Fire-fighting remote control device based on Internet of Things technology

Abstract

The invention relates to the technical field of fire-fighting informatization, in particular to a fire-fighting remote control device based on the Internet of Things technology, and the device comprises a distributed abnormal signal monitoring module which collects non-alarm data such as communication abnormality, response timeout and voltage fluctuation; the abnormal feature decoding and risk identification module extracts features such as signal drift and delay vectors and generates risk mode vectors; the equipment layer stability prediction module predicts the stability level and the fault type in the future 30 minutes in combination with historical data and environmental parameters; the linkage consistency evaluation module aggregates prediction results of all the nodes and identifies potential consistency risks under a communication link or a power bus; and the intelligent disposal strategy generation module outputs a disposal suggestion package based on the consistency report. According to the invention, early recognition, trend prediction and cross-node linkage disposal of the fire control system on non-alarm faults are realized, and the system stability and the remote control intelligence level are significantly improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of fire-fighting information technology, and in particular to a fire-fighting remote control device based on Internet of Things technology. BACKGROUND

[0002] With the in-depth application of Internet of Things technology in the field of fire-fighting, more and more fire-fighting control systems have functions of remote monitoring, centralized management and data linkage, effectively improving the fire warning and disaster response capability. In modern buildings and industrial scenes, fire-fighting equipment is generally distributed in multiple functional areas, forming a distributed control structure with the help of communication links and power supply buses. Traditional fire-fighting systems mainly rely on alarm mechanisms of explicit abnormalities such as smoke, temperature or open fire. However, with the complication of system structure and diversification of abnormal performance, higher requirements are put forward for the identification and linkage disposal of hidden dangers in non-alarm states such as equipment micro-failure, communication anomaly and power disturbance.

[0003] The prior art still has obvious limitations in abnormal monitoring and remote control. First, most systems mainly rely on whether an alarm is triggered as the main judgment basis, lacking the ability to collect and analyze implicit indicators such as communication signal stability, keyboard response delay, voltage fluctuation and capacitor breakdown precursor, which easily misses early signs of failure. Second, the device state analysis mostly uses simple threshold judgment, which cannot realize trend prediction or structured analysis of fault evolution, and cannot perform consistent risk research and grouping linkage management of multiple nodes. Finally, the remote disposal strategy generally uses static plans, which do not have the ability to dynamically match response templates according to actual risk types, making it difficult to achieve flexible response and resource optimization in complex systems. SUMMARY

[0004] The present application provides a fire-fighting remote control device based on Internet of Things technology, which can effectively make up for the shortcomings of existing technologies in non-alarm fault identification, short-cycle prediction analysis, node group risk aggregation and dynamic template response, significantly improving the stability and emergency response efficiency of fire-fighting systems under multi-point deployment, complex communication and power structure.

[0005] A fire-fighting remote control device based on Internet of Things technology, comprising a distributed abnormal signal monitoring module, an abnormal feature decoding and risk pattern recognition module, a device layer stability prediction module, a cross-node linkage consistency evaluation module and an intelligent disposal strategy generation module, wherein;

[0006] The distributed abnormal signal monitoring module is deployed at each fire-fighting control node for real-time collection of various non-alarm abnormal signals, including control board communication anomaly, keyboard response timeout, power supply voltage fluctuation and capacitor weak breakdown precursor feature, generating multi-source abnormal data;

[0007] The abnormal feature decoding and risk pattern recognition module receives multi-source abnormal data, decodes the multi-source abnormal data based on an abnormal feature library and a micro-fault evolution model, extracts abnormal features including signal drift amplitude, response delay vector, and voltage disturbance spectrum, and generates a structured risk pattern vector;

[0008] The device layer stability prediction module receives the structured risk pattern vector, combines historical running states of the target device and background parameters of the current sampling period, predicts the stability level and potential fault type of the device in a future predetermined monitoring period (30 minutes), and outputs the prediction result and a confidence label;

[0009] The cross-node linkage consistency evaluation module aggregates and analyzes the stability prediction results of multiple fire-fighting nodes, identifies potential consistency risks existing under the same communication link or power bus, and generates a node group linkage consistency report;

[0010] The intelligent disposal strategy generation module dynamically matches a preset risk disposal template based on the node group linkage consistency report, and generates an intelligent disposal suggestion package including early warning levels, recommended operation windows, communication isolation ranges, and permission control strategies.

[0011] Optionally, the distributed abnormal signal monitoring module includes:

[0012] Monitoring node deployment and initialization configuration: an embedded monitoring unit is integrated on each fire control node, and the monitoring unit completes initialization configuration of monitoring parameters including data sampling frequency setting, timestamp synchronization, and channel mapping identification configuration through access to a control board interface, a keyboard response line, a power input terminal, and a capacitor monitoring point;

[0013] Multi-source abnormal signal real-time acquisition: set the sampling frequency f s , record the number of communication success frames N succ (t) and the total number of sent frames N total (t) in real time, and count the number of frame losses N lost (t) = N total (t) - N succ (t) in a unit sampling period, listen to user key operation events and control board response signals, calculate the response delay time T key (t) for each time, and sample the power input voltage sequence S vol (t) in real time, record the capacitor leakage current sequence S cap( t), and adjacent sampling values S cap (t-δ), and form a multi-channel data structure body D(t) = {N lost (t), N total (t), T key (t), S vol (t), Scap (t)};

[0014] Abnormal feature discrimination: Based on the constructed data structure D(t), perform abnormal feature discrimination. For the communication channel, calculate the communication frame loss rate P per unit time. loss (t), if P loss (t) exceeds the communication channel threshold θ com , it is marked as communication abnormality. For the keyboard channel, its response time T is determined. key (t) Whether it exceeds the keyboard channel threshold θ key , to identify response timeout events, for voltage data, calculate the voltage standard deviation σ vol , if σ vol Exceeding voltage θ vol , then it is a voltage anomaly. For the capacitance monitoring data, calculate the current change rate per unit time ΔI cap (t), if ΔI cap (t) exceeds the capacitance threshold θ cap , it is considered as a precursor to breakdown and is a capacitor abnormality;

[0015] Structured output of multi-source anomaly data: All multi-source anomaly signals that meet the anomaly feature discrimination conditions are labeled separately and structured anomaly data entries are generated, which are then packaged into multi-source anomaly data.

[0016] Optionally, the abnormal feature decoding and risk pattern recognition module includes:

[0017] Abnormal feature decoding: Standardizes the received multi-source abnormal data and uses preset rules based on the abnormality type to extract key feature information, including signal drift amplitude in communication anomalies, delay vectors in keyboard responses, and disturbance spectrum features in voltage data. This converts the multi-source abnormal data into a set of feature vectors.

[0018] Risk pattern recognition: Analyze the feature vector set and, combined with the preset micro-fault evolution model, identify the corresponding risk pattern type and output a structured risk pattern vector including the fault prediction category, risk level, and confidence score.

[0019] Optionally, the abnormal feature decoding includes:

[0020] Multi-source abnormal data preprocessing and channel normalization: Perform classification processing, extract communication, keyboard and voltage data channels respectively, perform channel normalization, and construct a standardized channel data sequence set in, is the normalized communication status time series, is the normalized response delay sequence, is the normalized voltage waveform sequence;

[0021] Communication anomaly feature extraction: Based on Calculate the signal drift amplitude to describe the communication stability trend;

[0022] Keyboard response feature extraction: based on Extract the response delay vector and construct the delay mean and delay standard deviation of the response time;

[0023] Voltage disturbance spectrum feature extraction: Perform fast Fourier transform (FFT) to extract the voltage disturbance spectrum amplitude vector;

[0024] Feature vector concatenation: The extracted signal drift amplitude, delay mean, delay standard deviation, and voltage disturbance spectrum amplitude vectors are concatenated to form a feature vector set X.

[0025] Optionally, the risk pattern identification includes:

[0026] Fault mode similarity calculation: the feature vector set X and the fault template library in the micro-fault evolution model Each template M in k Calculate the Euclidean distance S k ;

[0027] Fault category matching and confidence scoring: Determine the best matching template index k * , and calculate the matching confidence score C of the current risk pattern risk ;

[0028] Risk level classification: Based on matching confidence score C risk Risk levels are divided into high risk, medium risk and low risk, which are expressed as:

[0029]

[0030] Structured risk pattern vector generation: The identified fault prediction categories (formulated by template The risk level RiskLevel and the matching confidence score C risk Integrate into a structured risk pattern vector R.

[0031] Optionally, the device layer stability prediction module includes:

[0032] State fusion: Receives structured risk pattern vectors and combines them with historical operating data of the target device and current environmental parameters, including operating time, historical fault frequency, temperature and humidity, and power supply status level, to generate a comprehensive device state vector.

[0033] Stability prediction: based on the device state comprehensive vector, predict the stability level of the device in the future predetermined monitoring period (30 minutes), including the level labels of stable operation, slight abnormal trend, high-risk failure tendency, while identifying the corresponding potential failure type, and output the prediction result with confidence label.

[0034] Optionally, the state fusion includes:

[0035] Data source receiving and standardization processing: receiving the structured risk mode vector R, synchronously calling the historical operation data and current environment parameters of the target device, including the operation time T run , historical failure frequency N fault , temperature T env , humidity H env , power supply state level G power , and performing normalization operation on the historical operation data and current environment parameters to generate a standardized feature set

[0036] Feature weighted fusion: using a linear weighted model to weight and integrate the standardized feature set and the risk level and confidence score in the structured risk mode vector to generate a device state comprehensive vector Z.

[0037] Optionally, the stability prediction includes:

[0038] Stability score calculation: using a prediction function to infer the device state comprehensive vector Z to obtain the stability score value S stab of the device in the future 30 minutes.

[0039] Stability level classification and failure type identification: according to the interval where the stability score S stab is located, divide it into stability level StabLevel, including stable operation, slight abnormal trend, high-risk failure tendency, and identify the corresponding potential failure type FaultCat, represented as:

[0040]

[0041] Prediction result packaging output: encapsulate the stability level StabLevel, potential failure type FaultCat and confidence label P conf =S stab together, and output the structured prediction result P result .

[0042] Optionally, the cross-node linkage consistency evaluation module includes:

[0043] Stability prediction result summary: collect the structured prediction results generated by multiple fire-fighting nodes in the same monitoring period The construction result set And through the device configuration mapping table, the communication link identification T corresponding to each node is obtained i With the power bus identification M i , and the structural completion of the prediction result is represented as:

[0044]

[0045] Among them, is the total number of fire nodes participating in the linkage evaluation in the current monitoring period, is the prediction result after structural completion;

[0046] Node grouping according to link and bus: based on Link identification T i and bus identification M i , the nodes are divided into multiple groups G t ={i|T i =t}, G m ={i|M i =m}, respectively forming a communication link group set and a power supply bus group set

[0047] Consistency risk score calculation: consistency analysis is performed on the prediction result in each group G x , and the consistency risk score R x is calculated, if R x ≥0.7, it is determined that the group G x is a high-consistency-risk node group, if 0.4≤R x <0.7, it is determined that the group G x is a moderate-consistency-risk node group, if R x <0.4, it is determined that the group G x is a low-consistency-risk node group;

[0048] Generating a linkage consistency report: for each high-consistency-risk node group, the FaultCat i with the highest frequency of occurrence in it is taken as the fault mode, and a consistency report R group is generated.

[0049] Optionally, the intelligent disposal strategy generation module comprises:

[0050] Risk disposal template dynamic matching: based on the consistency report R groupIf the grouping type given in the consistency report is a communication link, and the dominant fault type is abnormal drop-off, a communication isolation type template is matched, and if the dominant fault type is power disturbance, a delay intervention and frequent monitoring type template is matched;

[0051] Generating an intelligent treatment suggestion package: according to the matching result of the risk treatment template, an intelligent treatment suggestion package is generated, including a warning level (such as a first-level warning, a second-level warning), a suggestion operation window (immediate processing, priority intervention within 5 minutes), a communication isolation range (identifying nodes or link numbers that need to be temporarily logically isolated), and a permission control strategy (including whether to limit remote writing or adjust the control permission level).

[0052] The beneficial effects of the present application are:

[0053] The present application significantly improves the pre-identification ability of the fire remote control system for potential hazards by constructing a fusion perception mechanism for multiple source non-alarm abnormal signals. The existing system relies on an alarm threshold triggering mechanism, which is difficult to respond to weak fault signs such as communication anomalies and power fluctuations, resulting in a lack of perception ability for slow-changing abnormalities or early faults. The distributed embedded monitoring module can collect micro-feature signals such as communication frame loss rate, response delay, and capacitor breakdown trend under the non-alarm state of the device, effectively constructs structured abnormal data across nodes and multiple channels through channel normalization and standardization processing.

[0054] The present application realizes short-period trend prediction and multi-dimensional risk decoding of the fire equipment operating state by introducing a structured risk mode vector and a stability prediction score mechanism. Unlike traditional single-point discrimination logic, the micro-fault evolution model matches the feature vector and risk template, quantifies the current risk state of the system through matching confidence and risk level, and constructs a state comprehensive vector by fusing historical operation data of the device and current environmental parameters. Combined with linear weighting and deep learning prediction function, the device stability level and potential fault type within the next 30 minutes can be accurately output, effectively making up for the shortcomings of existing systems in trend analysis and early treatment, and improving the response foresight and decision basis reliability of the system for heterogeneous faults.

[0055] The present application identifies a high-consistency risk node group through consistent aggregation analysis of the prediction results of multiple nodes under the communication link or power bus, and dynamically matches the communication isolation type or delay intervention type strategy template according to the dominant fault type. The finally output treatment suggestion package clearly gives the warning level, operation window, communication isolation range and permission control strategy, realizes closed-loop control from data perception, intelligent analysis to accurate treatment, effectively deals with low-frequency high-risk scenarios such as communication failure and power disturbance, and provides a new type of solution with forward-looking perception and intelligent cooperation for fire remote control in the Internet of Things environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0057] Fig. 1 Schematic diagram of the functional modules of the device according to an embodiment of the present invention;

[0058] Fig. 2 Schematic diagram of a cross-node linkage consistency assessment module according to an embodiment of the present invention. DETAILED DESCRIPTION

[0059] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. Those skilled in the art may also adopt other alternatives to implement some known technologies; and the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0060] like Figs. 1-2 As shown, a fire remote control device based on Internet of Things technology includes a distributed abnormal signal monitoring module, an abnormal feature decoding and risk pattern recognition module, an equipment layer stability prediction module, a cross-node linkage consistency assessment module and an intelligent disposal strategy generation module, wherein;

[0061] The distributed abnormal signal monitoring module is deployed at each fire control node to collect various non-alarm abnormal signals in real time, including control panel communication anomalies, keyboard response timeouts, power supply voltage fluctuations, and capacitor weak breakdown precursor characteristics, generating multi-source abnormal data.

[0062] The abnormal feature decoding and risk pattern recognition module receives multi-source abnormal data, decodes the multi-source abnormal data based on the abnormal feature library and micro-fault evolution model, extracts abnormal features including signal drift amplitude, response delay vector, and voltage disturbance spectrum, and generates a structured risk pattern vector.

[0063] The device-level stability prediction module receives the structured risk pattern vector, combines the historical operating status of the target device with the background parameters of the current sampling period, predicts the stability level and potential fault type of the device within the next scheduled monitoring period (30 minutes), and outputs the prediction result and credibility label;

[0064] The cross-node linkage consistency assessment module aggregates and analyzes the stability prediction results between multiple fire protection nodes, identifies potential consistency risks under the same communication link or power bus, and generates a node group linkage consistency report;

[0065] The intelligent disposal strategy generation module generates an intelligent disposal suggestion package including early warning level, suggested operation window, communication isolation range and permission control strategy based on the node group linkage consistency report, dynamically matches the preset risk disposal template.

[0066] The distributed abnormal signal monitoring module comprises:

[0067] The monitoring node deployment and initialization configuration: an embedded monitoring unit is integrated on each fire control node, the monitoring unit completes the initialization configuration of monitoring parameters including data sampling frequency setting, timestamp synchronization and channel mapping identifier configuration through access to the control board interface, keyboard response line, power input end and capacitance monitoring point;

[0068] Multi-source abnormal signal real-time acquisition: set the sampling frequency f s , record the number of communication success frames N succ (t) and the total number of sent frames N total (t) in real time, and count the number of frame loss N lost (t) = N total (t) - N succ (t), listen to user key operation events and control board response signals, calculate the response delay time T key (t) for each time, sample the power input voltage sequence S vol (t) in real time, record the capacitance leakage current sequence S cap (t) and the adjacent sampling value S cap (t-δ), and form a multi-channel data structure body D(t) = {N lost (t), N total (t), T key (t), S vol (t), S cap (t)};

[0069] Abnormal feature discrimination: based on the constructed data structure body D(t), execute abnormal feature discrimination, for the communication channel, calculate the communication frame loss rate P loss (t) in unit time, if P loss (t) exceeds the communication channel threshold θ com , it is marked as communication abnormality, for the keyboard channel, judge whether its response time T key (t) exceeds the keyboard channel threshold θ key to identify the response timeout event, for the voltage data, calculate the voltage standard deviation σ vol , if σ vol exceeds the voltage θ vol , it is voltage abnormality, for the capacitance monitoring data, calculate the current change rate ΔI cap (t) in unit time, if ΔIcap (t) exceeds the capacitance threshold θ cap If so, it is considered as a precursor of breakdown, and is denoted as:

[0070]

[0071] σ vol = StdDev(S vol (t));

[0072]

[0073] wherein StdDev is a standard deviation calculation function performed on an input set of numerical values;

[0074] Multi-source anomaly data structured output: all multi-source anomaly signals meeting the abnormal feature discrimination condition are respectively labeled and structured anomaly data entries are generated, and are packaged as multi-source anomaly data.

[0075] Abnormal feature decoding and risk pattern recognition module includes:

[0076] Abnormal feature decoding: standardizing the received multi-source anomaly data, calling preset rules to extract key feature information according to the type of anomaly, including signal drift amplitude in communication anomaly, delay vector in keyboard response and disturbance frequency spectrum feature in voltage data, so as to convert multi-source anomaly data into a feature vector set;

[0077] Risk pattern recognition: analyzing the feature vector set, combining the preset micro-fault evolution model, identifying the corresponding risk pattern type, and outputting a structured risk pattern vector including fault prediction category, risk level and confidence score.

[0078] Abnormal feature decoding includes:

[0079] Multi-source anomaly data preprocessing and channel normalization: the received multi-source anomaly data are classified and processed, communication, keyboard and voltage data channels are extracted respectively, channel normalization is performed, time granularity, amplitude unit of different data sources are uniformly processed, and a standardized channel data sequence set is constructed wherein, is the normalized communication state time sequence, is the normalized response delay sequence, is the normalized voltage waveform sequence;

[0080] Communication anomaly feature extraction: based on the signal drift amplitude is calculated to describe the communication stability trend, and the signal drift amount is represented as:

[0081] Dcom (t) = |μ com (t) - μ com (t - Δt) |;

[0082] where D com (t) is the communication signal drift amplitude in the current time window, μ com (t) is the mean of the communication efficiency in the current window, and Δt is the window interval time.

[0083] Keyboard response feature extraction: based on The response delay vector is extracted, and the delay mean and delay standard deviation of the response time are constructed, denoted as:

[0084]

[0085] wherein, is the response delay mean, σ key is the response delay standard deviation, and V key is the keyboard response delay vector.

[0086] Voltage disturbance spectrum feature extraction: perform fast Fourier transform (FFT) on The voltage disturbance spectrum amplitude vector is extracted, denoted as:

[0087]

[0088] wherein, is the voltage value of the nth sampling point in the normalized voltage sequence, N is the total number of sampling points, k is the frequency index, F vol is the voltage disturbance spectrum amplitude vector with length n, and j is the imaginary unit.

[0089] Feature vector splicing: the extracted signal drift amplitude, delay mean, delay standard deviation, and voltage disturbance spectrum amplitude vector are spliced to form a feature vector set X, denoted as:

[0090]

[0091] Risk pattern recognition includes:

[0092] Fault mode similarity calculation: for each template M k in the feature vector set X and the fault template library in the micro-fault evolution model, k the Euclidean distance S is calculated, denoted as:

[0093]

[0094] wherein, x i is the ith component of the feature vector set, and m k,iS k is the i-th component of the template M k is the distance to the k-th failure template, d is the vector dimension;

[0095] Failure class matching and confidence score: determine the best matching template index k * and calculate the matching confidence score C risk of the current risk pattern, denoted as:

[0096] k * = argmin k S k ;

[0097]

[0098] where K is the total number of preset micro-failure pattern templates;

[0099] Risk level classification: classify the risk level according to the matching confidence score C risk , including high risk, medium risk, and low risk, denoted as:

[0100]

[0101] Structured risk pattern vector generation: integrate the identified failure prediction class (given by the label of the template ), risk level RiskLevel, and matching confidence score C risk into a structured risk pattern vector R, denoted as:

[0102]

[0103] where is the failure type represented by the most similar risk pattern.

[0104] The device layer stability prediction module includes:

[0105] State fusion: receive the structured risk pattern vector and combine the historical operation data of the target device and the current environmental parameters, including the running time, historical failure frequency, temperature and humidity, power supply state level, to generate a device state comprehensive vector;

[0106] Stability prediction: based on the device state comprehensive vector, predict the stability level of the device within a predetermined monitoring period (30 minutes), including stable operation, slight abnormal trend, high risk failure tendency level labels, and identify the corresponding potential failure type, and output the prediction result with confidence label.

[0107] State fusion includes:

[0108] Data source receiving and standardization processing: receiving structured risk pattern vector R, synchronously calling historical running data and current environment parameters of target device, including running time T run , historical failure frequency N fault , temperature T env , humidity H env , power supply state level G power , and performing normalization operation on historical running data and current environment parameters to generate standardized feature set

[0109] Feature weighting fusion: using linear weighting model to weight and integrate standardized feature set and risk level and confidence score in structured risk pattern vector, generating device state comprehensive vector Z, expressed as:

[0110]

[0111] Where, f i is the i-th standardized feature, α i is the feature weight corresponding to f i , L risk is the risk level mapping value (high risk = 1.0, medium risk = 0.6, low risk = 0.3), β1 and β2 are the fusion weights of risk level and confidence.

[0112] Stability prediction includes:

[0113] Stability score calculation: using prediction function to infer device state comprehensive vector Z to obtain stability score value S stab of the device in the next 30 minutes, expressed as:

[0114] S stab = f pred (Z) = W·Z T + b;

[0115] Where, W is the weight vector of the prediction function, and b is the bias term;

[0116] Stability level classification and fault type identification: according to the interval of stability score S stab , it is divided into stability level StabLevel, including stable operation, slight abnormal trend, high risk fault tendency, and the corresponding potential fault type FaultCat is identified, expressed as:

[0117]

[0118] FaultCat = argmax j∈{1,2,...,C} (P j );

[0119]

[0120] where C is the total number of fault categories, P j is the probability output of the prediction function for the jth fault category, S stab,j is the raw output score of the prediction function for the jth fault category or stability level, denotes the normalization factor;

[0121] Prediction result packaging output: The stability level StabLevel, the potential fault type FaultCat and the confidence label P conf = S stab are jointly packaged, outputting the structured prediction result P result , which is expressed as:

[0122] P result = <StabLevel, FaultCat, P conf >;

[0123] where P result is the confidence score, and the closer the value is to 0 or 1, the more credible it is.

[0124] The cross-node linkage consistency evaluation module includes:

[0125] Stability prediction result summary: Collect the structured prediction results generated by multiple fire-fighting nodes in the same monitoring period Construct a result set and obtain the communication link identifier T i and the power bus identifier M i of each node through the device configuration mapping table, and perform structure completion on the prediction results, which is expressed as:

[0126]

[0127] where, is the total number of fire-fighting nodes participating in the linkage evaluation in the current monitoring period, is the prediction result after structure completion;

[0128] Node grouping according to links and buses: Based on the link identifier T i and the bus identifier M i in the , the nodes are divided into multiple groups G t = {i | T i = t}, G m = {i | M i = m}, respectively forming a communication link group set and a power supply bus group set

[0129] Consistency risk score calculation: For each group G x The prediction results were analyzed for consistency and the consistency risk score R was calculated. x , if R x ≥0.7, then group G is determined x For a high-consistency risk node group, if 0.4≤R x <0.7, then group G is determined x is a moderate consistency risk node group, if R x <0.4, then group G is determined x is a low consistency risk node group, expressed as:

[0130]

[0131] Among them, |G x | is the number of nodes in the group, is an indicator function, which is 1 when the condition is met, otherwise it is 0;

[0132] Generate linkage consistency report: For each high consistency risk node group, count the FaultCat with the highest frequency i As a failure mode, generate a consistency report R group .

[0133] The intelligent disposal strategy generation module includes:

[0134] Dynamic matching of risk treatment templates: Based on consistency report R group , the matching template is retrieved from the preset risk treatment template library. If the group type given in the consistency report is communication link and the dominant fault type is abnormal disconnection, the communication isolation template is matched. If the dominant fault type is power disturbance, the delayed intervention and frequent monitoring templates are matched;

[0135] Generate intelligent disposal suggestion packages: Based on the results of dynamic matching of risk disposal templates, generate intelligent disposal suggestion packages, including warning levels (such as Level I warning, Level II warning), recommended operation windows (immediate processing, priority intervention within 5 minutes), communication isolation scope (identifying the node or link number that requires temporary logical isolation), and permission control policies (including whether to restrict remote writing or adjust the control permission level);

[0136] Communication isolation templates (used for communication link disconnection faults) include:

[0137] Include all nodes in the group in the communication isolation list;

[0138] Limit the issuance of remote commands to isolated nodes;

[0139] Start the link health diagnosis and loopback test module;

[0140] The delay intervention and frequent monitoring type template (for power disturbance or moderate risk) includes:

[0141] Issue a level II early warning;

[0142] Set the observation window to "5-10 minutes";

[0143] Start the timing heartbeat sampling and voltage fluctuation tracking mechanism;

[0144] Allow partial low-privilege remote operation, but open the early warning prompt pop-up window.

[0145] The present application encompasses any substitutions, modifications, equivalent methods and solutions made to the essence and scope of the present application. In order for the public to have a thorough understanding of the present application, specific details are described in the following preferred embodiments of the present application, and the present application can also be fully understood without the description of these details by those skilled in the art. In addition, in order to avoid unnecessary confusion to the essence of the present application, well-known methods, processes, procedures, elements and circuits, etc. are not described in detail.

[0146] The above is only the preferred embodiment of the present application, and it should be pointed out that for ordinary skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, and these improvements and refinements should be considered as the protection scope of the present application.

Claims

1. A fire remote control device based on Internet of Things technology, characterized in that: It includes a distributed abnormal signal monitoring module, an abnormal feature decoding and risk pattern recognition module, an equipment layer stability prediction module, a cross-node linkage consistency assessment module, and an intelligent disposal strategy generation module, among which; The distributed abnormal signal monitoring module is deployed at each fire control node to collect various non-alarm abnormal signals in real time, including control panel communication abnormalities, keyboard response timeouts, power supply voltage fluctuations, and capacitor weak breakdown precursor characteristics, and generate multi-source abnormal data; The abnormal feature decoding and risk pattern recognition module receives multi-source abnormal data, decodes the multi-source abnormal data based on the abnormal feature library and the micro-fault evolution model, extracts abnormal features including signal drift amplitude, response delay vector, and voltage disturbance spectrum, and generates a structured risk pattern vector; The device layer stability prediction module receives the structured risk pattern vector, combines the historical operating status of the target device with the background parameters of the current sampling period, predicts the stability level and potential fault type of the device within the future scheduled monitoring period, and outputs the prediction result and credibility label; The cross-node linkage consistency assessment module aggregates and analyzes the stability prediction results between multiple fire protection nodes, identifies potential consistency risks under the same communication link or power bus, and generates a node group linkage consistency report; The intelligent disposal strategy generation module dynamically matches the preset risk disposal template based on the node group linkage consistency report to generate an intelligent disposal suggestion package including the warning level, recommended operation window, communication isolation range and permission control strategy.

2. The fire remote control device based on Internet of Things technology according to claim 1, characterized in that: The distributed abnormal signal monitoring module includes: Monitoring node deployment and initial configuration: An embedded monitoring unit is integrated into each fire control node. The monitoring unit completes the initial configuration of monitoring parameters by accessing the control panel interface, keyboard response line, power input terminal and capacitor monitoring point, including data sampling frequency setting, timestamp synchronization and channel mapping identification configuration; Real-time acquisition of multi-source abnormal signals: set the sampling frequency f s , real-time record of the number of successful communication frames N succ (t) and the total number of frames sent N total (t), the number of frames lost in the statistical unit sampling period N lost (t) = N lost (t)-N succ (t), monitor user key operation events and control panel response signals, and calculate each response delay time T key (t), real-time sampling of the power input voltage sequence S vol (t), record the capacitor leakage current sequence S cap (t) and adjacent sample values ​​S cap (t-δ), and form a multi-channel data structure D(t)={N lost (t),N total (t),T key (t),S vol (t),S cap (t)}; Abnormal feature discrimination: Based on the constructed data structure D(t), perform abnormal feature discrimination. For the communication channel, calculate the communication frame loss rate P per unit time. loss (t), if P loss (t) exceeds the communication channel threshold θ com , it is marked as communication abnormality. For the keyboard channel, its response time T is determined. key (t) Whether it exceeds the keyboard channel threshold θ key , to identify response timeout events, for voltage data, calculate the voltage standard deviation σ vol , if σ vol Exceeding voltage θ vol , then it is a voltage anomaly. For the capacitance monitoring data, calculate the current change rate per unit time ΔI cap (t), if ΔI cap (t) exceeds the capacitance threshold θ cap , it is considered as a precursor to breakdown and is a capacitor abnormality; Structured output of multi-source anomaly data: All multi-source anomaly signals that meet the anomaly feature discrimination conditions are labeled separately and structured anomaly data entries are generated, which are then packaged into multi-source anomaly data.

3. The fire remote control device based on Internet of Things technology according to claim 1, characterized in that: The abnormal feature decoding and risk pattern recognition module includes: Abnormal feature decoding: Standardizes the received multi-source abnormal data and uses preset rules based on the abnormality type to extract key feature information, including signal drift amplitude in communication anomalies, delay vectors in keyboard responses, and disturbance spectrum features in voltage data. This converts the multi-source abnormal data into a set of feature vectors. Risk pattern recognition: Analyze the feature vector set and, combined with the preset micro-fault evolution model, identify the corresponding risk pattern type and output a structured risk pattern vector including the fault prediction category, risk level, and confidence score.

4. The fire remote control device based on Internet of Things technology according to claim 3, characterized in that: The abnormal feature decoding includes: Multi-source abnormal data preprocessing and channel normalization: Perform classification processing, extract communication, keyboard and voltage data channels respectively, perform channel normalization, and construct a standardized channel data sequence set in, is the normalized communication status time series, is the normalized response delay sequence, is the normalized voltage waveform sequence; Communication anomaly feature extraction: Based on Calculate the signal drift amplitude to describe the communication stability trend; Keyboard response feature extraction: based on Extract the response delay vector and construct the delay mean and delay standard deviation of the response time; Voltage disturbance spectrum feature extraction: Perform fast Fourier transform to extract the voltage disturbance spectrum amplitude vector; Feature vector concatenation: The extracted signal drift amplitude, delay mean, delay standard deviation, and voltage disturbance spectrum amplitude vectors are concatenated to form a feature vector set X.

5. The fire remote control device based on Internet of Things technology according to claim 4, characterized in that: The risk pattern identification includes: Fault mode similarity calculation: the feature vector set X and the fault template library in the micro-fault evolution model Each template M in k Calculate the Euclidean distance S k ; Fault category matching and confidence scoring: Determine the best matching template index k * , and calculate the matching confidence score C of the current risk pattern risk ; Risk level classification: Based on matching confidence score C risk Risk levels are divided into high risk, medium risk and low risk, which are expressed as: Structured risk pattern vector generation: The identified fault prediction categories (formulated by template The risk level RiskLevel and the matching confidence score C risk Integrate into a structured risk pattern vector R.

6. The fire remote control device based on Internet of Things technology according to claim 5, characterized in that: The device layer stability prediction module includes: State fusion: Receives structured risk pattern vectors and combines them with historical operating data of the target device and current environmental parameters, including operating time, historical fault frequency, temperature and humidity, and power supply status level, to generate a comprehensive device state vector. Stability prediction: Based on the comprehensive device status vector, the stability level of the device within the future scheduled monitoring period is predicted, including level labels such as stable operation, slight abnormal trends, and high-risk failure tendencies. The corresponding potential failure types are also identified, and the prediction results with confidence labels are output.

7. The fire remote control device based on Internet of Things technology according to claim 6, characterized in that: The state fusion includes: Data source reception and standardization processing: Receive the structured risk pattern vector R, and synchronously call the historical operation data and current environment parameters of the target device, including the operation time T run , Historical fault frequency N falult , temperature T env 、Humidity H env , power supply status level G power , and perform normalization operations on historical operation data and current environmental parameters to generate a standardized feature set Weighted feature fusion: A linear weighted model is used to weight the standardized feature set with the risk level and confidence score in the structured risk pattern vector to generate a comprehensive device status vector Z.

8. The fire remote control device based on Internet of Things technology according to claim 7, characterized in that: The stability prediction includes: Stability score calculation: Use the prediction function to infer the comprehensive vector Z of the device state to obtain the stability score S of the device in the next 30 minutes. stab ; Stability level classification and fault type identification: According to the stability score S stab The interval is divided into stability levels StabLevel, including stable operation, slight abnormal trend, and high-risk fault tendency, and the corresponding potential fault type FaultCat is identified, which is expressed as: The prediction results are packaged and output: the stability level StabLevel, potential fault type FaultCat and confidence label P conf =S stab Together, they package and output the structured prediction results P result .

9. The fire remote control device based on Internet of Things technology according to claim 8, characterized in that: The cross-node linkage consistency assessment module includes: Stability prediction result summary: collects the structured prediction results generated by multiple fire nodes in the same monitoring period Build result collection And obtain the communication link identifier T corresponding to each node through the device configuration mapping table i With power bus mark M i , and the predicted results are completed as follows: in, is the total number of fire protection nodes participating in the linkage assessment during the current monitoring period, is the prediction result after structure completion; Node grouping by link and bus: based on Link identifier T in i With busbar mark M i , divide the nodes into multiple groups G t ={i|T i =t}, G m ={i|M i =m}, forming communication link group sets respectively Grouped with power busbars Consistency risk score calculation: For each group G x The prediction results were analyzed for consistency and the consistency risk score R was calculated. x , if R x ≥0.7, then group G is determined x For a high-consistency risk node group, if 0.4≤R x <0.7, then group G is determined x is a moderate consistency risk node group, if R x <0.4, then group G is determined x A node group with low consistency risk; Generate linkage consistency report: For each high consistency risk node group, count the FaultCat with the highest frequency i As a failure mode, generate a consistency report R group .

10. The fire remote control device based on Internet of Things technology according to claim 9, characterized in that: The intelligent disposal strategy generation module includes: Dynamic matching of risk treatment templates: Based on consistency report R group , the matching template is retrieved from the preset risk treatment template library. If the group type given in the consistency report is communication link and the dominant fault type is abnormal disconnection, the communication isolation template is matched. If the dominant fault type is power disturbance, the delayed intervention and frequent monitoring templates are matched; Generate intelligent disposal suggestion package: Based on the results of dynamic matching of risk disposal templates, generate intelligent disposal suggestion package, including warning level, recommended operation window, communication isolation range and permission control strategy.

Images