Intelligent building fire prevention management system and method

Abstract

The application discloses an intelligent building fireproofing management system and method, and particularly relates to the technical field of building fireproofing; by dividing the building fireproofing system into multiple monitoring links, collecting delay accumulation, server computing load and network congestion fluctuation data of each monitoring link in different time periods, determining the weight of delay accumulation data of each monitoring link and calculating the overall delay accumulation interference index, dividing the delay situation into three levels of high, medium and low according to the index result, and taking corresponding management measures for different levels, for the high delay level, the timeliness of the fire detection and alarm notification link is preferentially ensured, and for the medium delay level, the computing load and bandwidth resources are dynamically adjusted by analysis, which not only improves the response speed and safety of the building fireproofing system in the early stage of fire, but also reduces the risk of personnel retention in high-rise or closed space.

Description

Technical Field

[0001] The present invention relates to the technical field of building fire protection, and in particular to an intelligent building fire protection management system and method. Background Art

[0002] Intelligent building fire prevention and fire management refers to a systematic approach that leverages modern intelligent technologies to monitor, manage, and control fire safety within buildings in real time. By deploying various sensors, monitoring equipment, and an intelligent management platform, this system enables early detection of fire hazards, automatic alarms, intelligent linkage with firefighting equipment, and data analysis. Its goal is to improve the efficiency and accuracy of fire prevention and control, minimize fire damage, and ensure personal safety.

[0003] In intelligent building fire management systems, technologies such as the Internet of Things (IoT), big data analytics, and artificial intelligence (AI) are being applied to monitor and evaluate various fire protection system parameters, such as temperature, smoke concentration, and flame identification. When the system detects an anomaly, it triggers an early warning mechanism, automatically notifying relevant personnel or the fire department and activating relevant fire protection facilities, such as sprinklers, smoke exhaust systems, and evacuation indicators, to achieve automated fire emergency response.

[0004] The existing technology has the following shortcomings:

[0005] Because building fire protection systems typically involve multi-layered feedback and command transmission processes, abnormal data from sensors must first be analyzed and classified by a processor before triggering early warnings and coordinated actions. Delays at any level in the building fire protection system (e.g., excessive server computing load, network data congestion) can lead to cumulative delays, resulting in feedback delays. In this scenario, the system detects a fire, but the triggered linkage devices and notification systems are delayed in activation, and personnel and the fire department may not receive the alert in a timely manner. Furthermore, due to the lack of real-time performance of building fire protection systems, alarm signals may be delayed in reaching the evacuation system and fire department, preventing evacuation from occurring in the early stages of a fire. People in the building, particularly those on upper floors or in confined spaces, may not receive alerts and evacuation instructions during this delay, putting them at risk of being trapped by smoke and flames. Summary of the Invention

[0006] The purpose of the present invention is to provide an intelligent building fire prevention and fire management system and method to solve the deficiencies in the background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: an intelligent building fire prevention and firefighting management method, comprising the following steps:

[0008] S1: Divide the building fire protection system into several monitoring links and collect server computing load data and network data congestion fluctuation status data for each monitoring link in different time periods;

[0009] S2: Determine the weight of the accumulated delay data of each monitoring link in different time periods based on the obtained server computing load data and network data congestion fluctuation status data, and calculate the weighted average of the accumulated delay data of each monitoring link in different time periods to obtain the overall delay cumulative interference index of the building fire protection system;

[0010] S3: Compare and analyze the calculated overall delay cumulative interference index of the building fire protection system with the gradient delay threshold set based on pre-acquired historical data, and classify the delay cumulative interference of the building fire protection system into different levels based on the analysis results;

[0011] S4: Further analyze the cumulative interference of the building fire protection system delay within a fixed time period, and dynamically adjust the distribution of the building fire protection system computing load and the use of bandwidth resources based on the analysis results.

[0012] Preferably, in S2, the memory usage saturation in the acquired server computing load data is analyzed to generate a memory usage saturation anomaly index. The memory usage saturation anomaly index is obtained by:

[0013] Collect and define historical data on memory usage and build a first time series dataset ,in Represents the memory usage at the nth time point. The model is trained using the first time series dataset D, and the memory usage is assumed to conform to the normal distribution. ,Right now: ;in: is the average memory usage under normal circumstances, is the variance of memory usage under normal circumstances; according to the historical distribution of memory usage, set the prior probability P(normal); , represents the prior probability of abnormal state, for the memory usage x at the current time point, calculate the probability under normal conditions and the probability under abnormal conditions , assuming that the memory usage in abnormal state is uniformly distributed U(a,b), the calculation expression is: ; where [a, b] is the memory usage interval under abnormal conditions. The posterior probability is calculated according to Bayes' theorem, and the expression is: ;in, Indicates the normal probability weight when the memory usage is x, LK represents the abnormal probability weight when the memory usage is x, and LK is the posterior probability, which is taken as the memory usage saturation abnormality index.

[0014] Preferably, in S2, a network bandwidth occupation fluctuation index is generated according to the fluctuation of the network bandwidth occupation data in the obtained network data congestion fluctuation state data. The network bandwidth occupation fluctuation index is obtained by:

[0015] Collecting network bandwidth occupation data in a W time period to form a second time series data set , wherein represents the network bandwidth occupation data at time m. Based on the characteristics of time series, the autoregressive order p, the difference order d, and the moving average order q are selected, the network bandwidth occupation data is input into the ARIMA (p, d, q) model and is trained, and the model formula is:

[0016] ; wherein: the actual network bandwidth occupation data at time t, is an autoregressive coefficient, is a moving average coefficient, is a random error term at time t, the model parameters ϕ and θ are solved by the maximum likelihood estimation method, and the fitted ARIMA model is obtained;

[0017] The fitted ARIMA model is used to predict the network bandwidth occupation data at the current time period t , and the expression is: ; wherein, represents the random error term at time , the predicted value is compared with the actual observed value , and the residual error is calculated: ; the residual error reflects the deviation of the current network bandwidth occupation data from the predicted value, the residual error is standardized, and the network bandwidth occupation fluctuation index is calculated, that is, the absolute value of the standardized residual error is calculated.

[0018] Preferably, in S2, the memory usage saturation anomaly index and the network bandwidth occupation fluctuation index are converted into a comprehensive feature vector, the comprehensive feature vector is taken as an input of a machine learning model, the machine learning model takes a weight assignment label of delay cumulative data of each monitoring link in different time periods as a prediction target, a sum of prediction errors of the weight assignment label of delay cumulative data of each monitoring link in all different time periods is minimized as a training target, the machine learning model is trained until the sum of prediction errors converges, and the weight assignment of delay cumulative data of each monitoring link in different time periods is determined according to an output result of the model, wherein the machine learning model is a polynomial regression model, and a whole delay cumulative interference index of the building fire protection system is obtained by weighted average calculation of the weight assignment of delay cumulative data of each monitoring link in different time periods.

[0019] Preferably, in S3, the calculated whole delay cumulative interference index of the building fire protection system is compared with a gradient delay threshold value set according to historical data, and the comparison is specifically as follows:

[0020] The obtained whole delay cumulative interference index of the building fire protection system is compared with the gradient delay threshold value, the gradient delay threshold value includes a first delay threshold value and a second delay threshold value, the first delay threshold value is less than the second delay threshold value, and the whole delay cumulative interference index of the building fire protection system is compared with the first delay threshold value and the second delay threshold value respectively.

[0021] If the whole delay cumulative interference index of the building fire protection system is greater than the second delay threshold value, it indicates that the whole delay cumulative interference degree of the building fire protection system is high, and the delay cumulative interference situation of the building fire protection system is divided into a high delay cumulative interference level; if the whole delay cumulative interference index of the building fire protection system is greater than or equal to the first delay threshold value and less than or equal to the second delay threshold value, it indicates that the whole delay cumulative interference degree of the building fire protection system is moderate, and the delay cumulative interference situation of the building fire protection system is divided into a moderate delay cumulative interference level; if the whole delay cumulative interference index of the building fire protection system is less than the first delay threshold value, it indicates that the whole delay cumulative interference degree of the building fire protection system is low, and the delay cumulative interference situation of the building fire protection system is divided into a low delay cumulative interference level.

[0022] Preferably, in S4, for the medium delay cumulative interference level, that is, the overall delay cumulative interference index of the building fire protection system generated within a fixed time period is greater than or equal to the first delay threshold and less than or equal to the second delay threshold, the overall delay cumulative interference index of the building fire protection system generated within a subsequent fixed time period that is greater than or equal to the first delay threshold and less than or equal to the second delay threshold is collected, and a corresponding data set is established, the mean and standard deviation of the data set are calculated, the degree of abnormality of the overall delay cumulative interference of the building fire protection system within the fixed time period is analyzed, and the distribution of the computing load of the building fire protection system and the use of bandwidth resources are dynamically adjusted according to the analysis results.

[0023] Preferably, if the mean of the overall delay cumulative interference index in the data set is greater than or equal to the reference threshold of the mean of the overall delay cumulative interference index, and the standard deviation of the overall delay cumulative interference index is less than the reference threshold of the standard deviation of the overall delay cumulative interference index, it indicates that the delay accumulation anomaly is stable and continuous with small fluctuations. At this time, a secondary warning signal is generated, and it is necessary to optimize the current computing load distribution and bandwidth resource configuration, concentrate on improving resources, and focus on alleviating bottlenecks.

[0024] If the mean of the overall delay cumulative interference index is greater than or equal to the reference threshold of the mean of the overall delay cumulative interference index, and the standard deviation of the overall delay cumulative interference index is greater than or equal to the reference threshold of the standard deviation of the overall delay cumulative interference index, it indicates that the delay accumulation anomaly is unstable and fluctuates greatly. At this time, a first-level warning signal is generated, and a load balancing mechanism and dynamic bandwidth management are introduced;

[0025] If the mean of the overall cumulative interference index is less than the reference threshold of the mean of the overall cumulative interference index, and the standard deviation of the overall cumulative interference index is greater than or equal to the reference threshold of the standard deviation of the overall cumulative interference index, it indicates that the abnormality of cumulative delay fluctuates greatly, but the overall abnormality is mild. In this case, a third-level warning signal is generated to monitor load changes, automatically allocate more bandwidth resources when the load is high, and maintain overall monitoring to prevent further aggravation of delay fluctuations;

[0026] If the mean of the overall delay cumulative interference index is less than the reference threshold of the mean of the overall delay cumulative interference index, and the standard deviation of the overall delay cumulative interference index is less than the reference threshold of the standard deviation of the overall delay cumulative interference index, it means that the delay accumulation anomaly is low and stable. At this time, no warning signal is generated. The existing configuration is maintained, and resource allocation is regularly monitored and maintained to maintain smooth operation of the system.

[0027] The present invention also provides an intelligent building fire prevention and firefighting management system, comprising a data collection module, a calculation module, a delay interference level classification module, and a dynamic adjustment module;

[0028] Data collection module: This module divides the building fire protection system into several monitoring links and collects the accumulated delay data of each monitoring link in different time periods. It also obtains the server computing load data and network data congestion fluctuation status data of each monitoring link in different time periods.

[0029] Calculation module: Determines the weight of the cumulative delay data of each monitoring link in different time periods based on the acquired server computing load data and network data congestion fluctuation status data, and calculates the weighted average of the cumulative delay data of each monitoring link in different time periods to obtain the overall delay cumulative interference index of the building fire protection system;

[0030] Low-latency and high-latency processing modules: If the building fire protection system's cumulative delay interference is classified as low, no additional measures are required, and regular system monitoring and maintenance are required. If the building fire protection system's cumulative delay interference is classified as high, the system process is immediately adjusted to prioritize delays in fire detection and alarm notification.

[0031] Dynamic adjustment module: further analyzes the cumulative interference of the building fire protection system delay within a fixed time period, and dynamically adjusts the distribution of the building fire protection system computing load and the use of bandwidth resources based on the analysis results.

[0032] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0033] 1. This invention collects and analyzes real-time accumulated delay data, server load, and network congestion data from various monitoring links of the building fire protection system. It dynamically determines the weights of accumulated delays within different time periods and the overall accumulated delay interference index. It then classifies delays into different levels and responds in real time based on thresholds from historical data. This method rapidly identifies different levels of accumulated system delay (high, medium, and low). For high-delay scenarios, it immediately optimizes processes to prioritize the real-time performance of fire detection and alarm notification. For medium-delay scenarios, it adaptively adjusts computing load and bandwidth resources to ensure that system delays remain within a manageable range, improving the system's overall responsiveness.

[0034] 2. By generating a memory usage saturation anomaly index and a network bandwidth usage fluctuation index, and assigning latency weights based on a machine learning model, this invention accurately identifies the key factors influencing latency accumulation within each time period. Furthermore, through three levels of early warning signals (level 1, 2, and 3), the system can promptly adjust load distribution and bandwidth resources to address unstable or minor latency fluctuations. Ultimately, this ensures the real-time response and operational stability of the building fire protection system in emergency situations, effectively reducing response delays in the early stages of a fire, and improving building fire safety and emergency evacuation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments described in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0036] Figure 1 Flow chart of the method of the present invention.

[0037] Figure 2 It is a system module diagram of the present invention. DETAILED DESCRIPTION

[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0039] Example 1, please refer to Figure 1 As shown, the intelligent building fire prevention and firefighting management method described in this embodiment includes the following steps:

[0040] S1: Divide the building fire protection system into several monitoring links and collect server computing load data and network data congestion fluctuation status data for each monitoring link in different time periods;

[0041] S2: Determine the weight of the accumulated delay data of each monitoring link in different time periods based on the obtained server computing load data and network data congestion fluctuation status data, and calculate the weighted average of the accumulated delay data of each monitoring link in different time periods to obtain the overall delay cumulative interference index of the building fire protection system;

[0042] S3: Compare and analyze the calculated overall delay cumulative interference index of the building fire protection system with the gradient delay threshold set based on pre-acquired historical data, and classify the delay cumulative interference of the building fire protection system into different levels based on the analysis results;

[0043] S4: Further analyze the cumulative interference of the building fire protection system delay within a fixed time period, and dynamically adjust the distribution of the building fire protection system computing load and the use of bandwidth resources based on the analysis results.

[0044] In S1, the building fire protection system is divided into several monitoring links, and the accumulated delay data of each monitoring link in different time periods is collected. In addition, the server computing load data and network data congestion fluctuation status data of each monitoring link in different time periods are obtained. Specifically:

[0045] The workflow of a building fire protection system can be divided into the following five main monitoring links, each of which represents a key node in the system's fire protection function:

[0046] Monitoring Stage 1: Fire Detection and Data Collection: This includes sensors for smoke, temperature, and flames, used to monitor potential fire signals in the environment in real time. Sources of delay: The time it takes for sensor data to be collected and initially transmitted to the central processing unit.

[0047] Monitoring Phase 2: Data Processing and Analysis: After receiving sensor data, the system server analyzes, classifies, and identifies anomalies to determine whether alarm conditions have been met. Sources of delay: The computing time consumed by the server during data processing, which may be affected by server load.

[0048] Monitoring Stage 3: Communication Transmission: Transmits analysis results and instructions from the server to connected devices and alarm systems, notifying relevant personnel and the fire department. Source of Delay: Data transmission time within the building's internal communication network, which can be affected by network congestion or bandwidth limitations.

[0049] Monitoring Link 4: Linkage Control: This function controls the activation of linked devices such as sprinklers, smoke exhaust, and emergency lighting, ensuring timely response in the event of a fire. Sources of Delay: The time it takes for commands to reach the control system from the server and activate linked devices can be delayed due to the complexity of the device control program or the response time of the hardware.

[0050] Monitoring Link 5: Alarm and Notification: This function transmits fire information to building management and the fire department, including information such as the fire location and severity. Sources of Delay: The time it takes for information to be sent to external recipients, including the fire department and building management, can be affected by network connectivity or external communication systems.

[0051] Analyze the delay of each monitoring link in different time periods, specifically including: dividing a day into multiple time periods (such as morning peak, noon, evening peak, night, etc.) to capture the impact of different load conditions on system delay. Deploy a delay monitoring program or timestamp recording module at the key nodes of each monitoring link. Timestamp recording: Record timestamps at the start and end of each link to calculate the delay. For example, record the time at nodes from fire detection to data processing completion, from data processing to the start of communication transmission, etc. Set the frequency of delay collection, such as once every minute or hour, and record the delay data to the database. Calculate the delay time of each link by the difference between the start and end timestamps, and accumulate the average delay value for different time periods.

[0052] The server computing load data reflects the pressure of the server processing tasks in each monitoring link. Install load monitoring software (such as Prometheus, Nagios, Zabbix, etc.) on the system server to record the CPU, memory and hard disk usage in real time. According to the time period of the aforementioned delayed collection, collect server load data in the corresponding time period to ensure that the load data matches the delay data. Key indicators collected: CPU usage: reflects the execution status of the server computing tasks. High CPU usage may cause computing delays. Memory usage: When the memory is close to saturation, the system will perform operations such as paging, resulting in a decrease in data processing speed. Storage read and write rate: The read and write speed during the data processing process directly affects the analysis and classification speed. The load data is aggregated according to the collection time period and stored in the database for subsequent analysis.

[0053] Network data congestion will affect data transmission and system response speed. Deploy network monitoring tools (such as Wireshark, SolarWinds, NetFlow Analyzer, etc.) to record network status in real time. Select key network indicators: Network bandwidth utilization: Monitor the network bandwidth usage of each monitoring link. When the bandwidth is close to saturation, the transmission speed decreases. Data packet loss rate: A high packet loss rate indicates an unstable network connection, which affects the success rate of data transmission. Network delay (Ping or RTT): Regularly ping network nodes or measure the round-trip time (RTT) to obtain network delay fluctuations. Set network collection for different time periods: According to the collection time period of delay data, record the network status in the corresponding time period to ensure data consistency. Store congestion status data by time period, and regularly calculate the average and peak values ​​to capture network status fluctuations.

[0054] Accumulated latency data, server computing load data, and network congestion status data for different time periods are stored in a unified database. Using data analysis tools or scripts, latency data is correlated with load and network status data to identify the causes of latency accumulation in different time periods. Latency reports are generated based on the data from different time periods, providing data support for latency cause analysis and real-time performance improvements.

[0055] S2: Determine the weight assignment of the cumulative delay data of each monitoring link in different time periods based on the obtained server computing load data and network data congestion fluctuation status data, and perform weighted average calculation on the weight assignment of the cumulative delay data of each monitoring link in different time periods to obtain the overall delay cumulative interference index of the building fire protection system.

[0056] After analyzing the memory usage saturation in the obtained server computing load data, a memory usage saturation anomaly index is generated. The memory usage saturation anomaly index is obtained as follows:

[0057] Collect and define historical data on memory usage and build a first time series dataset ,in Represents the memory usage at the nth time point. The model is trained using the first time series dataset D, and the memory usage is assumed to conform to the normal distribution. ,Right now: ;in: is the average memory usage under normal circumstances, is the variance of memory usage under normal circumstances; according to the historical distribution of memory usage, set the prior probability P(normal): for example, assuming it is 0.95 (or 95%), it means that most memory usage should be normal under no special circumstances; , represents the prior probability of abnormal state, which is used for initial hypothesis and represents subjective belief about normal and abnormal situations before observing data. For the memory usage x at the current time point, calculate the probability under normal conditions and the probability under abnormal conditions , the memory usage in abnormal situations presents a wide distribution (can be uniform distribution or wide normal distribution). Assuming the memory usage in abnormal situations is uniformly distributed U(a,b), the calculation expression is: Where [a, b] is the memory usage interval under abnormal conditions. The posterior probability (i.e., the probability of an abnormal state after observing memory usage x) is calculated using Bayes' theorem. The expression is: ;in, Indicates the normal probability weight when the memory usage is x, represents the abnormal probability weight when the memory usage rate is x, LK is the posterior probability, and the posterior probability is taken as the memory usage saturation abnormality index.

[0058] The greater the memory usage saturation abnormality index, the closer the server memory usage in the current time period to the normal saturation value or the greater the server memory usage, which indicates that the server bears a high load pressure in data processing or computing tasks. The saturation of memory resources will lead to a decrease in data processing efficiency, thereby increasing the delay accumulation of each monitoring link, especially in the fire detection and data analysis links. This delay may cause the response speed of alarm and linkage control to decrease. The high memory saturation abnormality index thus indicates that the system real-time performance is poor, and the delay accumulation of each monitoring link is serious, and immediate optimization of memory resources or load balancing measures need to be taken to reduce the delay impact.

[0059] On the contrary, the smaller the memory usage saturation abnormality index, the more normal or low the load level of the memory usage, which indicates that the computing resources of the server in the current time period are sufficient, and the data processing and task execution efficiency is high. In this case, the delay accumulation of each monitoring link is relatively light, the real-time responsiveness of the system is good, and the alarm and linkage control can be started in time. The low memory saturation abnormality index usually indicates that the system is running in a stable state, and the delay abnormality is not significant, and the memory usage can be monitored regularly to ensure the continuous efficient operation of the system.

[0060] The network bandwidth occupation fluctuation index is generated by analyzing the fluctuation of the network bandwidth occupation data in the obtained network data congestion fluctuation state data, and the acquisition method of the network bandwidth occupation fluctuation index is as follows:

[0061] Collect network bandwidth occupation data in a W time period to form a second time series data set , wherein represents the network bandwidth occupation data at time m, based on the characteristics of time series, the autoregressive order p, the difference order d, and the moving average order q are selected, which can be assisted by ACF (autocorrelation function) and PACF (partial autocorrelation function) diagrams or using Akaike information criterion (AIC) for model optimization.

[0062] The network bandwidth occupation data is input into the ARIMA (p, d, q) model and trained, and the model formula is as follows: , wherein the actual network bandwidth occupation data at time t, is an autoregressive coefficient (used to explain the historical dependence relationship of bandwidth occupation). is a moving average coefficient (used to explain the historical dependence relationship of error), is the random error term at time t. The model parameters ϕ and θ are solved by the maximum likelihood estimation method to obtain the fitted ARIMA model.

[0063] Using the fitted ARIMA model, the network bandwidth usage data for the current time period t is predicted To make a prediction, the expression is: Where, Indicates time The random error term will predict the value Compared with the actual observation value For comparison, calculate the residuals: ; residual This residual reflects the deviation between the current network bandwidth usage data and the predicted value. A larger residual indicates abnormal fluctuations in bandwidth usage. To quantify the degree of fluctuation, the residual is normalized to more clearly identify anomalies. The network bandwidth usage fluctuation index is calculated by taking the absolute value of the normalized residual.

[0064] A higher network bandwidth usage fluctuation index indicates more dramatic fluctuations in network bandwidth usage during the current time period, exceeding normal ranges. This abnormal fluctuation can lead to cumulative data transmission delays, particularly for critical processes that rely on real-time data transmission, such as fire detection and alarm notification. A high fluctuation index indicates severe network congestion during that time period, increasing data transmission delays across various monitoring links. This delays the response time for alarms, linkages, and notifications, increasing security risks. Therefore, a high fluctuation index requires immediate network optimization measures to mitigate the impact of latency.

[0065] Conversely, a smaller network bandwidth usage fluctuation index indicates relatively stable network bandwidth usage within normal fluctuations. Bandwidth stability reduces data transmission latency across all monitoring links and minimizes significant delay accumulation, helping to ensure real-time responsiveness across all monitoring links in the building fire protection system. During periods of low fluctuation index, alarms and linkage control operate efficiently, with minimal delays. The system operates at peak efficiency, eliminating the need for additional network adjustments.

[0066] The memory usage saturation anomaly index and the network bandwidth occupancy fluctuation index are converted into comprehensive feature vectors, which are used as the input of the machine learning model. The machine learning model uses each set of comprehensive feature vectors to predict the weighted assignment labels of the accumulated delay data of each monitoring link in different time periods as the prediction target, and takes minimizing the sum of the prediction errors of the weighted assignment labels of the accumulated delay data of each monitoring link in all different time periods as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the model training is stopped. The weight assignment of the accumulated delay data of each monitoring link in different time periods is determined according to the model output results. Among them, the machine learning model is a polynomial regression model, and the weighted average calculation of the weight assignment of the accumulated delay data of each monitoring link in different time periods is used to obtain the overall delay cumulative interference index of the building fire protection system.

[0067] The method for obtaining the weight assignment of the accumulated data of the delay of each monitoring link in different time periods is as follows: from the comprehensive feature vector training data of the trained machine learning model, the corresponding function expression is obtained: Where, is the output function of the model, LK is the memory usage saturation anomaly index, GH is the network bandwidth occupancy fluctuation index, Assign weights to the delayed accumulated data of each monitoring link in different time periods.

[0068] S3: The calculated overall delay cumulative interference index of the building fire protection system is compared and analyzed with the gradient delay threshold set based on pre-acquired historical data. Based on the analysis results, the delay cumulative interference conditions of the building fire protection system are divided into different levels, including a high delay cumulative interference level, a medium delay cumulative interference level, and a low delay cumulative interference level.

[0069] Comparing the obtained overall delay cumulative interference index of the building fire protection system with a gradient delay threshold, where the gradient delay threshold includes a first delay threshold and a second delay threshold, and the first delay threshold is less than the second delay threshold, and comparing the overall delay cumulative interference index of the building fire protection system with the first delay threshold and the second delay threshold respectively;

[0070] If the building fire protection system's overall cumulative delay interference index exceeds the second delay threshold, it indicates a high level of cumulative delay interference and is classified as high. In this case, system latency is severe, indicating significant cumulative delays due to abnormal fluctuations in memory usage and network bandwidth. This condition can significantly impact fire alarms, evacuations, and coordinated responses, requiring immediate resource optimization or policy adjustments to minimize delays.

[0071] If the overall cumulative delay interference index of the building fire protection system is greater than or equal to the first delay threshold and less than or equal to the second delay threshold, the overall cumulative delay interference level of the building fire protection system is medium, and the system's cumulative delay interference is classified as medium. At this point, the system's cumulative delay is significant, but within a reasonable range. Monitoring and optimization measures, such as load balancing and bandwidth adjustment, can be implemented to prevent further deterioration of cumulative delay.

[0072] If the overall cumulative delay interference index of the building fire protection system is less than the first delay threshold, it indicates a low level of cumulative delay interference and is classified as low. In this case, the system's cumulative delay is low, indicating sufficient memory and network resources, timely data transmission and response across all monitoring links, and a stable and efficient system. No additional measures are required; normal monitoring and maintenance are sufficient.

[0073] If the delay cumulative interference situation of the building fire protection system is classified as a low delay cumulative interference level, no additional measures are required to regularly monitor and maintain the system; if the delay cumulative interference situation of the building fire protection system is classified as a high delay cumulative interference level, the system process should be adjusted immediately to give priority to delays in fire detection and alarm notification links.

[0074] When the building fire protection system's cumulative latency interference is classified as low, it indicates that system latency is within a normal and stable range, with minimal cumulative latency. Data processing and transmission across all monitoring links are smooth, meeting the real-time requirements of the building fire protection system. In this state, the system operates smoothly, with sufficient resource load and network bandwidth, enabling timely responses for key processes such as fire detection, alarm notification, and coordinated control. Therefore, no additional intervention is required, requiring only regular system monitoring and maintenance to ensure consistently low latency and mitigate potential impacts from hardware aging or increased load.

[0075] When a building fire protection system's cumulative delay interference is classified as high, it means the system's cumulative delay has exceeded the safety threshold, seriously impacting the timeliness of fire detection and alarm notification. This accumulated delay significantly slows the transmission of fire detection signals and alarm information, potentially delaying fire alarm response. This prevents building occupants, management, and the fire department from being notified quickly in the early stages of a fire, increasing the risk of fire spread and occupants being trapped. To ensure the real-time nature of fire detection and alarm notification, the system must immediately adjust its processes, prioritizing these two processes to the highest level and ensuring resources are focused on these critical tasks.

[0076] Specific adjustments include prioritizing memory, CPU, and network bandwidth resources for the fire detection and alarm notification modules, while enabling dedicated data transmission channels to ensure rapid processing and transmission of critical data even in high-latency situations. Furthermore, non-critical tasks can be paused or postponed to free up system resources and prevent secondary tasks from interfering with fire detection and alarm notification. If delays persist, a switch to a backup communication link or the activation of an edge computing node can be made to process fire signals locally as much as possible, reducing reliance on the primary server and further ensuring system reliability and responsiveness under high-latency conditions.

[0077] S4: Further analyze the cumulative interference of the building fire protection system delay within a fixed time period, and dynamically adjust the distribution of the building fire protection system computing load and the use of bandwidth resources based on the analysis results.

[0078] For the medium delay cumulative interference level, that is, the overall delay cumulative interference index of the building fire protection system generated within a fixed time period is greater than or equal to the first delay threshold and less than or equal to the second delay threshold, the overall delay cumulative interference index of the building fire protection system generated within a subsequent fixed time period that is greater than or equal to the first delay threshold and less than or equal to the second delay threshold is collected, and a corresponding data set is established, the mean and standard deviation of the data set are calculated, the abnormal degree of the overall delay cumulative interference of the building fire protection system within the fixed time period is analyzed, and the distribution of the computing load and the use of bandwidth resources of the building fire protection system are dynamically adjusted according to the analysis results.

[0079] If the mean of the overall delay cumulative interference index in the data set is greater than or equal to the reference threshold of the mean of the overall delay cumulative interference index, and the standard deviation of the overall delay cumulative interference index is less than the reference threshold of the standard deviation of the overall delay cumulative interference index, it means that the delay accumulation anomaly is stable and continuous with small fluctuations, and there is a bottleneck in the system load and network resources at a fixed level. At this time, a secondary warning signal is generated, and it is necessary to optimize the current computing load distribution and bandwidth resource configuration, concentrate on improving resources, and focus on alleviating the bottleneck part to improve the delay accumulation situation.

[0080] If the mean of the overall delay cumulative interference index is greater than or equal to the reference threshold of the mean of the overall delay cumulative interference index, and the standard deviation of the overall delay cumulative interference index is greater than or equal to the reference threshold of the standard deviation of the overall delay cumulative interference index, it indicates that the delay accumulation anomaly is unstable and fluctuates greatly, and the system load and bandwidth resources vary greatly in different time periods. At this time, a first-level warning signal is generated, and a load balancing mechanism and dynamic bandwidth management are introduced to flexibly allocate resources and adapt to delay fluctuations to ensure that the system has sufficient resources to cope with delays during both peak and off-peak periods.

[0081] If the mean of the overall delay cumulative interference index is less than the reference threshold of the mean of the overall delay cumulative interference index, and the standard deviation of the overall delay cumulative interference index is greater than or equal to the reference threshold of the standard deviation of the overall delay cumulative interference index, it indicates that the delay accumulation anomaly fluctuates greatly, but the overall anomaly is mild, and system resources are under pressure during certain periods. At this time, a third-level warning signal is generated to monitor load changes, automatically allocate more bandwidth resources when the load is high, and maintain overall monitoring to prevent further aggravation of delay fluctuations.

[0082] If the mean of the overall delay cumulative interference index is less than the reference threshold of the mean of the overall delay cumulative interference index, and the standard deviation of the overall delay cumulative interference index is less than the reference threshold of the standard deviation of the overall delay cumulative interference index, it means that the delay accumulation anomaly is low and stable, and the system resource configuration basically meets the needs. At this time, no warning signal is generated, the existing configuration is maintained, and resource allocation is regularly monitored and maintained to maintain smooth operation of the system.

[0083] It should be noted here that the importance of the first-level warning signal is greater than that of the second-level warning signal, and the importance of the second-level warning signal is greater than that of the third-level warning signal. Relevant personnel can take corresponding management measures according to different warning signal levels.

[0084] When the cumulative abnormalities in a building fire protection system's latency become unstable and fluctuate significantly, the system generates a Level 1 warning signal, indicating that latency varies significantly over time, making it difficult to predict and control. This typically indicates that server computing load and network bandwidth fluctuate dramatically over time. This may be due to accumulated latency during peak hours or increased network congestion, leading to unstable response times for critical tasks such as fire detection, alarms, and evacuation guidance. Therefore, to ensure timely system responses under varying load conditions, load balancing mechanisms and dynamic bandwidth management are necessary to stabilize system performance.

[0085] The load balancing mechanism involves dynamically allocating computing tasks to less-loaded server nodes based on the server's current load, preventing latency accumulation caused by overloading a single node. This effectively distributes pressure and ensures balanced load across nodes. In the Level 1 alert state, the system prioritizes fire detection and alarm notification tasks, deferring or shifting non-urgent tasks to a lower priority level to ensure the real-time performance of critical tasks. If a node is detected to be overloaded, the system automatically activates a backup server or computing resource to share the high-load tasks, thereby reducing latency fluctuations during peak periods.

[0086] Dynamic bandwidth management specifically involves setting bandwidth priorities for critical processes (such as alarm transmission and fire signal processing) to ensure that important data is transmitted first during network congestion, minimizing latency. The system monitors network bandwidth usage and dynamically allocates bandwidth resources during periods of significant network fluctuation to prevent data transmission from being impacted by momentary congestion. In the event of severe network fluctuations, multipath transmission can be enabled to distribute critical data across multiple network paths, ensuring that at least one path can transmit important information immediately.

[0087] In this embodiment, the building fire protection system is divided into multiple monitoring segments. Data on cumulative delay, server computing load, and network congestion fluctuations for each monitoring segment over different time periods is collected and analyzed. The cumulative delay weight for each monitoring segment is determined, and the overall cumulative delay interference index (CDI) for the system is calculated. This index is compared with a gradient threshold set based on historical data, and the system's latency is categorized as high, medium, or low. For low latency, regular monitoring and maintenance are sufficient. For high latency, immediate process adjustments are made to prioritize the real-time performance of fire detection and alarm notification. For medium latency, further analysis of latency anomalies is performed, and the system's computing load and bandwidth resource allocation are dynamically adjusted to optimize system responsiveness.

[0088] Example 2, please refer to Figure 2 As shown, the intelligent building fire prevention and fire protection management system described in this embodiment includes a data collection module, a calculation module, a delay interference level classification module and a dynamic adjustment module;

[0089] Data collection module: This module divides the building fire protection system into several monitoring links and collects the accumulated delay data of each monitoring link in different time periods. It also obtains the server computing load data and network data congestion fluctuation status data of each monitoring link in different time periods.

[0090] Calculation module: Determines the weight of the cumulative delay data of each monitoring link in different time periods based on the acquired server computing load data and network data congestion fluctuation status data, and calculates the weighted average of the cumulative delay data of each monitoring link in different time periods to obtain the overall delay cumulative interference index of the building fire protection system;

[0091] Delay interference level classification module: This module compares and analyzes the calculated overall delay cumulative interference index of the building fire protection system with the gradient delay threshold set based on pre-acquired historical data. Based on the analysis results, the delay cumulative interference of the building fire protection system is classified into different levels, including high delay cumulative interference level, medium delay cumulative interference level, and low delay cumulative interference level.

[0092] The dynamic adjustment module: further analyzes the delay cumulative interference of the building fire protection system in a fixed time period, and dynamically adjusts the distribution of the building fire protection system calculation load and the use of bandwidth resources according to the analysis result.

[0093] The above formulas are dimensionless values calculated, and the formulas are obtained by collecting a large amount of data to simulate a formula of the latest real situation, and the preset parameters in the formula are set by the person skilled in the art according to the actual situation.

[0094] It should be understood that the term "and / or" herein merely describes an association relationship of associated objects, and indicates that there can be three relationships, for example, A and / or B can represent three cases of A alone, A and B together, and B alone, wherein A and B can be singular or plural. In addition, the character " / " herein generally represents that the associated objects before and after it are in an "or" relationship, but it can also represent an "and / or" relationship, which can be understood in combination with the context.

[0095] Those skilled in the art can realize that the units and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0096] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed in the present application, which should be covered within the protection scope of the present application.

Claims

1. An intelligent building fire prevention and firefighting management method, characterized by: The following steps are involved: S1: Divide the building fire protection system into several monitoring links and collect server computing load data and network data congestion fluctuation status data for each monitoring link in different time periods; S2: Determine the weight of the accumulated delay data of each monitoring link in different time periods based on the obtained server computing load data and network data congestion fluctuation status data, and calculate the weighted average of the accumulated delay data of each monitoring link in different time periods to obtain the overall delay cumulative interference index of the building fire protection system; S3: Compare and analyze the calculated overall delay cumulative interference index of the building fire protection system with the gradient delay threshold set based on pre-acquired historical data, and classify the delay cumulative interference of the building fire protection system into different levels based on the analysis results; S4: Further analyze the cumulative interference of the building fire protection system delay within a fixed time period, and dynamically adjust the distribution of the building fire protection system computing load and the use of bandwidth resources based on the analysis results.

2. The intelligent building fire prevention and firefighting management method according to claim 1, characterized in that: In S2, the memory usage saturation in the obtained server computing load data is analyzed to generate a memory usage saturation anomaly index. The memory usage saturation anomaly index is obtained as follows: Collect and define historical data on memory usage and build a first time series dataset ,in Represents the memory usage at the nth time point. The model is trained using the first time series dataset D, and the memory usage is assumed to conform to the normal distribution. ,Right now: ;in: is the average memory usage under normal circumstances, is the variance of memory usage under normal circumstances; according to the historical distribution of memory usage, set the prior probability P(normal); , represents the prior probability of abnormal state, for the memory usage x at the current time point, calculate the probability under normal conditions and the probability under abnormal conditions , assuming that the memory usage in abnormal state is uniformly distributed U(a,b), the calculation expression is: ; where [a, b] is the memory usage interval under abnormal conditions. The posterior probability is calculated according to Bayes' theorem, and the expression is: ;in, Indicates the normal probability weight when the memory usage is x, It represents the abnormal probability weight when the memory usage is x, LK is the posterior probability, and the posterior probability is used as the memory usage saturation abnormality index.

3. The intelligent building fire prevention and firefighting management method according to claim 2, characterized in that: In S2, the fluctuation of the network bandwidth occupancy data in the obtained network data congestion fluctuation status data is analyzed to generate a network bandwidth occupancy fluctuation index. The method for obtaining the network bandwidth occupancy fluctuation index is as follows: Collect network bandwidth usage data within the W time period to form the second time series data set ,in The network bandwidth usage data at time m is represented by the autoregressive order p, the difference order d, and the moving average order q based on the characteristics of time series. The network bandwidth usage data is input into the ARIMA (p, d, q) model and trained. The model formula is: ;in: The actual network bandwidth usage data at time t, is the autoregressive coefficient, is the moving average coefficient, is the random error term at time t, and the model parameters ϕ and θ are solved by the maximum likelihood estimation method to obtain the fitted ARIMA model; Using the fitted ARIMA model, the network bandwidth usage data for the current time period t is predicted To make a prediction, the expression is: Where, Indicates time The random error term will predict the value Compared with the actual observation value For comparison, calculate the residuals: ; residual It reflects the deviation between the current network bandwidth usage data and the predicted value. The residual is standardized and the network bandwidth usage fluctuation index is calculated, that is, the absolute value of the standardized residual is calculated.

4. The intelligent building fire prevention and firefighting management method according to claim 3 is characterized by: In S2, the memory usage saturation anomaly index and the network bandwidth occupancy fluctuation index are converted into comprehensive feature vectors, and the comprehensive feature vectors are used as the input of the machine learning model. The machine learning model uses each group of comprehensive feature vectors to predict the weighted assignment labels of the accumulated delay data of each monitoring link in different time periods as the prediction target, and takes minimizing the sum of the prediction errors of the weighted assignment labels of the accumulated delay data of each monitoring link in all different time periods as the training target. The machine learning model is trained until the sum of the prediction errors reaches convergence, and the model training is stopped. The weight assignment of the accumulated delay data of each monitoring link in different time periods is determined according to the model output results. Among them, the machine learning model is a polynomial regression model, and the weighted average calculation of the weight assignment of the accumulated delay data of each monitoring link in different time periods is used to obtain the overall delay cumulative interference index of the building fire protection system.

5. The intelligent building fire prevention and firefighting management method according to claim 4, characterized in that: In S3, the calculated overall delay cumulative interference index of the building fire protection system is compared and analyzed with the gradient delay threshold set based on the pre-acquired historical data, specifically: Comparing the obtained overall delay cumulative interference index of the building fire protection system with a gradient delay threshold, where the gradient delay threshold includes a first delay threshold and a second delay threshold, and the first delay threshold is less than the second delay threshold, and comparing the overall delay cumulative interference index of the building fire protection system with the first delay threshold and the second delay threshold respectively; If the overall delay cumulative interference index of the building fire protection system is greater than the second delay threshold, it means that the overall delay cumulative interference level of the building fire protection system is high, and the delay cumulative interference situation of the building fire protection system is classified as a high delay cumulative interference level; if the overall delay cumulative interference index of the building fire protection system is greater than or equal to the first delay threshold and less than or equal to the second delay threshold, it means that the overall delay cumulative interference level of the building fire protection system is at a medium level, and the delay cumulative interference situation of the building fire protection system is classified as a medium delay cumulative interference level; if the overall delay cumulative interference index of the building fire protection system is less than the first delay threshold, it means that the overall delay cumulative interference level of the building fire protection system is low, and the delay cumulative interference situation of the building fire protection system is classified as a low delay cumulative interference level.

6. The intelligent building fire prevention and firefighting management method according to claim 5, characterized in that: In S4, for the medium delay cumulative interference level, that is, the overall delay cumulative interference index of the building fire protection system generated within a fixed time period is greater than or equal to the first delay threshold and less than or equal to the second delay threshold, the overall delay cumulative interference index of the building fire protection system generated within a subsequent fixed time period that is greater than or equal to the first delay threshold and less than or equal to the second delay threshold is collected, and a corresponding data set is established, the mean and standard deviation of the data set are calculated, the degree of abnormality of the overall delay cumulative interference of the building fire protection system within the fixed time period is analyzed, and the distribution of the computing load and the use of bandwidth resources of the building fire protection system are dynamically adjusted according to the analysis results.

7. The intelligent building fire prevention and firefighting management method according to claim 6, characterized in that: If the mean of the overall delay cumulative interference index in the data set is greater than or equal to the reference threshold of the overall delay cumulative interference index mean, and the standard deviation of the overall delay cumulative interference index is less than the reference threshold of the overall delay cumulative interference index standard deviation, it means that the delay accumulation anomaly is stable and continuous with small fluctuations. At this time, a second-level warning signal is generated, and it is necessary to optimize the current computing load distribution and bandwidth resource configuration, concentrate on improving resources, and focus on alleviating bottlenecks. If the mean of the overall delay cumulative interference index is greater than or equal to the reference threshold of the mean of the overall delay cumulative interference index, and the standard deviation of the overall delay cumulative interference index is greater than or equal to the reference threshold of the standard deviation of the overall delay cumulative interference index, it indicates that the delay accumulation anomaly is unstable and fluctuates greatly. At this time, a first-level warning signal is generated, and a load balancing mechanism and dynamic bandwidth management are introduced; If the mean of the overall cumulative interference index is less than the reference threshold of the mean of the overall cumulative interference index, and the standard deviation of the overall cumulative interference index is greater than or equal to the reference threshold of the standard deviation of the overall cumulative interference index, it indicates that the abnormality of cumulative delay fluctuates greatly, but the overall abnormality is mild. In this case, a third-level warning signal is generated to monitor load changes, automatically allocate more bandwidth resources when the load is high, and maintain overall monitoring to prevent further aggravation of delay fluctuations; If the mean of the overall delay cumulative interference index is less than the reference threshold of the mean of the overall delay cumulative interference index, and the standard deviation of the overall delay cumulative interference index is less than the reference threshold of the standard deviation of the overall delay cumulative interference index, it means that the delay accumulation anomaly is low and stable. At this time, no warning signal is generated. The existing configuration is maintained, and resource allocation is regularly monitored and maintained to maintain smooth operation of the system.

8. An intelligent building fire prevention and fire management system, used to implement an intelligent building fire prevention and fire management method according to any one of claims 1 to 7, characterized in that: It includes data collection module, calculation module, delay interference level classification module and dynamic adjustment module; Data collection module: This module divides the building fire protection system into several monitoring links and collects the accumulated delay data of each monitoring link in different time periods. It also obtains the server computing load data and network data congestion fluctuation status data of each monitoring link in different time periods. Calculation module: Determines the weight of the cumulative delay data of each monitoring link in different time periods based on the acquired server computing load data and network data congestion fluctuation status data, and calculates the weighted average of the cumulative delay data of each monitoring link in different time periods to obtain the overall delay cumulative interference index of the building fire protection system; Delay interference level classification module: This module compares and analyzes the calculated overall delay cumulative interference index of the building fire protection system with the gradient delay threshold set based on pre-acquired historical data. Based on the analysis results, the delay cumulative interference of the building fire protection system is classified into different levels, including high delay cumulative interference level, medium delay cumulative interference level, and low delay cumulative interference level. Dynamic adjustment module: further analyzes the cumulative interference of the building fire protection system delay within a fixed time period, and dynamically adjusts the distribution of the building fire protection system computing load and the use of bandwidth resources based on the analysis results.

Images