building structure, this invention is proposed.

[0005] Therefore, the problem that this invention aims to solve is that existing building structure monitoring systems suffer from isolated dependent variables, fragmented logic, delayed diagnosis, and lack of closed-loop linkage with control execution, making it difficult to achieve early identification, accurate location, and proactive intervention of structural faults under the coupling of multiple physical fields.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, embodiments of the present invention provide a multivariable logic monitoring and fault diagnosis system for an active operation and maintenance control loop of a building structure, which includes a multi-source heterogeneous sensor fusion module for collecting state variables under the action of multiple physical fields, performing multi-source data collaborative preprocessing on raw data of different sensor types, and constructing a unified multivariable input data stream.

[0008] The multivariate logical association modeling module is used to construct logical relationships that reflect the dynamic evolution of structural behavior based on multivariate input data streams. By introducing a dynamic logical coupling reasoning mechanism, it identifies dynamic anomaly propagation paths between state variables.

[0009] The closed-loop active operation and maintenance control loop module is used to map the abnormal risk mapping mechanism output by the multivariate logical correlation modeling module to the structural operation and maintenance execution unit to generate adaptive control instructions;

[0010] The intelligent fault diagnosis early warning module combines multivariate degradation correlation features and adopts an incremental learning online model update mechanism to locate and predict the trend of diagnosed faults, triggering graded early warning signals before the diagnosed fault occurs.

[0011] As a preferred embodiment of the multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structure described in this invention, the multi-source heterogeneous sensing fusion module includes a multi-physics field state perception submodule, a multi-source temporal alignment submodule, a spatial topology mapping submodule, and a feature normalization fusion submodule.

[0012] The multi-physics field state perception submodule is used to deploy a multi-type sensor network on the building structure to collect raw data of multi-dimensional state variables;

[0013] The multi-source timing alignment submodule is used to perform timing collaborative calibration on multi-variable input data streams and align various state variables under a unified time reference.

[0014] The spatial topology mapping submodule is used to map the state variables collected by each sensor to the corresponding physical location nodes based on the structural sensing spatial registration benchmark, and to establish a multivariate distribution framework.

[0015] The feature normalization fusion submodule is used to perform structured weighted normalization fusion on the original data of multidimensional state variables to generate a unified multivariate input data stream.

[0016] As a preferred embodiment of the multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structure described in this invention, the multivariable logic association modeling module includes a multivariable temporal feature extraction submodule, a dynamic coupling graph construction submodule, a logic rule reasoning submodule, and an anomaly propagation path identification submodule;

[0017] The multivariate temporal feature extraction submodule is used to perform dynamic slicing encoding of temporal features on a unified multivariate input data stream, extract the temporal morphological features of each state variable in the time dimension, and form a temporal feature vector set.

[0018] The dynamic coupling graph construction submodule is used to obtain a multi-dimensional dynamic coupling metric between different state variables based on a time-series feature vector set, construct a multi-variable dynamic coupling process that evolves over time, and express a multi-field interaction mechanism for structural behavior.

[0019] The logical rule reasoning submodule is used to introduce a dynamic logical coupling reasoning mechanism on the basis of the dynamic coupling process. Through the hybrid logical reasoning framework, the existing structural behavioral logical topological relationship is symbolically modeled to generate a structural behavioral logical relationship network.

[0020] The anomaly propagation path identification submodule is used to traverse the activation paths in the structural behavioral logic relationship network, detect behavioral chains that deviate from the expected logic, track the anomaly propagation timing characteristics of abnormal signals between variable nodes, and output the dynamic anomaly propagation path.

[0021] As a preferred embodiment of the multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structure described in this invention, the closed-loop active operation and maintenance control loop module includes an abnormal risk semantic parsing submodule, an execution unit capability matching submodule, an adaptive strategy generation submodule, and a closed-loop instruction issuance feedback verification submodule.

[0022] The anomaly risk semantic parsing submodule is used to receive the dynamic anomaly propagation path output by the multivariate logical association modeling module, perform structured parsing, and generate anomaly risk description vectors.

[0023] The execution unit capability matching submodule is used to construct an execution unit capability representation based on the comprehensive capability characteristics of various operation and maintenance execution units deployed in the building structure, and to perform multi-dimensional matching between the abnormal risk description vector and the execution unit capability representation to screen out a set of candidate execution units that can participate in regulation.

[0024] The adaptive strategy generation submodule is used to dynamically combine or generate control strategy rules that are adapted to the current abnormal risk scenario based on the control strategy knowledge base and the current environmental boundary conditions, and convert them into parameterized instruction templates.

[0025] The closed-loop instruction issuance feedback verification submodule is used to encapsulate the parameterized instruction template into a structured control instruction and issue it to the corresponding operation and maintenance execution unit in real time. It collects the response status and structured feedback signal of the execution unit, verifies whether the control effect meets the expected logic, and triggers the strategy replanning process when the deviation exceeds the threshold.

[0026] As a preferred embodiment of the multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structure described in this invention, the fault intelligent diagnosis early warning module includes a multivariable degradation feature fusion submodule, an incremental degradation mode learning submodule, a fault location trend inference submodule, and a graded early warning triggering submodule.

[0027] The multivariate degradation feature fusion submodule is used to receive dynamic anomaly propagation paths, combine structural service environment parameters, extract and fuse multi-dimensional correlation features to form a unified multivariate degradation correlation feature set.

[0028] The incremental degradation pattern learning submodule is used to continuously update the fault pattern library based on a multivariate degradation association feature set using an incremental learning mechanism, and to dynamically expand and optimize the fault representation space by identifying degradation trajectory clusters online.

[0029] The fault location trend inference submodule is used to update the matching results in the fault mode library based on the multivariate degradation association feature set, determine the fault development stage, and combine the temporal extrapolation physical constraint fusion mechanism to make multi-step predictions on the future evolution path of fault indicators and generate a fault development trend curve.

[0030] The graded early warning triggering submodule is used to compare the fault development trend curve with the deviation exceeding the threshold. When the predicted degree of deterioration will cross any early warning threshold within the time window, the corresponding level of early warning signal will be automatically activated, the early warning level will be divided, and an early warning command will be output.

[0031] As a preferred embodiment of the multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structure described in this invention, the structure operation and maintenance execution unit is used to receive parameterized instruction templates issued by the closed-loop active operation and maintenance control loop module and intervene in the mechanical state of the building structure.

[0032] During the intervention process, the execution structure coupling feedback data is collected and sent back to the closed-loop instruction issuance feedback verification submodule to complete the execution feedback.

[0033] As a preferred embodiment of the multivariate logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structure described in this invention, the fault location trend inference submodule includes updating the matching results in the fault mode library based on the multivariate degradation association feature set, determining the fault development stage, and combining the temporal extrapolation physical constraint fusion mechanism to perform multi-step prediction of the future evolution path of fault indicators.

[0034] Fault indicators at the prediction time status Calculate using the following formula:

[0035]

[0036] in, This indicates the state value of the fault indicator at the predicted time. This represents the multivariate degenerate association feature vector at the current moment. This represents the temporal extrapolation mapping function that incorporates physical constraints. This represents the set of updatable parameters for the extrapolation mapping function. This represents the residual perturbation term that satisfies the physical boundary conditions.

[0037] Secondly, embodiments of the present invention provide a multivariable logic monitoring and fault diagnosis method for an active operation and maintenance control loop of a building structure, which includes: collecting state variables under the action of multiple physical fields, performing multi-source data collaborative preprocessing on raw data of different sensor types, and constructing a unified multivariable input data stream;

[0038] Based on multivariate input data streams, logical relationships reflecting the dynamic evolution of structural behavior are constructed. By introducing a dynamic logical coupling reasoning mechanism, dynamic anomaly propagation paths between state variables are identified.

[0039] The abnormal risk mapping mechanism output by the multivariate logical correlation modeling module is mapped to the structural operation and maintenance execution unit to generate adaptive control instructions;

[0040] By combining multivariate degradation association features, an incremental learning online model update mechanism is adopted to predict the location trend of diagnosed faults and trigger graded early warning signals before the diagnosed faults occur.

[0041] Thirdly, embodiments of the present invention provide a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement any step of the above-described multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structures.

[0042] Fourthly, embodiments of the present invention provide a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the above-described multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structures.

[0043] The beneficial effects of this invention are as follows: By constructing a closed-loop intelligent system integrating perception, modeling, decision-making, execution, and early warning, this invention achieves a fundamental shift in building structure operation and maintenance from passive response to proactive intervention. The multi-source heterogeneous sensor fusion module effectively eliminates variable isolation and information fragmentation, ensuring the spatiotemporal consistency and fusionability of multi-physics state data; the multi-variable logical association modeling module, through a dynamic logical coupling reasoning mechanism, accurately characterizes the nonlinear coupling relationships and anomaly propagation paths between variables, significantly improving the logical completeness and mechanism interpretability of fault identification; the closed-loop proactive operation and maintenance control loop module breaks down the barriers between monitoring and diagnosis and execution control, realizing risk-driven adaptive command generation and real-time feedback verification; the fault intelligent diagnosis early warning module, combining incremental learning and physical constraint extrapolation, can complete fault trend prediction and graded early warning before significant deterioration of structural performance. Overall, this invention significantly improves the safety, reliability, and intelligent operation and maintenance level of building structures in complex service environments, providing key technical support for the full life-cycle health management of smart city infrastructure. Attached Figure Description

[0044] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0045] Figure 1 This is a schematic diagram of a multivariable logic monitoring and fault diagnosis system for an active operation and maintenance control loop of a building structure, provided as an embodiment of the present invention.

[0046] Figure 2 The flowchart illustrates a method for a multivariable logic monitoring and fault diagnosis system for an active operation and maintenance control loop of a building structure, as provided in an embodiment of the present invention. Detailed Implementation

[0047] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0048] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0049] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0050] This invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of this invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not adhering to the usual scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of this invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.

[0051] Furthermore, in the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used solely for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In addition, the terms "first," "second," or "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0052] Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0053] Example

[0054] Reference Figure 1 and Figure 2 This is the first embodiment of the present invention, which provides a multivariable logic monitoring and fault diagnosis system for an active operation and maintenance control loop of a building structure, comprising:

[0055] S1: Multi-source heterogeneous sensing fusion module, used to collect state variables under the action of multiple physical fields, perform multi-source data collaborative preprocessing on raw data from different sensor types, and construct a unified multivariable input data stream.

[0056] The multi-source heterogeneous sensing fusion module includes a multi-physics field state perception submodule, a multi-source temporal alignment submodule, a spatial topology mapping submodule, and a feature normalization fusion submodule.

[0057] The multi-physics state sensing submodule is used to deploy a multi-type sensor network in building structures to collect raw data of multi-dimensional state variables;

[0058] The multi-source timing alignment submodule is used to perform timing-coordinated calibration on multivariable input data streams, aligning various state variables under a unified time reference.

[0059] The spatial topology mapping submodule is used to map the state variables collected by each sensor to the corresponding physical location nodes based on the structural sensing spatial registration benchmark, and to establish a multivariate distribution framework.

[0060] The feature normalization fusion submodule is used to perform structured weighted normalization fusion on the original data of multidimensional state variables to generate a unified multivariate input data stream.

[0061] Furthermore, the multi-source heterogeneous sensing fusion module systematically addresses the heterogeneity of multi-sensor data across time, space, and semantics through four collaborative sub-modules. First, the multi-physics state perception sub-module deploys sensors of various types, including strain, acceleration, temperature, humidity, and displacement, on key stress areas, environmentally sensitive parts, and vulnerable components of the building structure. This comprehensively collects and reflects the original state variables of the structure under the coupled effects of force, heat, humidity, and vibration. Second, the multi-source timing alignment sub-module performs high-precision timestamp synchronization and dynamic interpolation compensation to address time discrepancies caused by sampling frequency, triggering mechanisms, or communication delays among different sensors. This ensures that all state variables are aligned on a unified timeline, avoiding logical misjudgments caused by timing misalignments. Next, the spatial topology mapping... Based on the geometric topology of the building structure and the actual installation location of the sensors, the shooting module constructs a structure-sensor spatial registration benchmark, accurately mapping each dimension of sensor data to its corresponding physical node, forming a spatially consistent multivariate distribution framework. Finally, the feature normalization and fusion submodule performs dimensional unification, amplitude normalization, and noise suppression on the spatiotemporally aligned data, and combines the influence weights of each physical field on structural safety to implement structured weighted fusion, ultimately outputting a standardized, highly consistent unified multivariate input data stream, providing a high-quality, associative input foundation for subsequent logical modeling and fault diagnosis.

[0062] Furthermore, the multi-source heterogeneous sensing fusion module constructs a complete data processing chain from raw sensing to standardized input through four closely connected and progressively advanced sub-modules. First, the multi-physics state sensing sub-module deploys a heterogeneous sensor network at key parts of the building structure, such as beam-column joints, seismic isolation layers, curtain wall connections, and temperature and humidity sensitive areas, simultaneously acquiring multi-dimensional raw signals such as strain, acceleration, temperature, humidity, and displacement, comprehensively covering the structural response characteristics under the combined action of multiple physical fields such as mechanical loads, environmental erosion, and vibration excitation. Subsequently, the multi-source time-series alignment sub-module performs fine-grained time alignment on these raw data streams from different hardware platforms with different sampling rates and communication protocols. Through high-precision timestamp calibration and dynamic interpolation compensation technology, it eliminates sampling offsets caused by device asynchrony or network latency, ensuring that all variables strictly correspond to the same physical moment. Based on this, the spatial topology mapping sub-module relies on the building structure... The three-dimensional geometric information of the structure and the sensor layout drawings are used to establish a structure-sensor spatial registration benchmark. Each set of time-aligned sensor data is precisely bound to its actual physical location in the structure, thereby constructing a multivariate distribution framework with clear spatial topological relationships. Finally, the feature normalization fusion submodule performs unified dimension conversion, amplitude normalization and noise filtering on the spatiotemporally coordinated data. Based on the contribution of each physical field (such as stress field dominating safety and temperature and humidity field dominating durability) to the structural performance degradation, differentiated weights are assigned, and structured weighted fusion is implemented to output a standardized multivariate input data stream with consistent dimensions, unified semantics and controllable noise. This provides high-fidelity and strongly correlated basic data support for subsequent multivariate logical modeling and fault diagnosis.

[0063] S2: Multivariable logical association modeling module, used to construct logical relationships that reflect the dynamic evolution of structural behavior based on multivariable input data streams. By introducing a dynamic logical coupling reasoning mechanism, it identifies dynamic anomaly propagation paths between state variables.

[0064] The multivariate logical association modeling module includes a multivariate temporal feature extraction submodule, a dynamic coupling graph construction submodule, a logical rule reasoning submodule, and an anomaly propagation path identification submodule.

[0065] The multivariate temporal feature extraction submodule is used to perform dynamic slicing encoding of temporal features on a unified multivariate input data stream, extract the temporal morphological features of each state variable in the time dimension, and form a temporal feature vector set.

[0066] The dynamic coupling graph construction submodule is used to obtain a multi-dimensional dynamic coupling measure between different state variables based on the temporal feature vector set, construct a multi-variable dynamic coupling process that evolves over time, and express a multi-field interaction mechanism for structural behavior;

[0067] The logical rule reasoning submodule is used to introduce a dynamic logical coupling reasoning mechanism on the basis of the dynamic coupling process. Through the hybrid logical reasoning framework, it symbolically models the existing structural behavioral logical topological relationships and generates a structural behavioral logical relationship network.

[0068] The anomaly propagation path identification submodule is used to traverse the activation paths in the structural behavioral logic relationship network, detect behavioral chains that deviate from the expected logic, track the anomaly propagation timing characteristics of abnormal signals between variable nodes, and output the dynamic anomaly propagation path.

[0069] Furthermore, the multivariate logical correlation modeling module systematically extracts the intrinsic evolutionary laws and anomaly transmission mechanisms of structural behavior from the original fused data through four progressively advancing, logically closed-loop sub-modules. First, the multivariate temporal feature extraction sub-module performs a sliding window-style dynamic slicing of the unified multivariate input data stream. Combining morphological features such as temporal trends, fluctuation patterns, and local extrema, it encodes each state variable in a high-dimensional temporal sequence, forming a temporal feature vector set that characterizes its dynamic behavior. Subsequently, the dynamic coupling graph construction sub-module, based on this feature vector set, calculates the mutual information, phase synchronization, and nonlinear correlation between different state variables at multiple time scales, quantifies their coupling strength and direction, and constructs a multivariate dynamic coupling graph that evolves over time, intuitively expressing the interaction mechanism of multiple physical fields such as force, heat, humidity, and vibration in the structural response. Building upon this foundation, the logical rule reasoning submodule introduces a dynamic logical coupling reasoning mechanism, integrating Boolean logic constraints and fuzzy causal rules. It symbolically abstracts the causal dependencies, feedback loops, and collaborative failure modes implicit in the dynamic coupling diagram, generating an interpretable structural behavioral logical relationship network. Finally, the anomaly propagation path identification submodule traverses this logical relationship network in real time, identifies activation paths caused by external disturbances or internal degradation, detects behavioral chains that deviate from normal logic, and tracks the transmission order and time delay characteristics of abnormal signals between variable nodes. Ultimately, it outputs a complete dynamic anomaly propagation path, providing accurate logical basis for fault tracing and risk mapping.

[0070] Furthermore, the multivariate logical correlation modeling module constructs a complete reasoning path from raw time-series data to an interpretable anomaly logic chain through four highly collaborative and progressively abstract sub-modules. First, the multivariate time-series feature extraction sub-module, based on a unified multivariate input data stream, employs a sliding window mechanism to continuously slice each state variable over time, extracting key time-series morphological features such as trends (e.g., rate of increase / decrease), fluctuation patterns (e.g., periodicity, random oscillation), and local extrema (e.g., peaks, valleys) within each time window. This forms a structured high-dimensional time-series feature vector set, accurately characterizing the dynamic behavior of each physical quantity. Next, the dynamic coupling graph construction sub-module, based on this feature vector set, calculates the time-varying mutual information, phase synchronization, and nonlinear dependencies between different state variables at multiple time scales. This quantifies their real-time interaction strength and direction under the coupling effects of multiple physical fields such as force, heat, humidity, and vibration, and constructs a multivariate dynamic coupling graph that evolves over service time, making the intrinsic correlation of the overall structural response visible and traceable. Finally, logical rule reasoning... Based on the dynamic coupling diagram, the submodule introduces a hybrid logical reasoning framework that integrates Boolean hard constraints (such as a sudden increase in beam end displacement if the support fails) and fuzzy causal rules (such as the possibility that a continuous increase in temperature and humidity may accelerate steel corrosion). It symbolically models the causal dependencies, positive or negative feedback loops, and multi-factor collaborative failure modes between variables, generating a semantically clear and mechanistically traceable structural behavior logical relationship network. Finally, the anomaly propagation path identification submodule performs real-time activation path traversal of this logical relationship network. Once a node state is detected to deviate from the historical normal logical pattern, the anomaly propagation tracking mechanism is activated. It analyzes how the abnormal signal is transmitted, amplified, or transformed between variable nodes along the logical edges, and accurately records its propagation order and time delay characteristics. This outputs a dynamic anomaly propagation path with temporal causality and physical rationality, providing a high-confidence logical basis for subsequent risk mapping, fault location, and proactive control.

[0071] S3: Closed-loop active operation and maintenance control loop module, used to map the abnormal risk mapping mechanism output by the multivariable logical correlation modeling module to the structural operation and maintenance execution unit to generate adaptive control instructions.

[0072] The closed-loop active operation and maintenance control loop module includes an anomaly risk semantic parsing submodule, an execution unit capability matching submodule, an adaptive strategy generation submodule, and a closed-loop instruction issuance feedback verification submodule.

[0073] The anomaly risk semantic parsing submodule is used to receive the dynamic anomaly propagation path output by the multivariate logical association modeling module, perform structured parsing, and generate anomaly risk description vectors.

[0074] The execution unit capability matching submodule is used to construct execution unit capability representations based on the comprehensive capability characteristics of various operation and maintenance execution units deployed in the building structure, and to perform multi-dimensional matching between the abnormal risk description vector and the execution unit capability representations to select a set of candidate execution units that can participate in regulation.

[0075] The adaptive strategy generation submodule is used to dynamically combine or generate control strategy rules that are adapted to the current abnormal risk scenario based on the control strategy knowledge base and the current environmental boundary conditions, and convert them into parameterized instruction templates.

[0076] The closed-loop instruction issuance feedback verification submodule is used to encapsulate parameterized instruction templates into structured control instructions and issue them to the corresponding operation and maintenance execution units in real time. It collects the response status and structured feedback signals of the execution units, verifies whether the control effect meets the expected logic, and triggers the strategy replanning process when the deviation exceeds the threshold.

[0077] Furthermore, the closed-loop active operation and maintenance control loop module, through four closely connected sub-modules forming a complete feedback loop, realizes fully automated decision-making from anomaly identification to execution control and effect verification. First, the anomaly risk semantic parsing submodule receives the dynamic anomaly propagation path output by the multivariate logical association modeling module. Combined with a predefined risk ontology system, it performs structured parsing of the anomaly types (e.g., support slippage, crack propagation, damping failure), impact range, evolution stage, and potential consequences, transforming them into standardized, machine-readable anomaly risk description vectors. Subsequently, the execution unit capability matching submodule constructs a unified execution unit capability representation based on the comprehensive capability characteristics of various operation and maintenance execution units (e.g., adjustable dampers, intelligent tensioning devices, environmental control equipment) actually deployed in the building structure, including functional attributes, response speed, control amplitude, and current availability. This representation is then intelligently matched with the anomaly risk description vector across multiple dimensions, such as functional dimension, spatial location, and response time, to select a set of candidate execution units capable of effectively addressing the current risk. Building upon this, the adaptive strategy generation submodule calls upon the built-in control strategy knowledge base, combining current environmental boundary conditions (e.g., wind load level, temperature and humidity levels, personnel occupancy). The system dynamically combines or generates control strategy rules in real time to adapt to the current abnormal scenario (e.g., when the displacement at the beam end exceeds the limit, it simultaneously activates the magnetorheological dampers of the two adjacent spans and reduces the air conditioning load to reduce thermal stress). This strategy is then transformed into a parameterized instruction template for specific execution units, specifying key parameters such as action type, target value, and duration. Finally, the closed-loop instruction issuance feedback verification submodule encapsulates the parameterized instruction template into structured control instructions and issues them to the corresponding operation and maintenance execution units in real time through the edge controller. At the same time, it simultaneously collects the execution unit's own operating status (such as current, stroke, start / stop status) and structural feedback signals from the adjacent area (i.e., execution-structure coupling feedback data) and compares them with the expected control logic. If the actual response deviates from the expected response by more than a preset threshold, the strategy replanning process is immediately triggered, and the system re-enters the matching and generation stage, thereby ensuring that the entire control process is always in a controllable, calibrable, and optimizable closed-loop state.

[0078] S4: Fault Intelligent Diagnosis Early Warning Module, which combines multivariate degradation correlation features and adopts an incremental learning online model update mechanism to locate and predict the trend of diagnosed faults, and triggers graded warning signals before the diagnosed fault occurs.

[0079] The fault intelligent diagnosis early warning module includes a multivariate degradation feature fusion submodule, an incremental degradation pattern learning submodule, a fault location trend inference submodule, and a graded early warning triggering submodule.

[0080] The multivariate degradation feature fusion submodule is used to receive dynamic anomaly propagation paths, combine them with structural service environment parameters, extract and fuse multi-dimensional correlation features to form a unified multivariate degradation correlation feature set;

[0081] The incremental degradation pattern learning submodule is used to continuously update the fault pattern library based on a multivariate degradation association feature set using an incremental learning mechanism. It identifies degradation trajectory clusters online and dynamically expands and optimizes the fault representation space.

[0082] The fault location trend inference submodule is used to update the matching results in the fault mode library based on the multivariate degradation association feature set, determine the fault development stage, and combine the temporal extrapolation physical constraint fusion mechanism to make multi-step predictions on the future evolution path of fault indicators and generate fault development trend curves.

[0083] The graded early warning triggering submodule is used to compare the fault development trend curve with the deviation exceeding the threshold. When the predicted degree of deterioration will cross any early warning threshold within the time window, the corresponding level of early warning signal will be automatically activated, the early warning level will be divided, and early warning instructions will be output.

[0084] Furthermore, the fault intelligent diagnosis early warning module constructs a full-process intelligent diagnosis mechanism through four organically connected and progressively advanced sub-modules, from degradation feature fusion to trend prediction and then to hierarchical early warning. First, the multivariate degradation feature fusion submodule receives dynamic anomaly propagation paths from the multivariate logical association modeling module and integrates the current service environment parameters of the structure (such as temperature, humidity, corrosive gas concentration, load history, etc.). It extracts multi-dimensional correlation features related to typical failure modes such as material aging, connection loosening, and stiffness degradation, including anomaly propagation intensity, variable coupling deviation, and environmental acceleration factors, forming a unified, high-dimensional multivariate degradation correlation feature set. Subsequently, the incremental degradation mode learning submodule, based on this feature set, employs an incremental learning mechanism that does not require full retraining to identify newly emerging degradation behavior trajectory clusters online. These clusters are then compared and clustered with existing failure modes to dynamically expand and optimize the representation space in the failure mode library, enabling the system to continuously adapt to new degradation patterns evolving during long-term service. Finally, the fault location trend inference submodule combines the current degradation features with the updated failure modes. The system matches data with a database to accurately determine the location, type, and development stage (e.g., initial emergence, mid-term expansion, critical instability) of potential faults. It then incorporates a time-series extrapolation mechanism that integrates physical constraints (e.g., material strength limits, allowable variable values) to perform multi-step predictions of the evolution path of key fault indicators (e.g., crack width, frequency drift, damping loss factor) over a future period, generating a fault development trend curve with confidence intervals. Finally, the tiered early warning triggering submodule compares this trend curve with preset multi-level safety thresholds (e.g., attention level, early warning level, alarm level). Once the prediction result indicates that the degree of degradation will cross a certain threshold within a specific time window, the corresponding level of early warning signal is automatically activated. The early warning level is then finely divided based on the urgency of the risk, the scope of impact, and the intervention window, outputting structured early warning instructions that are pushed to the operation and maintenance platform or control loop. This achieves the proactive health management goal of early detection, early warning, and early intervention.

[0085] Among them, the structural operation and maintenance execution unit is used to receive parameterized instruction templates issued by the closed-loop active operation and maintenance control loop module and intervene in the mechanical state of the building structure;

[0086] During the intervention process, the execution structure coupling feedback data is collected and sent back to the closed-loop instruction issuance feedback verification submodule to complete the execution feedback.

[0087] The fault location trend inference submodule includes updating the matching results in the fault mode library based on the multivariate degradation association feature set, determining the fault development stage, and combining the temporal extrapolation physical constraint fusion mechanism to make multi-step predictions on the future evolution path of fault indicators.

[0088] Fault indicators at the prediction time status Calculate using the following formula:

[0089]

[0090] in, This indicates the state value of the fault indicator at the predicted time. This represents the multivariate degenerate association feature vector at the current moment. This represents the temporal extrapolation mapping function that incorporates physical constraints. This represents the set of updatable parameters for the extrapolation mapping function. This represents the residual perturbation term that satisfies the physical boundary conditions.

[0091] Furthermore, the structural operation and maintenance execution unit, as the physical execution end of the closed-loop active operation and maintenance control loop, is not only responsible for receiving parameterized instruction templates (such as target stiffness value, damping force magnitude, tension force setpoint, etc.) issued by the adaptive strategy generation submodule, and driving adjustable dampers, intelligent tension cables, magnetorheological supports, or environmental adjustment devices to intervene in the mechanical state or service environment of the building structure in real time, but also synchronously collects its own operating status (such as action arrival signal, energy consumption, response delay) and dynamic response data (such as displacement, acceleration, strain changes) of adjacent structural areas during the execution process, forming execution-structure coupling feedback data, and transmitting this data back to the closed-loop instruction issuance feedback verification submodule in real time to verify whether the actual control effect is consistent with the expected logic, thereby supporting the dynamic correction and closed-loop optimization of the strategy. Meanwhile, during the diagnostic phase, the fault location trend extrapolation submodule, based on a continuously updated multivariate degradation correlation feature set, performs high-precision matching with an expanded fault mode library after incremental learning. This accurately identifies the current development stage of the fault (such as the nascent stage, accelerated degradation stage, or critical failure precursor). On this basis, it integrates prior physical knowledge such as structural material performance limits, geometric constraints, and load boundaries to construct an engineering-rational time-series extrapolation mechanism. This mechanism robustly predicts the evolution trend of key fault indicators over multiple future time steps, generating a fault development trend curve containing a reasonable confidence interval. This provides a reliable basis for graded early warning, ensuring that early warnings neither lead to false alarms too early nor miss the intervention window too late.

[0092] In a preferred embodiment, a multivariable logic monitoring and fault diagnosis method for an active operation and maintenance control loop of a building structure includes: collecting state variables under the action of multiple physical fields; performing multi-source data collaborative preprocessing on raw data from different sensor types to construct a unified multivariable input data stream; constructing logical relationships reflecting the dynamic evolution of structural behavior based on the multivariable input data stream; identifying dynamic anomaly propagation paths between state variables by introducing a dynamic logic coupling reasoning mechanism; mapping the anomaly risk mapping mechanism output by the multivariable logic association modeling module to the structural operation and maintenance execution unit to generate adaptive control instructions; and combining multivariable degradation association characteristics with an incremental learning online model update mechanism to locate and predict the trend of diagnosed faults, triggering graded early warning signals before diagnosed faults occur.

[0093] The above-mentioned unit modules can be embedded in the processor of the computer device in hardware form or independent of it, or they can be stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of the above modules.

[0094] In one embodiment, a computer device is provided, which may be a terminal. The computer device includes a processor, memory, a communication interface, a display screen, and an input device connected via a system bus. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage medium. The communication interface of the computer device is used for wired or wireless communication with external terminals. Wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen of the computer device may be an LCD screen or an e-ink display screen. The input device of the computer device may be a touch layer covering the display screen, or buttons, a trackball, or a touchpad located on the casing of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0095] In summary, this invention achieves a fundamental shift in building structure operation and maintenance from passive response to proactive intervention by constructing a closed-loop intelligent system integrating perception, modeling, decision-making, execution, and early warning. The multi-source heterogeneous sensor fusion module effectively eliminates variable isolation and information fragmentation, ensuring the spatiotemporal consistency and fusionability of multi-physics state data. The multi-variable logical association modeling module, through a dynamic logical coupling reasoning mechanism, accurately characterizes the nonlinear coupling relationships and anomaly propagation paths between variables, significantly improving the logical completeness and mechanistic interpretability of fault identification. The closed-loop proactive operation and maintenance control loop module breaks down the barriers between monitoring and diagnosis and execution control, enabling risk-driven adaptive command generation and real-time feedback verification. The intelligent fault diagnosis and early warning module, combining incremental learning and physical constraint extrapolation, can complete fault trend prediction and graded early warning before significant structural performance degradation. Overall, this invention significantly improves the safety, reliability, and intelligent operation and maintenance level of building structures in complex service environments, providing key technical support for the full life-cycle health management of smart city infrastructure.

[0096] 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 multivariable logic monitoring and fault diagnosis system for an active operation and maintenance control loop of a building structure, characterized in that: include, The multi-source heterogeneous sensing fusion module is used to collect state variables under the action of multiple physical fields, perform multi-source data collaborative preprocessing on raw data from different sensor types, and construct a unified multivariable input data stream; The multivariate logical association modeling module is used to construct logical relationships that reflect the dynamic evolution of structural behavior based on multivariate input data streams. By introducing a dynamic logical coupling reasoning mechanism, it identifies dynamic anomaly propagation paths between state variables. The closed-loop active operation and maintenance control loop module is used to map the abnormal risk mapping mechanism output by the multivariate logical correlation modeling module to the structural operation and maintenance execution unit to generate adaptive control instructions; The intelligent fault diagnosis early warning module combines multivariate degradation correlation features and adopts an incremental learning online model update mechanism to locate and predict the trend of diagnosed faults, triggering graded early warning signals before the diagnosed fault occurs.

2. The multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structures as described in claim 1, characterized in that: The multi-source heterogeneous sensing fusion module includes a multi-physics field state perception submodule, a multi-source temporal alignment submodule, a spatial topology mapping submodule, and a feature normalization fusion submodule. The multi-physics field state perception submodule is used to deploy a multi-type sensor network on the building structure to collect raw data of multi-dimensional state variables; The multi-source timing alignment submodule is used to perform timing collaborative calibration on multi-variable input data streams and align various state variables under a unified time reference. The spatial topology mapping submodule is used to map the state variables collected by each sensor to the corresponding physical location nodes based on the structural sensing spatial registration benchmark, and to establish a multivariate distribution framework. The feature normalization fusion submodule is used to perform structured weighted normalization fusion on the original data of multidimensional state variables to generate a unified multivariate input data stream.

3. The multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structures as described in claim 2, characterized in that: The multivariate logical association modeling module includes a multivariate temporal feature extraction submodule, a dynamic coupling graph construction submodule, a logical rule reasoning submodule, and an anomaly propagation path identification submodule. The multivariate temporal feature extraction submodule is used to perform dynamic slicing encoding of temporal features on a unified multivariate input data stream, extract the temporal morphological features of each state variable in the time dimension, and form a temporal feature vector set. The dynamic coupling graph construction submodule is used to obtain a multi-dimensional dynamic coupling metric between different state variables based on the time-series feature vector set, construct a multi-variable dynamic coupling process that evolves over time, and express a multi-field interaction mechanism for structural behavior. The logical rule reasoning submodule is used to introduce a dynamic logical coupling reasoning mechanism on the basis of the dynamic coupling process. Through the hybrid logical reasoning framework, the existing structural behavioral logical topological relationship is symbolically modeled to generate a structural behavioral logical relationship network. The anomaly propagation path identification submodule is used to traverse the activation paths in the structural behavioral logic relationship network, detect behavioral chains that deviate from the expected logic, track the anomaly propagation timing characteristics of abnormal signals between variable nodes, and output the dynamic anomaly propagation path.

4. The multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structures as described in claim 3, characterized in that: The closed-loop active operation and maintenance control loop module includes an anomaly risk semantic parsing submodule, an execution unit capability matching submodule, an adaptive strategy generation submodule, and a closed-loop instruction issuance feedback verification submodule. The anomaly risk semantic parsing submodule is used to receive the dynamic anomaly propagation path output by the multivariate logical association modeling module, perform structured parsing, and generate anomaly risk description vectors. The execution unit capability matching submodule is used to construct an execution unit capability representation based on the comprehensive capability characteristics of various operation and maintenance execution units deployed in the building structure, and to perform multi-dimensional matching between the abnormal risk description vector and the execution unit capability representation to screen out a set of candidate execution units that can participate in regulation. The adaptive strategy generation submodule is used to dynamically combine or generate control strategy rules that are adapted to the current abnormal risk scenario based on the control strategy knowledge base and the current environmental boundary conditions, and convert them into parameterized instruction templates. The closed-loop instruction issuance feedback verification submodule is used to encapsulate the parameterized instruction template into a structured control instruction and issue it to the corresponding operation and maintenance execution unit in real time. It collects the response status and structured feedback signal of the execution unit, verifies whether the control effect meets the expected logic, and triggers the strategy replanning process when the deviation exceeds the threshold.

5. The multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structures as described in claim 4, characterized in that: The fault intelligent diagnosis early warning module includes a multivariate degradation feature fusion submodule, an incremental degradation pattern learning submodule, a fault location trend inference submodule, and a graded early warning triggering submodule. The multivariate degradation feature fusion submodule is used to receive dynamic anomaly propagation paths, combine structural service environment parameters, extract and fuse multi-dimensional correlation features to form a unified multivariate degradation correlation feature set. The incremental degradation pattern learning submodule is used to continuously update the fault pattern library based on a multivariate degradation association feature set using an incremental learning mechanism, and to dynamically expand and optimize the fault representation space by identifying degradation trajectory clusters online. The fault location trend inference submodule is used to update the matching results in the fault mode library based on the multivariate degradation association feature set, determine the fault development stage, and combine the temporal extrapolation physical constraint fusion mechanism to make multi-step predictions on the future evolution path of fault indicators and generate a fault development trend curve. The graded early warning triggering submodule is used to compare the fault development trend curve with the deviation exceeding the threshold. When the predicted degree of deterioration will cross any early warning threshold within the time window, the corresponding level of early warning signal will be automatically activated, the early warning level will be divided, and an early warning command will be output.

6. The multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structures as described in claim 5, characterized in that: The structural operation and maintenance execution unit is used to receive parameterized instruction templates issued by the closed-loop active operation and maintenance control loop module and intervene in the mechanical state of the building structure. During the intervention process, the execution structure coupling feedback data is collected and sent back to the closed-loop instruction issuance feedback verification submodule to complete the execution feedback.

7. The multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structures as described in claim 6, characterized in that: The fault location trend inference submodule includes updating the matching results in the fault mode library based on the multivariate degradation association feature set, determining the fault development stage, and combining the temporal extrapolation physical constraint fusion mechanism to perform multi-step prediction of the future evolution path of fault indicators. Fault indicators at the prediction time status Calculate using the following formula: in, This indicates the state value of the fault indicator at the predicted time. This represents the multivariate degenerate association feature vector at the current moment. This represents the temporal extrapolation mapping function that incorporates physical constraints. This represents the set of updatable parameters for the extrapolation mapping function. This represents the residual perturbation term that satisfies the physical boundary conditions.

8. A method for multivariable logic monitoring and fault diagnosis of an active operation and maintenance control loop for building structures, based on the multivariable logic monitoring and fault diagnosis system for an active operation and maintenance control loop for building structures as described in any one of claims 1 to 7, characterized in that: include, Collect state variables under the action of multi-physics fields, perform multi-source data collaborative preprocessing on raw data from different sensor types, and construct a unified multivariable input data stream; Based on multivariate input data streams, logical relationships reflecting the dynamic evolution of structural behavior are constructed. By introducing a dynamic logical coupling reasoning mechanism, dynamic anomaly propagation paths between state variables are identified. The abnormal risk mapping mechanism output by the multivariate logical correlation modeling module is mapped to the structural operation and maintenance execution unit to generate adaptive control instructions; By combining multivariate degradation association features, an incremental learning online model update mechanism is adopted to predict the location trend of diagnosed faults and trigger graded early warning signals before the diagnosed faults occur.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structures as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the multivariable logic monitoring and fault diagnosis system for the active operation and maintenance control loop of building structures as described in any one of claims 1 to 7.

Images