An emergency status alarm system

Abstract

The application relates to the technical field of disaster monitoring. By providing an emergency state alarm system, including: performing clock calibration processing on all sensor nodes through a synchronous clock protocol to generate a calibrated unified reference clock source; collecting and processing multi-source data of a target scene to generate a sensor data stream with a time stamp and an environmental interference parameter set; constructing a medium propagation delay prediction model according to the environmental interference parameter set and sensor physical characteristic parameters; performing dynamic compensation processing on the sensor data stream to generate a time-aligned multi-sensor data set; performing multi-modal fusion analysis on the time-aligned multi-sensor data set to trigger a hierarchical alarm signal, so as to solve the problems of insufficient dynamic coupling modeling of sensor degradation and environmental interference in a multi-modal conflict resolution mechanism and insufficient space-time feature driving caused by cross-sensor decision dependence on static weights, and realize accurate identification of abnormal events and optimization of hierarchical alarm.

Description

Technical Field

[0001] The present invention relates to the technical field of disaster monitoring, and in particular to an emergency state alarm system. Background Art

[0002] With the rapid development of the Internet of Things (IoT) and intelligent sensing technologies, emergency alarm systems are becoming increasingly important in areas such as natural disaster warning, industrial safety monitoring, and urban public safety. By collecting real-time environmental data from multiple sources (such as temperature, humidity, air pressure, vibration, and electromagnetic signals) and combining it with multimodal data analysis, these systems can effectively identify potential risks and trigger graded alarms, providing critical decision-making support for disaster prevention and emergency response.

[0003] However, the relevant emergency alarm systems still have a multimodal conflict resolution mechanism that lacks dynamic coupling analysis of sensor degradation status and environmental interference intensity, resulting in misjudgment of abnormal events; cross-sensor collaborative decision-making relies on static weight distribution rules, making it difficult to construct adaptive alarm triggering logic driven by spatiotemporal correlation features, resulting in insufficient accuracy of alarm decisions. Summary of the Invention

[0004] Based on this, it is necessary to provide an emergency alarm system to address the above technical problems, so as to solve the problem of insufficient dynamic coupling modeling of sensor degradation and environmental interference in the multimodal conflict resolution mechanism, as well as the problem of insufficient spatiotemporal feature driving caused by cross-sensor decision-making relying on static weights, so as to achieve accurate identification of abnormal events and hierarchical alarm optimization.

[0005] The present application provides an emergency alarm system, which includes:

[0006] The clock calibration module is used to perform clock calibration on all sensor nodes through the synchronous clock protocol to generate a calibrated unified reference clock source;

[0007] The multi-source data acquisition module is used to collect and process multi-source data of the target scene based on a calibrated unified reference clock source, generating a sensor data stream with a timestamp and an environmental interference parameter set;

[0008] The delay prediction modeling module is used to build a medium propagation delay prediction model based on the environmental interference parameter set and the sensor physical characteristic parameters;

[0009] The dynamic compensation alignment module is used to dynamically compensate the sensor data stream based on the medium propagation delay prediction model and the calibrated unified reference clock source to generate a time-aligned multi-sensor data set;

[0010] The multimodal alarm trigger module is used to perform multimodal fusion analysis on time-aligned multi-sensor data sets and trigger hierarchical alarm signals.

[0011] Furthermore, a medium propagation delay prediction model is constructed based on the environmental interference parameter set and the sensor physical characteristic parameters, including:

[0012] Perform feature extraction on the humidity and air pressure parameters in the environmental interference parameter set to generate environmental propagation impact factors;

[0013] Correlation analysis is performed on the signal delay coefficient and thermal stability coefficient in the physical characteristic parameters of the sensor to generate the inherent delay characteristics of the sensor;

[0014] The environmental propagation influencing factors and the inherent delay characteristics of the sensor are fused in multiple dimensions to generate a medium propagation delay prediction model.

[0015] Furthermore, the environmental propagation influencing factors and the inherent delay characteristics of the sensor are fused in multiple dimensions to generate a medium propagation delay prediction model, including:

[0016] Perform time-series alignment processing on environmental propagation influencing factors and sensor inherent delay characteristics to generate a spatiotemporal consistency feature vector;

[0017] Based on the dynamic environment weight allocation strategy, the environmental propagation influencing factors in the spatiotemporal consistency feature vector and the sensor's inherent delay characteristics are adaptively weighted and fused to generate a hybrid delay feature matrix.

[0018] The mixed delay characteristic matrix is ​​processed by nonlinear mapping to generate a medium propagation delay prediction model.

[0019] Furthermore, based on the dynamic environment weight allocation strategy, the environmental propagation influencing factors in the spatiotemporal consistency feature vector and the sensor inherent delay characteristics are adaptively weighted and fused to generate a hybrid delay feature matrix, including:

[0020] Perform interference intensity classification on the environmental propagation impact factors in the spatiotemporal consistency feature vector to generate an environmental interference intensity level;

[0021] Performing degradation sensitivity evaluation on the sensor inherent delay characteristics in the spatiotemporal consistency feature vector to generate a sensor degradation level;

[0022] Based on the real-time dynamic relationship between the environmental interference intensity level and the sensor degradation level, the fusion weights of the environmental propagation influencing factors and the sensor's inherent delay characteristics are assigned to generate a hybrid delay feature matrix.

[0023] Furthermore, multimodal fusion analysis is performed on the time-aligned multi-sensor dataset to trigger graded alarm signals, including:

[0024] Perform spatiotemporal feature enhancement on the time-aligned multi-sensor dataset to generate a multimodal feature matrix;

[0025] Based on the conflict resolution rules between sensors, the multimodal feature matrix is ​​logically prioritized to generate conflict-resolved fusion data.

[0026] According to the dynamic weight of environmental interference and the sensor health status score, adaptive weighted decision processing is performed on the fused data after conflict resolution to generate a graded alarm signal.

[0027] Furthermore, based on the dynamic weight of environmental interference and the sensor health status score, adaptive weighted decision processing is performed on the fused data after conflict resolution to generate graded alarm signals, including:

[0028] Perform interference intensity quantification on the dynamic weight of environmental interference to generate an environmental interference intensity level;

[0029] Perform degradation tolerance mapping on the sensor health status score to generate the sensor credibility coefficient;

[0030] Based on the dynamic coupling relationship between the environmental interference intensity level and the sensor credibility coefficient, the fusion data after conflict resolution is processed with real-time weight allocation to generate a dynamic weighted decision vector;

[0031] The dynamic weighted decision vector is processed by threat level threshold matching to generate a graded alarm signal.

[0032] Furthermore, based on the calibrated unified reference clock source, multi-source data of the target scene is collected and processed to generate a sensor data stream with a timestamp and an environmental interference parameter set, including:

[0033] Based on the calibrated unified reference clock source, dynamic synchronization trigger processing is performed on multi-source sensor nodes to generate clock-synchronized sensor acquisition instructions;

[0034] According to the predefined hierarchical acquisition strategy, the physical parameters and electromagnetic parameters of the target scene are collected and processed in a time-sharing manner to generate an environmental interference parameter set;

[0035] The time-space correlation verification process is performed on the multi-source data stream triggered by the clock-synchronized sensor acquisition instruction to generate the sensor data stream with time stamp and the environmental interference parameter set.

[0036] Furthermore, all sensor nodes are calibrated using a synchronous clock protocol to generate a calibrated unified reference clock source, including:

[0037] Based on the dynamic clock source selection strategy, the master clock nodes in the candidate clock source pool are evaluated for environmental interference resistance and the optimal master clock node is generated.

[0038] According to the clock signal of the optimal master clock node, the clock offset of the multi-source sensor nodes is calculated and processed in a consistent manner through a distributed consensus algorithm to generate a distributed consistent clock offset parameter;

[0039] Based on the distributed consistency clock offset parameter, the local clock of the sensor node is decentralized converged and adjusted to generate a calibrated unified reference clock source.

[0040] Furthermore, based on the dynamic clock source selection strategy, the master clock nodes in the candidate clock source pool are evaluated for environmental interference resistance to generate the optimal master clock node, including:

[0041] Perform real-time collection and processing of environmental interference parameter sets for the master clock nodes in the candidate clock source pool to generate node interference feature vectors.

[0042] Based on the node interference feature vector, the signal stability of the master clock node is processed for anti-interference scoring to generate a node stability score;

[0043] According to the dynamic ranking results of the node stability scores, the collaborative verification switching process of the master clock node is performed to generate the optimal master clock node.

[0044] Furthermore, based on the medium propagation delay prediction model and the calibrated unified reference clock source, the sensor data stream is dynamically compensated to generate a time-aligned multi-sensor dataset, including:

[0045] Performing initial delay compensation processing on the sensor data stream based on the medium propagation delay prediction model to generate an initial compensated data stream;

[0046] Generate real-time error feedback parameters based on the real-time comparison results between the calibrated unified reference clock source and the medium propagation delay prediction model;

[0047] Based on the real-time error feedback parameters, dynamic error correction processing is performed on the initial compensation data stream to generate a real-time correction data stream;

[0048] Multi-channel time series integration processing is performed on the real-time correction data stream to generate a time-aligned multi-sensor dataset.

[0049] The technical solution provided by the present application includes the following technical effects: by providing an emergency alarm system, including: a clock calibration module, which is used to perform clock calibration processing on all sensor nodes through a synchronous clock protocol to generate a calibrated unified reference clock source; a multi-source data acquisition module, which is used to collect and process multi-source data of the target scene based on the calibrated unified reference clock source to generate a sensor data stream with a timestamp and an environmental interference parameter set; a delay prediction modeling module, which is used to construct a medium propagation delay prediction model based on the environmental interference parameter set and the sensor physical characteristic parameters; a dynamic compensation alignment module, which is used to perform dynamic compensation processing on the sensor data stream based on the medium propagation delay prediction model and the calibrated unified reference clock source to generate a time-aligned multi-sensor data set; a multimodal alarm triggering module, which is used to perform multimodal fusion analysis on the time-aligned multi-sensor data set to trigger a hierarchical alarm signal, so as to solve the problem of insufficient dynamic coupling modeling of sensor degradation and environmental interference in the multimodal conflict resolution mechanism, and the problem of insufficient spatiotemporal feature driving caused by cross-sensor decision-making relying on static weights, thereby realizing accurate identification of abnormal events and hierarchical alarm optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following briefly introduces the drawings required for use in the embodiments or related technical descriptions. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0051] Figure 1 is a structural diagram of an emergency alarm system in one embodiment of the present invention;

[0052] Figure 2 This is a flowchart of generating a time-aligned multi-sensor data set by dynamically compensating sensor data streams based on a medium propagation delay prediction model and a calibrated unified reference clock source in one embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application are described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to fully understand the present application. However, the present application can be implemented in many other ways than those described herein, and those skilled in the art can make similar improvements without violating the connotation of the application. Therefore, the present application is not limited to the specific embodiments disclosed below.

[0054] like Figure 1 As shown, the present application provides an emergency alarm system 100, which includes:

[0055] The clock calibration module 101 is configured to perform clock calibration processing on all sensor nodes through a synchronous clock protocol to generate a calibrated unified reference clock source.

[0056] Specifically, the system collects environmental interference parameters from candidate nodes in real time to generate a node interference feature vector. Based on this feature vector, the node's signal stability is scored for interference resistance. The optimal master clock node is then dynamically selected based on the score. Using the optimal master clock node's clock signal as a benchmark, a distributed consensus algorithm is used to calculate the clock offsets of all sensor nodes, generating a global clock offset parameter. This process ensures that all nodes reach consensus on the clock offset, minimizing the impact of single-point failures or local errors. Based on the calculated global clock offset parameter, a distributed convergence adjustment is performed on each sensor node's local clock. Each node gradually adjusts its local clock according to the offset parameter, ultimately achieving global time consistency. After these steps, the local clocks of all sensor nodes are aligned with the optimal master clock node, generating a unified, calibrated reference clock source that ensures time synchronization for subsequent multi-source data acquisition and processing. This process effectively addresses hardware drift, communication delays, and environmental interference in complex environments, ensuring timestamp consistency across multi-source data and laying the foundation for subsequent dynamic compensation and multimodal fusion analysis.

[0057] The multi-source data acquisition module 102 is used to collect and process multi-source data of the target scene based on the calibrated unified reference clock source, and generate a sensor data stream with a time stamp and an environmental interference parameter set.

[0058] Specifically, a calibrated unified reference clock source is used to generate clock-synchronized sensor acquisition instructions, ensuring that all sensor nodes initiate data acquisition tasks based on the same time reference. Based on a predefined hierarchical acquisition strategy, physical parameters (such as temperature, humidity, air pressure, and vibration) and electromagnetic parameters (such as electromagnetic signal strength and frequency) of the target scene are collected and processed in a time-sharing manner to ensure comprehensive and accurate data collection. Multi-source data streams triggered by the clock synchronization instructions are marked in real time to generate timestamped sensor data streams. Furthermore, spatiotemporal correlation verification is performed on the data streams to ensure temporal and spatial consistency. During data collection, environmental interference parameters (such as humidity and air pressure) are extracted in real time to generate an environmental interference parameter set, providing basic data for subsequent dynamic compensation and multimodal fusion analysis. The timestamped sensor data streams and the environmental interference parameter set are integrated to ensure data integrity and availability, providing high-quality input data for subsequent delay prediction modeling and multimodal fusion analysis. This process ensures timestamp consistency for multi-source data, effectively addressing the challenges of data collection in complex environments and laying a solid foundation for subsequent dynamic compensation and multimodal fusion analysis.

[0059] The delay prediction modeling module 103 is used to construct a medium propagation delay prediction model based on the environmental interference parameter set and the sensor physical characteristic parameters.

[0060] Specifically, key parameters such as humidity and air pressure are extracted from the environmental interference parameter set to generate environmental propagation influencing factors. Sensor physical characteristics (such as signal delay coefficient and thermal stability coefficient) are analyzed to generate the sensor's intrinsic delay signature. The environmental propagation influencing factors and the sensor's intrinsic delay signature are time-aligned to generate a spatiotemporally consistent feature vector. Based on a dynamic environmental weight allocation strategy, the environmental propagation influencing factors and the sensor's intrinsic delay signature in the feature vector are adaptively weighted and fused to generate a mixed delay signature matrix. The mixed delay signature matrix is ​​modeled using nonlinear mapping methods (such as artificial neural networks or Gaussian process regression) to generate a medium propagation delay prediction model. The model's effectiveness is verified using long-term field data, and the model is optimized to improve prediction accuracy. This process enables dynamic prediction of medium propagation delay, effectively addressing nonlinear effects in complex environments and providing fundamental support for subsequent dynamic compensation and multimodal fusion analysis.

[0061] The dynamic compensation alignment module 104 is configured to perform dynamic compensation processing on the sensor data stream based on the medium propagation delay prediction model and the calibrated unified reference clock source to generate a time-aligned multi-sensor data set.

[0062] Specifically, using a pre-built medium propagation delay prediction model, initial delay compensation is performed on the sensor data stream to generate an initial compensated data stream. This process estimates the signal propagation delay using the prediction model and makes preliminary adjustments to the data stream's timestamps. Real-time error feedback parameters are generated based on the real-time comparison results between a calibrated unified reference clock source and the medium propagation delay prediction model. Dynamic error correction is applied to the initial compensated data stream using these parameters to generate a real-time corrected data stream, improving time alignment accuracy. Multi-channel time series integration is then performed on the real-time corrected data stream to ensure temporal consistency of all sensor data, ultimately generating a time-aligned multi-sensor dataset. This process uses methods such as interpolation and extrapolation to fuse data from different sensors using a unified time base. This process effectively addresses signal propagation delay issues in complex environments, ensures accurate time alignment of multi-sensor data, and provides high-quality data support for subsequent multimodal fusion analysis and alarm triggering.

[0063] The multimodal alarm triggering module 105 is used to perform multimodal fusion analysis on the time-aligned multi-sensor data set and trigger a graded alarm signal.

[0064] Specifically, the time-aligned multi-sensor dataset is subjected to spatiotemporal feature enhancement to generate a multimodal feature matrix. Feature extraction and enhancement techniques are used to extract key features from each sensor data and integrate them into a unified feature matrix, providing a foundation for subsequent fusion analysis. Based on inter-sensor conflict resolution rules, the multimodal feature matrix is ​​logically prioritized to generate conflict-resolved fused data. By analyzing the consistency and conflict points of sensor data, inconsistencies between the data are resolved, ensuring the reliability of the fused data. Adaptive weighted decision processing is performed on the conflict-resolved fused data based on the dynamic weights of environmental interference and sensor health scores. By quantifying environmental interference intensity and assessing sensor health, the weights of each sensor data are dynamically adjusted to generate a dynamic weighted decision vector. The dynamic weighted decision vector is then matched to a threat level threshold to generate a graded alarm signal. The decision vector is evaluated based on a preset threat level threshold, triggering the corresponding alarm signal, enabling accurate identification of abnormal events and graded alarms. This process effectively resolves conflicts in multi-source data, dynamically adjusts decision weights, and ensures the accuracy and timeliness of alarm signals, providing reliable support for disaster prevention and emergency response.

[0065] An embodiment of the present application provides an emergency alarm system, comprising: a clock calibration module, configured to perform clock calibration processing on all sensor nodes through a synchronous clock protocol to generate a calibrated unified reference clock source; a multi-source data acquisition module, configured to acquire and process multi-source data of a target scene based on the calibrated unified reference clock source to generate a sensor data stream with a timestamp and an environmental interference parameter set; a delay prediction modeling module, configured to construct a medium propagation delay prediction model based on the environmental interference parameter set and sensor physical characteristic parameters; a dynamic compensation alignment module, configured to perform dynamic compensation processing on the sensor data stream based on the medium propagation delay prediction model and the calibrated unified reference clock source to generate a time-aligned multi-sensor data set; and a multimodal alarm triggering module, configured to perform multimodal fusion analysis on the time-aligned multi-sensor data set to trigger a hierarchical alarm signal to resolve the problem of insufficient dynamic coupling modeling of sensor degradation and environmental interference in the multimodal conflict resolution mechanism, as well as the problem of insufficient spatiotemporal feature driving caused by cross-sensor decision-making relying on static weights, thereby achieving accurate identification of abnormal events and hierarchical alarm optimization.

[0066] Furthermore, a medium propagation delay prediction model is constructed based on the environmental interference parameter set and the sensor physical characteristic parameters, including:

[0067] Perform feature extraction on the humidity and air pressure parameters in the environmental interference parameter set to generate environmental propagation impact factors;

[0068] Correlation analysis is performed on the signal delay coefficient and thermal stability coefficient in the physical characteristic parameters of the sensor to generate the inherent delay characteristics of the sensor;

[0069] The environmental propagation influencing factors and the inherent delay characteristics of the sensor are fused in multiple dimensions to generate a medium propagation delay prediction model.

[0070] Specifically, key parameters such as humidity and air pressure are extracted from the environmental interference parameter set to generate environmental propagation impact factors. This step is implemented using a feature extraction algorithm to capture the dynamic impact of environmental parameters on signal propagation. The signal delay coefficient and thermal stability coefficient within the sensor's physical parameters are analyzed to generate the sensor's intrinsic delay signature. This step is implemented using correlation analysis to quantify the contribution of sensor hardware characteristics to signal propagation delay. The environmental propagation impact factors are then integrated with the sensor's intrinsic delay signature in a multi-dimensional manner. This step is implemented using time series alignment and adaptive weighted fusion methods to generate a hybrid delay signature matrix, ensuring that the model comprehensively accounts for the impact of environmental and hardware characteristics. The hybrid delay signature matrix is ​​modeled using nonlinear mapping methods (such as artificial neural networks or Gaussian process regression) to generate a medium propagation delay prediction model. This step is implemented using a machine learning algorithm to improve the model's adaptability to complex environments and its prediction accuracy. This process enables dynamic prediction of medium propagation delay, effectively addressing nonlinear effects in complex environments and providing foundational support for subsequent dynamic compensation and multimodal fusion analysis.

[0071] Furthermore, the environmental propagation influencing factors and the inherent delay characteristics of the sensor are fused in multiple dimensions to generate a medium propagation delay prediction model, including:

[0072] Perform time-series alignment processing on environmental propagation influencing factors and sensor inherent delay characteristics to generate a spatiotemporal consistency feature vector;

[0073] Based on the dynamic environment weight allocation strategy, the environmental propagation influencing factors in the spatiotemporal consistency feature vector and the sensor's inherent delay characteristics are adaptively weighted and fused to generate a hybrid delay feature matrix.

[0074] The mixed delay characteristic matrix is ​​processed by nonlinear mapping to generate a medium propagation delay prediction model.

[0075] Specifically, key parameters such as humidity and air pressure are extracted from the environmental interference parameter set to generate environmental propagation influencing factors. Simultaneously, signal delay coefficients and thermal stability coefficients are extracted from the sensor's physical characteristic parameters to generate the sensor's intrinsic delay signature. This step is implemented using a feature extraction algorithm, aiming to capture the dynamic impact of environmental and hardware characteristics on signal propagation. The extracted environmental propagation influencing factors and sensor intrinsic delay signatures are time-aligned to generate a spatiotemporally consistent feature vector. This step ensures temporal and spatial consistency of data from different sources, providing a foundation for subsequent fusion processing. Based on a dynamic environmental weight allocation strategy, the environmental propagation influencing factors and sensor intrinsic delay signatures in the spatiotemporally consistent feature vector are adaptively weighted and fused to generate a mixed delay signature matrix. This step is implemented using a weight allocation algorithm, aiming to dynamically adjust the importance of each feature to accommodate nonlinear changes in complex environments. A nonlinear mapping method (such as artificial neural networks or Gaussian process regression) is used to model the mixed delay signature matrix and generate a medium propagation delay prediction model. This step is implemented using a machine learning algorithm, aiming to improve the model's adaptability to complex environments and its prediction accuracy.

[0076] Furthermore, based on the dynamic environment weight allocation strategy, the environmental propagation influencing factors in the spatiotemporal consistency feature vector and the sensor inherent delay characteristics are adaptively weighted and fused to generate a hybrid delay feature matrix, including:

[0077] Perform interference intensity classification on the environmental propagation impact factors in the spatiotemporal consistency feature vector to generate an environmental interference intensity level;

[0078] Performing degradation sensitivity evaluation on the sensor inherent delay characteristics in the spatiotemporal consistency feature vector to generate a sensor degradation level;

[0079] Based on the real-time dynamic relationship between the environmental interference intensity level and the sensor degradation level, the fusion weights of the environmental propagation influencing factors and the sensor's inherent delay characteristics are assigned to generate a hybrid delay feature matrix.

[0080] Specifically, the environmental propagation impact factors in the spatiotemporal consistency feature vector are subjected to interference intensity classification to generate an environmental interference intensity level. This step analyzes real-time changes in environmental parameters (such as humidity and air pressure) to quantify their interference on signal propagation. The sensor's inherent delay features in the spatiotemporal consistency feature vector are subjected to degradation sensitivity assessment to generate a sensor degradation level. This step analyzes the sensor's signal delay coefficient and thermal stability coefficient to assess the sensor's degradation level in the current environment. Based on the real-time dynamic relationship between the environmental interference intensity level and the sensor degradation level, fusion weights are assigned to the environmental propagation impact factors and the sensor's inherent delay features. This step uses a dynamic weight allocation strategy to ensure that the weights of each feature can be adaptively adjusted under different environmental conditions to reflect their actual impact on signal propagation delay. Through these steps, a mixed delay feature matrix is ​​generated, providing the foundation for subsequent nonlinear mapping and model construction. This step uses matrix operations and feature fusion techniques to comprehensively consider environmental and sensor characteristics to generate a feature matrix that reflects the propagation characteristics of complex environments.

[0081] Furthermore, multimodal fusion analysis is performed on the time-aligned multi-sensor dataset to trigger graded alarm signals, including:

[0082] Perform spatiotemporal feature enhancement on the time-aligned multi-sensor dataset to generate a multimodal feature matrix;

[0083] Based on the conflict resolution rules between sensors, the multimodal feature matrix is ​​logically prioritized to generate conflict-resolved fusion data.

[0084] According to the dynamic weight of environmental interference and the sensor health status score, adaptive weighted decision processing is performed on the fused data after conflict resolution to generate a graded alarm signal.

[0085] Specifically, the time-aligned multi-sensor dataset is enhanced with spatiotemporal features to generate a multimodal feature matrix. Feature extraction and enhancement techniques are used to extract key features from each sensor data and integrate them into a unified feature matrix, providing a foundation for subsequent fusion analysis. Based on inter-sensor conflict resolution rules, the multimodal feature matrix is ​​logically prioritized to generate conflict-resolved fused data. By analyzing the consistency and conflict points of sensor data, inconsistencies between the data are resolved, ensuring the reliability of the fused data. Adaptive weighted decision processing is performed on the conflict-resolved fused data based on the dynamic weights of environmental interference and sensor health scores. By quantifying environmental interference intensity and assessing sensor health, the weights of each sensor data are dynamically adjusted to generate a dynamic weighted decision vector. The dynamic weighted decision vector is then matched to threat level thresholds to generate graded alarm signals. The decision vector is evaluated based on preset threat level thresholds, triggering corresponding alarm signals, enabling more accurate identification and graded alarming of abnormal events.

[0086] Furthermore, based on the dynamic weight of environmental interference and the sensor health status score, adaptive weighted decision processing is performed on the fused data after conflict resolution to generate graded alarm signals, including:

[0087] Perform interference intensity quantification on the dynamic weight of environmental interference to generate an environmental interference intensity level;

[0088] Perform degradation tolerance mapping on the sensor health status score to generate the sensor credibility coefficient;

[0089] Based on the dynamic coupling relationship between the environmental interference intensity level and the sensor credibility coefficient, the fusion data after conflict resolution is processed with real-time weight allocation to generate a dynamic weighted decision vector;

[0090] The dynamic weighted decision vector is processed by threat level threshold matching to generate a graded alarm signal.

[0091] Specifically, the dynamic weights of environmental interference are quantified to generate an environmental interference intensity level. This step analyzes real-time changes in environmental parameters (such as humidity and air pressure) to quantify their degree of interference with signal propagation. Degradation tolerance mapping is performed on the sensor health score to generate a sensor credibility coefficient. This step analyzes the sensor's health status to assess its reliability in the current environment. Based on the dynamic coupling relationship between the environmental interference intensity level and the sensor credibility coefficient, real-time weighting is assigned to the conflict-resolved fused data to generate a dynamic weighted decision vector. This step uses a dynamic weighting strategy to ensure that the weights of each feature are adaptively adjusted under different environmental conditions to reflect their actual impact on the abnormal event. The dynamic weighted decision vector is then matched to a threat level threshold to generate a graded alarm signal. The decision vector is evaluated based on the preset threat level threshold, triggering the corresponding alarm signal, achieving accurate identification of abnormal events and graded alarms.

[0092] Furthermore, based on the calibrated unified reference clock source, multi-source data of the target scene is collected and processed to generate a sensor data stream with a timestamp and an environmental interference parameter set, including:

[0093] Based on the calibrated unified reference clock source, dynamic synchronization trigger processing is performed on multi-source sensor nodes to generate clock-synchronized sensor acquisition instructions;

[0094] According to the predefined hierarchical acquisition strategy, the physical parameters and electromagnetic parameters of the target scene are collected and processed in a time-sharing manner to generate an environmental interference parameter set;

[0095] The time-space correlation verification process is performed on the multi-source data stream triggered by the clock-synchronized sensor acquisition instruction to generate the sensor data stream with time stamp and the environmental interference parameter set.

[0096] Specifically, a calibrated unified reference clock source is used to dynamically synchronize and trigger multi-source sensor nodes, generating clock-synchronized sensor acquisition instructions. This step ensures that all sensor nodes initiate data acquisition tasks based on the same time reference, ensuring temporal consistency for subsequent data fusion. Based on a predefined layered acquisition strategy, physical parameters (such as temperature, humidity, air pressure, and vibration) and electromagnetic parameters (such as electromagnetic signal strength and frequency) of the target scene are collected in a time-sharing manner to generate an environmental interference parameter set. This step ensures comprehensive and accurate data acquisition through layered and time-sharing acquisition. The multi-source data streams triggered by the clock synchronization instructions are verified for temporal and spatial correlation, generating a timestamped sensor data stream and an environmental interference parameter set. This step ensures data accuracy and reliability by verifying the temporal and spatial consistency of the data.

[0097] Furthermore, all sensor nodes are calibrated using a synchronous clock protocol to generate a calibrated unified reference clock source, including:

[0098] Based on the dynamic clock source selection strategy, the master clock nodes in the candidate clock source pool are evaluated for environmental interference resistance and the optimal master clock node is generated.

[0099] According to the clock signal of the optimal master clock node, the clock offset of the multi-source sensor nodes is calculated and processed in a consistent manner through a distributed consensus algorithm to generate a distributed consistent clock offset parameter;

[0100] Based on the distributed consistency clock offset parameter, the local clock of the sensor node is decentralized converged and adjusted to generate a calibrated unified reference clock source.

[0101] Specifically, the system collects and processes the environmental interference parameter set of candidate nodes in real time to generate a node interference feature vector. Based on this feature vector, the master clock node's signal stability is evaluated for interference resistance, generating a node stability score. Based on the dynamic ranking of the node stability scores, collaborative verification and switching of the master clock node is performed to determine the optimal clock node. Using the clock signal of the optimal master clock node as a benchmark, a distributed consensus algorithm is used to calculate the consistency of the clock offsets of all sensor nodes, generating a distributed consensus clock offset parameter. This step ensures that all nodes reach consensus on the clock offset, preventing the impact of single-point failures or local errors. Based on the calculated global clock offset parameter, a distributed convergence adjustment is performed on the local clock of each sensor node. Each node gradually adjusts its local clock according to the offset parameter, ultimately achieving global time consistency. After these steps, the local clocks of all sensor nodes are aligned with the optimal master clock node, generating a unified, calibrated reference clock source that ensures time synchronization for subsequent multi-source data acquisition and processing.

[0102] Furthermore, based on the dynamic clock source selection strategy, the master clock nodes in the candidate clock source pool are evaluated for environmental interference resistance to generate the optimal master clock node, including:

[0103] Perform real-time collection and processing of environmental interference parameter sets for the master clock nodes in the candidate clock source pool to generate node interference feature vectors.

[0104] Based on the node interference feature vector, the signal stability of the master clock node is processed for anti-interference scoring to generate a node stability score;

[0105] According to the dynamic ranking results of the node stability scores, the collaborative verification switching process of the master clock node is performed to generate the optimal master clock node.

[0106] Specifically, for each master clock node in the candidate clock source pool, environmental interference parameter sets (such as electromagnetic noise, temperature, and air pressure) are collected in real time to generate a node interference feature vector. This step uses a sensor network to monitor environmental parameters in real time, capturing the node's operating status in complex environments. Based on the node interference feature vector, the master clock node's signal stability is evaluated for interference resistance, generating a node stability score. This step quantifies the node's interference resistance in the current environment by analyzing metrics such as signal jitter, drift, and noise level. Based on the dynamic ranking of the node stability scores, collaborative verification and switching of the master clock node is performed. A distributed consensus algorithm (such as Raft or Paxos) is used to verify and confirm the optimal clock node across multiple nodes, ensuring that the selected master clock node has the highest global stability and interference resistance. Through these steps, an optimal master clock node is determined as the clock reference source. This ensures the continuous provision of high-precision time synchronization services in a dynamically changing environment.

[0107] In one embodiment, if Figure 2 As shown in the figure, based on the medium propagation delay prediction model and the calibrated unified reference clock source, the sensor data stream is dynamically compensated to generate a time-aligned multi-sensor dataset, including:

[0108] S201: Performing initial delay compensation processing on the sensor data stream based on a medium propagation delay prediction model to generate an initial compensated data stream;

[0109] S202: Generate a real-time error feedback parameter based on a real-time comparison result between the calibrated unified reference clock source and the medium propagation delay prediction model;

[0110] S203: Performing dynamic error correction processing on the initial compensation data stream based on the real-time error feedback parameter to generate a real-time correction data stream;

[0111] S204: Perform multi-channel time series integration processing on the real-time correction data stream to generate a time-aligned multi-sensor data set.

[0112] Specifically, the medium propagation delay prediction model is used to perform initial delay compensation on the sensor data stream to generate an initial compensated data stream. This step estimates the signal propagation delay using the prediction model and makes preliminary adjustments to the data stream's timestamps. Real-time error feedback parameters are generated based on the real-time comparison results between the calibrated unified reference clock source and the medium propagation delay prediction model. This step quantifies the compensation error by monitoring the difference between actual and predicted time. Based on the real-time error feedback parameters, dynamic error correction is performed on the initial compensated data stream to generate a real-time corrected data stream. This step reduces error and improves time alignment accuracy by adjusting the compensation model parameters. Multi-channel time series integration is performed on the real-time corrected data stream to generate a time-aligned multi-sensor dataset. This step ensures that data from different sensors are fused based on a unified time base through methods such as interpolation or extrapolation.

[0113] It should be understood that, although the various steps in the flowcharts involved in the various embodiments described above are displayed in sequence according to the instructions of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the flowcharts involved in the various embodiments described above can include multiple steps or multiple stages, and these steps or stages are not necessarily executed and completed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of steps or stages in other steps.

[0114] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to the partial description of the method embodiments. The device embodiments described above are merely illustrative, wherein the components described as separate parts may or may not be physically separated, and the parts displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the disclosed solution. A person of ordinary skill in the art can understand and implement it without expending creative work.

[0115] The above-described embodiments merely represent several implementation methods of the embodiments of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that a person skilled in the art may make various modifications and improvements without departing from the concept of the embodiments of the present application, and these modifications and improvements fall within the scope of protection of the embodiments of the present application.

Claims

1. An emergency alarm system, characterized in that: The system comprises: The clock calibration module is used to perform clock calibration on all sensor nodes through the synchronous clock protocol to generate a calibrated unified reference clock source; A multi-source data acquisition module is used to collect and process multi-source data of the target scene based on the calibrated unified reference clock source to generate a sensor data stream with a timestamp and an environmental interference parameter set; A delay prediction modeling module is used to build a medium propagation delay prediction model based on the environmental interference parameter set and the sensor physical characteristic parameters; a dynamic compensation alignment module, configured to perform dynamic compensation processing on the sensor data stream based on the medium propagation delay prediction model and the calibrated unified reference clock source to generate a time-aligned multi-sensor data set; a multimodal alarm triggering module, configured to perform multimodal fusion analysis on the time-aligned multi-sensor data set and trigger a graded alarm signal; The method of constructing a medium propagation delay prediction model based on the environmental interference parameter set and the sensor physical characteristic parameters includes: performing feature extraction processing on the humidity parameter and the air pressure parameter in the environmental interference parameter set to generate an environmental propagation influence factor; performing correlation analysis processing on the signal delay coefficient and the thermal stability coefficient in the sensor physical characteristic parameters to generate a sensor inherent delay feature; and performing multi-dimensional fusion processing on the environmental propagation influence factor and the sensor inherent delay feature to generate the medium propagation delay prediction model. The method of collecting and processing multi-source data of the target scene based on the calibrated unified reference clock source to generate a sensor data stream with a timestamp and an environmental interference parameter set includes: based on the calibrated unified reference clock source, dynamically synchronizing and triggering multi-source sensor nodes to generate clock-synchronized sensor collection instructions; collecting and processing the physical parameters and electromagnetic parameters of the target scene in a time-sharing manner according to a predefined layered collection strategy to generate an environmental interference parameter set; and performing spatiotemporal correlation verification on the multi-source data stream triggered by the clock-synchronized sensor collection instructions to generate the sensor data stream with a timestamp and the environmental interference parameter set.

2. An emergency alarm system according to claim 1, characterized in that: The performing multi-dimensional fusion processing on the environmental propagation influencing factors and the inherent delay characteristics of the sensor to generate the medium propagation delay prediction model includes: Performing time alignment processing on the environmental propagation influencing factor and the inherent delay characteristic of the sensor to generate a spatiotemporal consistency feature vector; Based on a dynamic environment weight allocation strategy, the environmental propagation influencing factors in the spatiotemporal consistency feature vector and the inherent delay characteristics of the sensor are adaptively weighted and fused to generate a hybrid delay feature matrix; Nonlinear mapping processing is performed on the mixed delay characteristic matrix to generate the medium propagation delay prediction model.

3. An emergency alarm system according to claim 2, characterized in that: The method of adaptively weighting and fusing the environmental propagation influencing factors and the inherent delay characteristics of the sensor in the spatiotemporal consistency feature vector based on the dynamic environmental weight allocation strategy to generate a hybrid delay feature matrix includes: Performing interference intensity grading processing on the environmental propagation impact factors in the spatiotemporal consistency feature vector to generate an environmental interference intensity level; performing degradation sensitivity evaluation processing on the sensor inherent delay feature in the spatiotemporal consistency feature vector to generate a sensor degradation level; Based on the real-time dynamic relationship between the environmental interference intensity level and the sensor degradation level, a fusion weight of the environmental propagation impact factor and the sensor inherent delay characteristic is allocated to generate the hybrid delay characteristic matrix.

4. An emergency alarm system according to claim 1, characterized in that: The performing multimodal fusion analysis on the time-aligned multi-sensor data set to trigger a graded alarm signal includes: performing spatiotemporal feature enhancement processing on the time-aligned multi-sensor dataset to generate a multimodal feature matrix; Based on the conflict resolution rules between sensors, the multimodal feature matrix is ​​subjected to logical priority processing to generate conflict-resolved fusion data; According to the dynamic weight of the environmental interference and the sensor health status score, the fused data after the conflict resolution is adaptively weighted and decided to generate the graded alarm signal.

5. An emergency alarm system according to claim 4, characterized in that: The adaptive weighted decision-making process is performed on the conflict-resolved fusion data according to the dynamic weight of the environmental interference and the sensor health status score to generate the graded alarm signal, including: Performing interference intensity quantification processing on the environmental interference dynamic weight to generate an environmental interference intensity level; Performing degradation tolerance mapping processing on the sensor health status score to generate a sensor credibility coefficient; Based on the dynamic coupling relationship between the environmental interference intensity level and the sensor credibility coefficient, performing real-time weight distribution processing on the fused data after conflict resolution to generate a dynamic weighted decision vector; The dynamic weighted decision vector is subjected to threat level threshold matching processing to generate the graded alarm signal.

6. An emergency alarm system according to claim 1, characterized in that: The clock calibration process is performed on all sensor nodes through the synchronous clock protocol to generate a calibrated unified reference clock source, including: Based on the dynamic clock source selection strategy, the master clock nodes in the candidate clock source pool are evaluated for environmental interference resistance and the optimal master clock node is generated. According to the clock signal of the optimal master clock node, the clock offsets of the multi-source sensor nodes are calculated and processed in a consistent manner through a distributed consensus algorithm to generate a distributed consistent clock offset parameter; A decentralized convergence adjustment process is performed on the local clock of the sensor node based on the distributed consistency clock offset parameter to generate the calibrated unified reference clock source.

7. An emergency alarm system according to claim 6, characterized in that: The method of performing environmental interference resistance evaluation on the master clock nodes in the candidate clock source pool based on the dynamic clock source selection strategy to generate the optimal master clock node includes: Performing real-time collection and processing of an environmental interference parameter set on a master clock node in the candidate clock source pool to generate a node interference feature vector; Based on the node interference feature vector, performing anti-interference scoring processing on the signal stability of the master clock node to generate a node stability score; According to the dynamic ranking result of the node stability score, the collaborative verification switching process of the master clock node is performed to generate the optimal master clock node.

8. An emergency alarm system according to claim 1, characterized in that: The method of performing dynamic compensation processing on the sensor data stream based on the medium propagation delay prediction model and the calibrated unified reference clock source to generate a time-aligned multi-sensor data set includes: Performing initial delay compensation processing on the sensor data stream based on the medium propagation delay prediction model to generate an initial compensated data stream; generating a real-time error feedback parameter according to a real-time comparison result between the calibrated unified reference clock source and the medium propagation delay prediction model; Based on the real-time error feedback parameter, dynamic error correction processing is performed on the initial compensation data stream to generate a real-time correction data stream; Multi-channel time series integration processing is performed on the real-time corrected data stream to generate the time-aligned multi-sensor data set.

Images