Intelligent fire-fighting service management system

Abstract

The invention relates to the technical field of fire-fighting management, and discloses an intelligent fire-fighting service management system, which comprises a data acquisition unit used for acquiring three basic parameters, namely environment data, working condition data and battery data, of a fire-fighting object in an office building, and the system acquires the environment, working condition and battery data of a fire extinguisher through the data acquisition unit; the fusion unit generates a dynamic composite abnormal threshold value, and the adjustment unit executes a three-stage sampling frequency adjustment mechanism according to the threshold value, the battery health degree score and the data deviation degree, which means that the sampling frequency can be reduced to reduce the battery energy consumption when the state of the fire extinguisher is normal and the risk is relatively low, and when the risk is increased or the battery health degree is reduced, the battery energy consumption is reduced. The whole system realizes accurate management and regulation of the battery according to the actual working condition of the fire extinguisher, energy waste is avoided, the service life of the battery is prolonged, the cost of frequently replacing the battery is reduced, and the sustainability of intelligent fire control management is enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of fire management, and particularly relates to a smart fire service management system. Background Art

[0002] Currently in office buildings, it is usually relied on a platform to achieve the collection of data of multiple fire extinguishers, so that property management personnel can remotely view the real-time status of the fire extinguishers. Coupled with regular personnel inspections, once the data of the fire extinguishers is abnormal, a real-time warning can be issued. However, in actual use, since temperature sensors, pressure sensors, and RFID tags all need to be powered by batteries, in order to ensure real-time management.

[0003] Currently, a continuous monitoring mode with a fixed sampling frequency is generally adopted, and the battery energy consumption will gradually decrease with the monitoring time, which leads to the need to replace the battery regularly. In the overall fire management, it is often impossible to effectively manage and control the battery according to the actual working conditions of the fire fighting equipment, resulting in a situation of insufficient energy utilization in the system and reducing the management ability of smart fire protection. Summary of the Invention

[0004] In view of the deficiencies of the prior art, the present invention provides a smart fire service management system to solve the above problems.

[0005] The above technical objectives of the present invention are achieved through the following technical solutions:

[0006] A smart fire service management system includes:

[0007] A data acquisition unit for acquiring three basic parameters of environmental data, working condition data, and battery data of fire fighting objects in the office building, where the fire fighting object is a fire extinguisher;

[0008] A fusion unit for fusing environmental data, working condition data, and battery data to generate a dynamic composite anomaly threshold for battery state perception;

[0009] An adjustment unit for comparing the working condition data, environmental data, and battery data with the dynamic composite anomaly threshold and implementing a three-level sampling frequency adjustment mechanism based on battery health;

[0010] A management unit for dividing the office building into fire protection areas managed by master and slave nodes, and implementing a battery priority hierarchical wake-up strategy in combination with the distribution characteristics of the fire protection areas and the dynamic composite anomaly threshold;

[0011] A balance unit for constructing a three-dimensional evaluation matrix of battery remaining life - monitoring efficiency - energy consumption cost based on the battery data and the three-level sampling frequency adjustment mechanism, and generating a fire management strategy based on the three-dimensional evaluation matrix.

[0012] Further, the environmental data, working condition data, and battery data are fused to generate a dynamic composite anomaly threshold for battery state perception, including:

[0013] The normalized environmental data, working condition data, and battery data are fused and weighted to obtain fused data;

[0014] The fused data is analyzed to generate a dynamic composite anomaly threshold for battery state perception, specifically including:

[0015] Calculate the average value of the fused data within the historical period to obtain the historical average value;

[0016] Calculate the standard deviation of the fused data within the historical period to obtain the historical standard deviation;

[0017] By scaling the product of the historical average value and the historical standard deviation, with the scaling ratio controlled by a dynamic adjustment coefficient, and finally adding the scaled result to the historical mean value to form a dynamic composite anomaly threshold;

[0018] The dynamic adjustment coefficient is composed of the superposition of three parts, and the three parts respectively reflect the dynamic impacts of battery health status, environmental fluctuation characteristics, and working condition stability on the threshold.

[0019] Further, the working condition data, environmental data, and battery data are compared with the dynamic composite anomaly threshold, and a three-level sampling frequency adjustment mechanism based on battery health is executed, including:

[0020] Calculate the battery data to obtain a battery health score;

[0021] Calculate the data deviation degrees of the working condition data, environmental data, and battery data from the dynamic composite anomaly threshold, specifically as follows:

[0022] Calculate the absolute difference between the working condition parameter and the dynamic composite anomaly threshold, and then divide it by the dynamic composite anomaly threshold to obtain the relative deviation degree of the working condition parameter;

[0023] Calculate the absolute difference between the environmental parameter and the dynamic composite anomaly threshold, and then divide it by the dynamic composite anomaly threshold to obtain the relative deviation degree of the environmental parameter;

[0024] Calculate the absolute difference between the battery parameter and the dynamic composite anomaly threshold, and then divide it by the dynamic composite anomaly threshold to obtain the relative deviation degree of the battery parameter;

[0025] Among them, the weight value range of the relative deviation degree of the working condition parameter is between 0.3 and 0.4, the weight value range of the relative deviation degree of the environmental parameter is between 0.2 and 0.3, and the weight value range of the relative deviation degree of the battery parameter is between 0.3 and 0.5. The sum of the three weights is equal to 1;

[0026] Add the relative deviation degrees of operating conditions parameters, environmental parameters, and battery parameters respectively according to their respective weights to obtain the data deviation degree;

[0027] According to the battery health score and the data deviation degree, divide the sampling frequency into three levels, namely the first-level sampling frequency, the second-level sampling frequency, and the third-level sampling frequency.

[0028] Furthermore, divide the office building into fire protection areas managed by master and slave nodes, including:

[0029] Divide the office building into multiple fire protection areas according to floors, and each floor is a fire protection area;

[0030] Each fire protection area contains all the fire extinguishers on this floor, and 1 master node is set in each fire protection area, and multiple slave nodes are set under each master node;

[0031] The master node is used to manage multiple slave nodes on this floor;

[0032] The slave nodes are associated with multiple fire extinguishers, collect the environmental data, operating conditions data, and battery data of the fire extinguishers in the area, and transmit the three basic parameters to the master node.

[0033] Furthermore, combine the fire protection area distribution characteristics and the dynamic composite anomaly threshold to execute the battery priority hierarchical wake-up strategy, including:

[0034] According to the fire protection areas divided by floors, extract the distribution characteristic parameters of each fire protection area. The distribution characteristic parameters are composed of the fire extinguisher deployment density and the master-slave node distance;

[0035] Among them, by dividing the number of fire extinguishers in the fire protection area of each floor by the area of the fire protection area of this floor, the fire extinguisher deployment density can be obtained;

[0036] Use the Euclidean distance formula to calculate the straight-line distance from each fire extinguisher to its affiliated master node to obtain the master-slave node distance;

[0037] Associate the dynamic composite anomaly threshold by area, and summarize the dynamic composite anomaly thresholds reported by all slave nodes on this floor through the master node to calculate the regional average dynamic threshold, specifically including:

[0038] Collect the dynamic composite anomaly thresholds reported by all slave nodes under this floor through the master node, use the interquartile range method to identify and eliminate the outlier dynamic composite anomaly thresholds, and perform spatial weighted averaging on the effective dynamic composite anomaly thresholds according to the fire extinguisher deployment density and distance weights of the areas associated with each slave node, and finally generate the regional average dynamic threshold reflecting the overall state of the fire protection system on this floor.

[0039] Furthermore, in combination with the fire area distribution characteristics and the dynamic composite anomaly threshold, implementing the battery priority hierarchical wake-up strategy further includes:

[0040] Combining the regional average dynamic threshold with the distribution characteristic parameters to generate a regional comprehensive risk score, specifically including: normalizing the fire extinguisher deployment density and the master-slave node distance, then establishing a multi-parameter coupling model, taking the regional average dynamic threshold as the core risk input item, the deployment density as the reverse adjustment factor, and the master-slave node distance as the efficiency compensation item, and then using the analytic hierarchy process to determine the weights of each parameter, and non-linearly coupling the regional average dynamic threshold and the distribution characteristic parameters through a weighted fusion algorithm to generate a quantitative risk score on a 0-100 scale;

[0041] When the regional comprehensive risk score is in the range of 0-40 points, the fire area is a low-risk area;

[0042] When the regional comprehensive risk score is in the range of 41-70 points, the fire area is a medium-risk area;

[0043] When the regional comprehensive risk score is in the range of 71-100 points, the fire area is a high-risk area;

[0044] Implement the battery priority hierarchical wake-up strategy based on the regional comprehensive risk score.

[0045] Furthermore, implementing the battery priority hierarchical wake-up strategy based on the regional comprehensive risk score includes:

[0046] Analyzing the regional comprehensive risk score and the battery health score to generate a wake-up priority index;

[0047] Implement the battery priority hierarchical wake-up strategy through the wake-up priority index and the regional comprehensive risk score, specifically including:

[0048] The battery priority hierarchical wake-up strategy includes primary wake-up, secondary wake-up, and tertiary wake-up.

[0049] Furthermore, according to the battery data and the three-level sampling frequency adjustment mechanism, construct a three-dimensional evaluation matrix of battery remaining life - monitoring efficiency - energy consumption cost:

[0050] Based on the battery data and the three-level sampling frequency adjustment mechanism, obtain the three normalized indicators of battery remaining life index, monitoring efficiency index, and energy consumption cost index, and construct a three-dimensional evaluation matrix of battery remaining life - monitoring efficiency - energy consumption cost based on the three indicators;

[0051] Dynamically evaluate the monitoring efficiency and data deviation degree to generate a joint evaluation value for each node;

[0052] Generate the monitoring priorities of the fire area nodes based on the combined evaluation values of each node in the fire area, specifically including:

[0053] When the combined evaluation value ≥ 0.85, it is determined as a high-priority node and real-time monitoring is started. When 0.5 ≤ combined evaluation value < 0.85, it is determined as a medium-priority node and the normal monitoring cycle is executed. When the combined evaluation value < 0.5, it is determined as a low-priority node and enters the standby monitoring mode;

[0054] Dynamically adjust the weights of the three-dimensional evaluation matrix according to the risk level of the fire area.

[0055] Furthermore, dynamically adjust the weights of the three-dimensional evaluation matrix according to the risk level of the fire area, including:

[0056] When the fire area is a high-risk area, increase the monitoring efficiency weight and reduce the energy consumption cost weight;

[0057] When the fire area is a medium-risk area, maintain the default weights;

[0058] When the fire area is a low-risk area, increase the energy consumption cost weight and reduce the monitoring efficiency weight.

[0059] Furthermore, generate fire management strategies based on the three-dimensional evaluation matrix, including:

[0060] Analyze the three-dimensional evaluation matrix to obtain the comprehensive score of the fire area nodes;

[0061] Generate fire management strategies based on the comprehensive score. The fire management strategies include: battery health management strategy and battery dynamic power consumption regulation strategy;

[0062] The battery health management strategy is as follows:

[0063] When the node is of high priority and the comprehensive score ≥ 0.85, the sampling frequency is increased to 1 minute / time at this time, and the battery is marked for replacement within 30 days;

[0064] When the node is of medium priority and 0.5 ≤ comprehensive score < 0.85, maintain sampling at 10 minutes / time;

[0065] When the node is of low priority and the comprehensive score < 0.5, the sampling frequency is reduced to 6 hours / time, and only the basic monitoring of the battery SOC is retained;

[0066] The battery dynamic power consumption regulation strategy is as follows:

[0067] For high-risk fire areas, the master node is fully active and the slave node wakes up once every 30 minutes;

[0068] For medium-risk fire protection areas, the master node works intermittently (waking up once every 10 minutes), and the slave nodes adjust sampling dynamically according to the battery health score.

[0069] For low-risk fire protection areas, the master node goes into deep sleep (activated twice a day), and the slave nodes are only woken up emergently when the temperature in the fire protection area > 60°C.

[0070] In summary, the present invention mainly has the following beneficial effects:

[0071] Through the multi-dimensional data fusion and dynamic threshold adjustment mechanism, precise scheduling of battery energy is achieved. In the traditional fixed sampling mode, a large amount of energy is wasted because it cannot sense the differences in the environment, working conditions, and battery status. This solution innovatively fuses and weights the environmental temperature and humidity, fire extinguisher pressure parameters, and battery health to generate a dynamic composite anomaly threshold, and combines a three-level sampling frequency adjustment mechanism, enabling the system to dynamically adjust the monitoring frequency according to the deviation degree between the battery health score and real-time data. When the data deviation degree is low, the sampling frequency is automatically reduced to once every 6 hours, while in high-risk scenarios, it is increased to once a minute, effectively balancing the monitoring accuracy and energy consumption cost. In addition, by constructing a three-dimensional evaluation matrix of battery remaining life - monitoring efficiency - energy consumption cost, the system can dynamically adjust the weight distribution of different risk areas. For example, in high-risk areas, the monitoring efficiency is prioritized, while in low-risk areas, energy consumption optimization is emphasized, thereby realizing gradient management of the batteries of fire protection equipment in the entire building and significantly reducing the labor and material costs of property management for battery replacement.

[0072] Through the master-slave node architecture and hierarchical wake-up strategy, the problem of resource allocation in traditional fire protection management is solved. Based on the fire protection areas divided by floors, combined with spatial characteristic parameters such as the deployment density of fire extinguishers and the distance between master and slave nodes, the system generates a comprehensive regional risk score (0 - 100 points), and accordingly executes a priority hierarchical wake-up strategy. In high-risk areas (71 - 100 points), the master node is fully active, and the slave nodes are woken up once every 30 minutes to ensure real-time response to abnormal events. In medium-risk areas (41 - 70 points), the master node adopts an intermittent working mode, and the slave nodes dynamically adjust the sampling period according to the battery health. In low-risk areas (0 - 40 points), they go into deep sleep, only retaining basic monitoring functions. Through differential control, the overall system energy consumption is reduced while the response speed of key areas is improved.

[0073] By constructing a three-dimensional evaluation matrix, the mode transformation from passive replacement to active maintenance is achieved. Through the quantitative analysis of the remaining battery life, monitoring efficiency, and energy consumption cost by the three-dimensional evaluation matrix, the system can automatically generate a comprehensive node score (0 - 1 point), and accordingly implement a hierarchical management strategy: high-priority nodes (≥0.85 points) are marked for battery replacement within 30 days, medium-priority nodes maintain routine monitoring, and low-priority nodes (<0.5 points) enter the standby mode. This mechanism can avoid the risk of sudden power outages. At the same time, the dynamic power consumption regulation strategy concentrates the node energy consumption in the high-risk area during the critical monitoring period. Combined with the emergency wake-up mechanism, on the premise of ensuring the safety bottom line, the annual wake-up times of devices in the low-risk area are reduced, achieving the intelligent fire protection service management effect for the office building. Brief Description of the Drawings

[0074] Figure 1 It is a block diagram of the intelligent fire protection service management system of the present invention. Detailed Embodiment

[0075] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

[0076] Refer to Figure 1 , an intelligent fire protection service management system, including:

[0077] A data acquisition unit for acquiring three basic parameters, namely environmental data, operating condition data, and battery data of fire protection objects in the office building, where the fire protection object is a fire extinguisher;

[0078] A fusion unit for fusing environmental data, operating condition data, and battery data to generate a dynamic composite anomaly threshold for battery state perception;

[0079] An adjustment unit for comparing the operating condition data, environmental data, and battery data with the dynamic composite anomaly threshold and implementing a three-level sampling frequency adjustment mechanism based on battery health;

[0080] A management unit for dividing the office building into fire protection areas managed by master-slave nodes and implementing a battery priority hierarchical wake-up strategy in combination with the distribution characteristics of the fire protection areas and the dynamic composite anomaly threshold;

[0081] A balance unit for constructing a three-dimensional evaluation matrix of battery remaining life - monitoring efficiency - energy consumption cost according to the battery data and the three-level sampling frequency adjustment mechanism, and generating a fire protection management strategy based on the three-dimensional evaluation matrix.

[0082] Through the coordinated operation of multiple units, a three-dimensional fire-fighting equipment monitoring and management system is constructed. Among them, the data acquisition unit realizes the full-dimensional real-time acquisition of the environment, working conditions, and battery data of fire extinguishers, laying a data foundation for accurate monitoring. The fusion unit dynamically generates composite anomaly thresholds based on multi-source data, breaking through the limitations of traditional fixed-threshold monitoring, and can be intelligently adapted to changes in environmental conditions, significantly improving the accuracy and timeliness of battery state perception. The adjustment unit constructs a three-level sampling frequency adjustment mechanism, which can dynamically optimize the data acquisition density according to the battery health, reducing redundant energy consumption while ensuring monitoring efficiency, and achieving an intelligent balance between monitoring accuracy and energy efficiency ratio. The management unit implements differential management through the master-slave node fire area division and priority-level wake-up strategy, combining regional distribution characteristics, which not only ensures key monitoring of high-risk areas but also reduces unnecessary energy consumption through the wake-up strategy, improving the overall response efficiency of the system. The balance unit constructs a three-dimensional evaluation matrix of battery remaining life - monitoring efficiency - energy consumption cost, providing a quantitative decision-making basis for the generation of fire-fighting management strategies, promoting the transformation of fire-fighting management from experience-driven to data-driven, and realizing the organic unity of the full-life-cycle health management of equipment, the optimal allocation of monitoring resources, and the refined control of operating costs. Through the deep coordination of data fusion, dynamic adjustment, intelligent management, and multi-dimensional evaluation, the intelligent level of fire-fighting equipment monitoring in office buildings is effectively improved, and then intelligent fire-fighting service management is realized for office buildings.

[0083] In one case of this embodiment, three basic parameters, namely, the environmental data, working condition data, and battery data of fire-fighting objects in the office building are obtained, including:

[0084] The environmental data are: temperature, humidity, and smoke concentration;

[0085] The working condition data are: pressure, position, and temperature and humidity (only representing the temperature and humidity of the fire extinguisher);

[0086] The battery data are: battery SOC and the health state of the battery.

[0087] In one case of this embodiment, the environmental data, working condition data, and battery data are fused to generate a dynamic composite anomaly threshold for battery state perception, including:

[0088] The normalized environmental data, working condition data, and battery data are fused and weighted to obtain the fusion data. The specific calculation formula is as follows:

[0089]

[0090] In the formula, F represents the fusion data, E represents the environmental data, E min and E max respectively represent the minimum and maximum values of the environmental data, R represents the working condition data, R min and R maxrespectively represent the minimum and maximum values of the operating condition data, B represents the battery data, B min and B max respectively represent the minimum and maximum values of the battery data, α, β, and γ respectively represent the weights corresponding to the environmental data E, the operating condition data R, and the battery data B, where the value ranges of α, β, and γ are all between 0.3 and 0.4, and α + β + γ = 1;

[0091] Analyze the fused data to generate a dynamic composite anomaly threshold for battery state perception. The specific calculation formula is as follows:

[0092] Y = μ + δσ·F;

[0093] In the formula, Y represents the dynamic composite anomaly threshold, μ represents the historical mean of the fused data F (the historical mean can be obtained by adding up all the past fused data and dividing by the number of fused data), σ represents the historical standard deviation of the fused data F (calculate the difference between each fused data and the historical mean, then square these differences respectively, add them up, divide by the number of fused data minus one, and finally take the square root to obtain the historical standard deviation), and δ represents the dynamic adjustment coefficient;

[0094] Calculate the dynamic adjustment coefficient δ. The calculation formula is as follows:

[0095]

[0096] In the formula, δ0 represents the initial dynamic adjustment coefficient, k1 and k2 are respectively the influence coefficients of the battery health and the comprehensive fluctuation factor, both of which are positive numbers. Among them, the comprehensive fluctuation factor mainly refers to the fluctuation characteristics and stability characteristics of the environmental data and the operating condition data. e represents the natural constant, HF represents the battery health score, and its value range is [0, 1]. The closer HF is to 1, the closer the battery performance is to the brand-new state. The closer HF is to 0, the more serious the battery aging and the worse the performance. γ1 represents the exponential parameter affected by the battery health, V represents the change rate of the fused data F, λ1 represents the adjustment parameter of the fused data change rate V, ΔE represents the change amount of the environmental data, represents the historical average value of the environmental data, ΔR represents the change amount of the operating condition data, represents the historical average value of the operating condition data, S represents the operating condition stability index, and λ2 represents the adjustment parameter of the operating condition stability;

[0097] By means of normalization and weighted fusion, comprehensively considering data from three aspects: environment, working conditions, and the battery, it can comprehensively reflect the actual operating state of the battery under different working conditions and environments, avoid misjudgment caused by a single data factor, make the subsequent evaluation of the battery state more accurate. Based on the fused data, a dynamic composite anomaly threshold is calculated, which can change with the data changes during the battery operation process, can timely reflect the dynamic characteristics of the battery state, effectively improve the sensitivity and detection effect for the abnormal state of the battery, help to timely discover potential battery problems, ensure the safe operation of the battery. The calculation of the dynamic adjustment coefficient comprehensively considers the influence of various factors on the battery, including the battery health, the fluctuation characteristics and stability of environmental data and working condition data, can more accurately adjust the dynamic composite anomaly threshold, make it more conform to the actual health status and performance of the battery under different situations, further improve the accuracy and reliability of battery state perception, and provide a strong basis for the maintenance and management of the battery.

[0098] In one case of this embodiment, the working condition data, environmental data, and battery data are compared with the dynamic composite anomaly threshold, and a three-level sampling frequency adjustment mechanism based on battery health is executed, including:

[0099] Calculate the battery data to obtain a battery health score, and the specific calculation formula is as follows:

[0100] HF = w SOC ·SOC + w HS ·HS;

[0101] In the formula, SOC represents the state of charge of the battery, and its value range is [0, 1], HS represents the health state of the battery, and its value range is [0, 1], w SOC and w HS are the weights corresponding to SOC and HS respectively, where the value range of w SOC is between 0.2 - 0.4, the value range of w HS is between 0.6 - 0.8, and w SOC + w HS = 1;

[0102] Calculate the data deviation degree of the working condition data, environmental data, and battery data from the dynamic composite anomaly threshold, and the specific calculation formula is as follows:

[0103]

[0104] Among them, D represents the data deviation degree, represents the relative deviation degree of the working condition data from the dynamic composite anomaly threshold, represents the relative deviation degree of the environmental data from the dynamic composite anomaly threshold, Indicates the relative deviation of battery data from the dynamic composite anomaly threshold, w R , w E and w B respectively represent the corresponding and weights, where the value range of w R is between 0.3 - 0.4, the value range of w E is between 0.2 - 0.3, the value range of w B is between 0.3 - 0.5, and w R + w E + w B = 1;

[0105] According to the battery health score and data deviation, the sampling frequency is divided into three levels, specifically as follows:

[0106] Set the battery health score threshold as HF th , and the data deviation threshold as D th ;

[0107] First - level sampling frequency: When HF ≥ HF th and D ≤ D th , the sampling frequency is the lowest, which is once every 30 minutes;

[0108] Second - level sampling frequency: When HF < HF th and D ≤ D th or HF ≥ HF th and D > D th , the sampling frequency is medium, which is once every 10 minutes;

[0109] Third - level sampling frequency: When HF < HF th and D > D th , the sampling frequency is the highest, which is once every 1 minute;

[0110] Through the weighted calculation based on the state of charge and health state, combined with the differential weight allocation, it not only highlights the core influence of the battery health state but also takes into account the dynamic characteristics of the real - time power level, and can more comprehensively reflect the actual performance of the battery. On this basis, the constructed three - level sampling frequency system can intelligently match the monitoring density according to the battery health: when the health is good, low - frequency sampling is adopted to effectively reduce the power consumption of data acquisition and storage pressure; when the health deteriorates, the sampling frequency is automatically increased to ensure high - frequency monitoring at the initial stage of battery performance degradation, providing data support for battery life prediction and preventive maintenance, and avoiding resource waste caused by over - sampling or monitoring blind spots caused by insufficient sampling;

[0111] Through multi-dimensional deviation analysis that integrates operating conditions, environment, and battery data, a dynamic anomaly recognition mechanism is constructed. The weight setting of battery data highlights the leading role of core monitoring indicators, while taking into account the synergistic effects of operating conditions and environmental factors. This enables the system to accurately capture abnormal fluctuations in external operating conditions and internal state parameters. The three-level adjustment strategy further enhances the adaptability of the monitoring mechanism: when the data deviation is low, the basic sampling frequency is maintained to balance monitoring efficiency and resource consumption; when an anomaly occurs in a single dimension (health or deviation), medium-frequency sampling is initiated to achieve key tracking of local risks; when double anomalies overlap, high-frequency sampling is triggered to ensure real-time capture of key data during battery performance degradation or drastic changes in the operating environment, providing high-density data support for battery safety warning and fault diagnosis, and effectively improving the reliability and response speed of the battery management system.

[0112] In one case of this embodiment, the office building is divided into fire protection areas managed by master and slave nodes, including:

[0113] The office building is divided into multiple fire protection areas according to floors, with each floor being a fire protection area;

[0114] Each fire protection area contains all the fire extinguishers on this floor, and 1 master node is set in each fire protection area, with multiple slave nodes set under each master node;

[0115] The master node is deployed in areas such as the computer room and power distribution room, and the master node is used to manage multiple slave nodes on this floor;

[0116] The slave nodes are deployed in areas such as corridors and office areas. The slave nodes are associated with multiple fire extinguishers, and collect the environmental data, operating condition data, and battery data of the fire extinguishers in the area, and transmit the three basic parameters to the master node;

[0117] By dividing the office building into fire protection areas managed by master and slave nodes according to floors, the precision and efficiency of fire protection management are realized through an intelligent hierarchical architecture. First, the independent fire protection area division based on floors, combined with the centralized control of slave nodes (corridors / office areas) by the master node (computer room / power distribution room), effectively achieves the purpose of fire protection management. The slave nodes collect the environmental data (temperature and humidity, smoke concentration), operating condition data (pressure, expiration date), and battery data (remaining power, abnormal status) of the fire extinguishers in real time, and through the summary and analysis of the master node, a dynamic monitoring network covering the whole building is constructed, enabling managers to identify potential hazards such as equipment aging and battery failure in advance, improving the ability of fire protection management. By hierarchical management, the system complexity is reduced. The master node is responsible for policy distribution and data integration, and the slave nodes focus on local data collection, which not only ensures communication stability but also reduces the impact range of single-point failures, realizes the full life cycle management of fire protection equipment, and achieves the fire safety management of high-rise buildings.

[0118] In one case of this embodiment, in combination with the distribution characteristics of the fire protection area and the dynamic composite anomaly threshold, a battery priority hierarchical wake-up strategy is executed, including:

[0119] According to the fire protection areas divided by floors, distribution characteristic parameters of each fire protection area are extracted, specifically including:

[0120] The distribution characteristic parameters are composed of the fire extinguisher deployment density and the master-slave node distance;

[0121] Among them, by dividing the number of fire extinguishers in each floor's fire protection area by the area of that floor's fire protection area, the fire extinguisher deployment density can be obtained;

[0122] Using the Euclidean distance formula to calculate the straight-line distance from each fire extinguisher to its affiliated master node, the master-slave node distance can be obtained;

[0123] Associate the dynamic composite anomaly threshold by region, and summarize the dynamic composite anomaly thresholds reported by all slave nodes on this floor through the master node to calculate the regional average dynamic threshold, specifically including: collecting the T values reported by all slave nodes under this floor through the master node, using the interquartile range method to identify and eliminate outlier T values, and performing spatial weighted averaging on the valid T values according to the fire extinguisher deployment density and distance weights of the regions associated with each slave node. Among them, the data weights of fire extinguishers with denser deployments and closer distances to the master node are higher, and finally a regional average dynamic threshold reflecting the overall state of the fire protection system on this floor is generated;

[0124] After obtaining all T values, enter the outlier identification and elimination stage, and use the interquartile range method to ensure the reliability of the data. The specific steps are as follows: First, arrange all the collected T values in ascending order, and then calculate the quartiles. Among them, the first quartile is the value at the 25% position in the sorted data, and the third quartile is the value at the 75% position. The interquartile range is the difference between the third quartile and the first quartile. According to the interquartile range method, the determination criterion for outliers is a value less than the first quartile minus 1.5 times the interquartile range, or greater than the third quartile plus 1.5 times the interquartile range. These outliers will be eliminated, and the remaining are the valid T values. This process can exclude the interference of abnormal data caused by factors such as equipment failures and data transmission anomalies;

[0125] Perform spatial weighted averaging on the valid T values according to the fire extinguisher deployment density and distance weights of the regions associated with each slave node. The specific process is: multiply the valid T value of each slave node by the corresponding comprehensive weight respectively, then add these products, and then divide by the sum of all comprehensive weights to obtain the regional average dynamic threshold;

[0126] Identify and eliminate outlier T-values through the interquartile range method, effectively excluding abnormal data interference caused by factors such as equipment failures and abnormal data transmissions, ensuring the reliability of the T-values used for calculations, and laying a solid foundation for subsequent analysis. In terms of reflecting the regional status, determine the weights by combining the fire extinguisher deployment density and the distance between the master and slave nodes, and perform spatial weighted averaging on the effective T-values, so that the data weights of fire extinguishers with denser deployments and closer distances to the master node are higher, and the fire protection status of key areas can be more accurately reflected. The generated regional average dynamic threshold can truly reflect the overall status of the floor fire protection system. In terms of resource utilization, based on the distribution characteristics of the floor fire protection areas and the dynamic composite anomaly threshold, execute battery priority hierarchical wake-up, which can reasonably allocate battery energy, preferentially wake up devices in key areas, while ensuring the monitoring effect, extend the device battery life, and improve the system operation efficiency and stability.

[0127] In one case of this embodiment, combining the fire protection area distribution characteristics and the dynamic composite anomaly threshold, and executing the battery priority hierarchical wake-up strategy, further includes:

[0128] Combine the regional average dynamic threshold with the distribution characteristic parameters to generate a regional comprehensive risk score, specifically including: normalize the fire extinguisher deployment density and the distance between the master and slave nodes to eliminate the dimension difference, use the regional average dynamic threshold as the core risk input item, the deployment density as the reverse adjustment factor, and the distance between the master and slave nodes as the efficiency compensation item, then use the analytic hierarchy process to determine the weights of each parameter, and non-linearly couple the regional average dynamic threshold and the distribution characteristic parameters through a weighted fusion algorithm to generate a quantitative risk score on a 0-100 scale;

[0129] The processing procedure for normalizing the fire extinguisher deployment density is as follows: Determine the maximum value of the fire extinguisher deployment density in all floor fire protection areas, and use it as the normalization reference value. This reference value represents the area with the densest fire extinguisher deployment in the entire office building. For each floor's fire protection area, calculate its actual fire extinguisher deployment density (i.e., the number of fire extinguishers on that floor divided by the floor area), and then compare this actual value with the reference value to obtain a ratio value. This ratio value is the normalized fire extinguisher deployment density, and its value range is between 0 and 1. For example, an office building has three floors, each with an area of 1000 square meters. The first floor has 20 fire extinguishers, and the deployment density is "20 / 1000 square meters = 0.02 fire extinguishers per square meter". The second floor has 25 fire extinguishers, and the deployment density is "25 / 1000 square meters = 0.025 fire extinguishers per square meter". The third floor has 30 fire extinguishers, and the deployment density is "30 / 1000 square meters = 0.03 fire extinguishers per square meter". At this time, 0.03 fire extinguishers per square meter on the third floor is the maximum deployment density and is used as the reference value. The normalized deployment density of the first floor is "0.02 / 0.03 ≈ 0.67", the normalized deployment density of the second floor is "0.025 / 0.03 ≈ 0.83", and the normalized deployment density of the third floor is "0.03 / 0.03 = 1". Through such normalization, the fire extinguisher deployment densities of the three floors are all converted to the range of 0 to 1, eliminating the influence of different floor area differences, enabling the fire extinguisher deployment density to be compared on a unified scale;

[0130] The processing procedure for normalizing the master-slave node distance is as follows: Determine the maximum value of the distances from all fire extinguishers to their respective master nodes, and use it as the normalization reference value. This reference value represents the farthest distance between the fire extinguisher and the master node. For each fire extinguisher, calculate its actual distance to its respective master node, and then compare this actual value with the reference value to obtain a ratio value. For example, a certain floor has three fire extinguishers, and their distances to their respective master nodes are 10 meters, 15 meters, and 20 meters respectively. Taking the maximum distance of 20 meters as the reference value, the normalized distance of the first fire extinguisher is "10 meters / 20 meters = 0.5", the normalized distance of the second fire extinguisher is "15 meters / 20 meters = 0.75", and the normalized distance of the third fire extinguisher is "20 meters / 20 meters = 1". Through this method, the distances from the fire extinguishers to the master nodes are also converted to the range of 0 to 1, eliminating the influence of different floor fire extinguisher layout differences, enabling the master-slave node distance to be compared on a unified scale;

[0131] When the comprehensive risk score of the area is between 0 and 40 points, then the fire protection area is a low-risk area;

[0132] When the regional comprehensive risk score is between 41 and 70 points, the fire protection area is a medium-risk area;

[0133] When the regional comprehensive risk score is between 71 and 100 points, the fire protection area is a high-risk area;

[0134] Execute the battery priority hierarchical wake-up strategy based on the regional comprehensive risk score;

[0135] Through parameters such as the fire extinguisher deployment density and the distance between the master and slave nodes, the maximum value benchmark normalization method is adopted to effectively eliminate the influence of dimensions such as area difference and layout difference, making the basic data of different fire protection areas comparable horizontally. For example, the fire extinguisher deployment density is converted into a proportional value in the range of 0-1, which not only retains the relative differences between regions but also avoids the interference of the area base on the evaluation results, ensuring the objectivity of data preprocessing. On this basis, with the regional average dynamic threshold as the core risk input, the weights are determined by combining the analytic hierarchy process, and the organic integration of multiple parameters is achieved through non-linear coupling to generate a quantitative score on a 0-100 scale. This evaluation mode that combines dynamic risks with static distribution characteristics breaks through the one-sidedness of single indicators, can comprehensively reflect the comprehensive risk status of the fire protection area, provides an accurate decision-making basis for subsequent hierarchical strategies, and makes the risk level division more scientific and credible;

[0136] The battery priority hierarchical wake-up strategy based on the quantitative risk score realizes the dynamic optimal allocation of fire protection system resources. By dividing the risk level into three levels: low, medium, and high, the system can intelligently adjust the wake-up mechanism according to the risk degree of different regions: in low-risk areas, due to high deployment density and excellent response efficiency, the battery wake-up frequency can be reduced, reducing energy consumption while ensuring basic monitoring. In high-risk areas, due to problems such as insufficient deployment or lagging response, high-frequency wake-up is ensured to achieve real-time monitoring and improve the timeliness of abnormal response. This differential strategy not only avoids resource waste but also solves the monitoring blind area problem in high-risk areas, achieving a balance between safety and economy. This mechanism can be continuously optimized with the dynamic change of regional parameters to adapt to complex environments such as personnel flow and equipment adjustment in office buildings and other scenarios, promoting the transformation of fire protection monitoring from passive response to active prevention and effectively improving the overall emergency management efficiency.

[0137] In one case of this embodiment, executing the battery priority hierarchical wake-up strategy based on the regional comprehensive risk score includes:

[0138] Analyze the regional comprehensive risk score and the battery health score HF to generate a wake-up priority index P, specifically including:

[0139]

[0140] In the formula, FD represents the regional comprehensive risk score, FD minand FD max are the minimum and maximum values of the regional comprehensive risk score FD respectively. wa1 and wa2 are weight coefficients respectively, and wa1 + wa2 = 1. The value range of wa1 is between 0.5 and 0.6, and the value range of wa2 is between 0.4 and 0.5. αa represents the exponential adjustment parameter of the regional comprehensive risk score, and βa represents the exponential adjustment parameter of the battery health score;

[0141] Execute the battery priority hierarchical wake-up strategy through the wake-up priority index and the regional comprehensive risk score;

[0142] Set the maximum threshold of the wake-up priority index to P max and the minimum threshold of the wake-up priority index to P min ;

[0143] The battery priority hierarchical wake-up strategy is as follows:

[0144] Level 1 wake-up: When P ≥ P max and the fire area is a high-risk area, wake up the slave nodes in this fire area every hour. The slave nodes collect three basic parameters: fire extinguisher environment data, working condition data, and battery data and upload them to the master node;

[0145] Level 2 wake-up: When P min < P < P max and the fire area is a medium-risk area, wake up the slave nodes in this fire area every 3 hours. The slave nodes only collect fire extinguisher environment data and working condition data;

[0146] Level 3 wake-up: When P ≤ P min and the fire area is a low-risk area, wake up the slave nodes in this fire area every 6 hours. The slave nodes only collect battery data;

[0147] By combining the regional comprehensive risk score FD with the battery health score HF to generate the wake-up priority index P, a comprehensive and scientific assessment of the battery wake-up demand is achieved. FD reflects the regional fire risk situation, and HF reflects the battery's own state. The integration of the two avoids the one-sidedness of a single indicator and makes the wake-up decision more in line with the actual risk situation. In terms of resource allocation, the hierarchical wake-up strategy realizes the precise utilization of resources. The first-level wake-up collects comprehensive data at high frequencies in high-risk areas to ensure timely capture of abnormal changes in the fire extinguisher environment, working conditions, and batteries, providing strong support for rapid emergency response. The second-level wake-up reasonably balances the monitoring intensity and energy consumption in medium-risk areas, only collecting key environmental and working condition data to reduce the data processing pressure. The third-level wake-up focuses on battery data in low-risk areas, minimizing the wake-up frequency to the greatest extent, saving system energy consumption and communication resources, and improving the overall operation efficiency. By setting clear wake-up priority index thresholds, a standardized and differentiated wake-up mechanism is constructed, which can not only ensure the safety monitoring accuracy in high-risk areas but also avoid resource waste in low-risk areas, significantly improving the flexibility of battery monitoring in fire protection management.

[0148] In one case of this embodiment, according to the battery data and the three-level sampling frequency adjustment mechanism, a three-dimensional evaluation matrix of battery remaining life - monitoring effectiveness - energy consumption cost is constructed:

[0149] Based on the battery data and the three-level sampling frequency adjustment mechanism, three normalized indicators, namely the battery remaining life index, the monitoring effectiveness index, and the energy consumption cost index, are obtained. Based on these three indicators, a three-dimensional evaluation matrix of battery remaining life - monitoring effectiveness - energy consumption cost is constructed. Specifically, the normalization process of the battery remaining life index is as follows: Obtain the historical remaining life data of the fire extinguisher battery, set the time unit as hours, determine the maximum remaining life value and the minimum remaining life value in the historical data. For the current battery remaining life value, subtract the minimum value from it and then divide by the difference between the maximum value and the minimum value to obtain the normalized battery remaining life index, making the result fall within the interval [0, 1]. For example, if the maximum remaining life is 1000 hours, the minimum is 200 hours, and the current battery remaining life is 600 hours, then the normalized value of the battery remaining life index is "(600 - 200) / (1000 - 200) = 0.5";

[0150] The normalization process of the monitoring effectiveness index is as follows: count the number of successful monitoring of fire abnormal events at the three-level sampling frequency, determine the maximum and minimum number of successful monitoring in the historical monitoring period, subtract the minimum value from the number of successful monitoring in the current monitoring period, and then divide by the difference between the maximum and minimum values to obtain the normalized monitoring effectiveness index, so that it falls into the interval [0, 1]. For example, the maximum number of successful monitoring is 100 times, the minimum is 10 times, and the current number of successful monitoring is 60 times, then the normalized value of the monitoring effectiveness index is "(60 - 10) / (100 - 10) ≈ 0.5556".

[0151] The normalization process of the energy consumption cost index is as follows: summarize the node energy consumption data at different sampling frequencies, determine the maximum and minimum energy consumption values in the historical operation period, subtract the minimum value from the energy consumption value in the current period, and then divide by the difference between the maximum and minimum values to obtain the normalized energy consumption cost index, so that it falls into the interval [0, 1]. For example, the maximum energy consumption is 1000 joules, the minimum is 100 joules, and the current energy consumption is 500 joules, then the normalized value of the energy consumption cost index is "(500 - 100) / (1000 - 100) ≈ 0.4444".

[0152] By normalizing the battery remaining life index, the monitoring effectiveness index, and the energy consumption cost index, the dimensional differences of the three indexes are eliminated.

[0153] Through the time series integration algorithm, process the data of the battery remaining life index, the monitoring effectiveness index, and the energy consumption cost index in the time series. Integrate the recent monitoring data of each index and convert it into a value that can reflect its overall state in a period of time. Then, use the integration results of the three indexes as the coordinate axes to construct a three-dimensional evaluation space.

[0154] According to the monitoring time series, form data points by the integration values of the three indexes in each period and input them into the fuzzy clustering algorithm. Based on the positional relationship of these data points in the three-dimensional space, considering the continuous change in time and the similarity in space, divide the data points into different categories, identify the association patterns of each index with changes in time and space, generate the clustering results, and construct a three-dimensional evaluation matrix according to the clustering results. The horizontal and vertical axes of the matrix correspond to the three indexes respectively, and each element in the matrix reflects the comprehensive evaluation value of the three indexes under specific space-time conditions, so as to comprehensively and quantitatively evaluate the battery-related performance.

[0155] Dynamically evaluate the monitoring effectiveness and data deviation degree, and generate the joint evaluation value of each node, specifically including: using the dynamic weighted coupling algorithm to jointly evaluate the monitoring effectiveness and data deviation degree. First, analyze the distribution characteristics of historical monitoring data through the entropy weight method, dynamically calculate the weight value of the monitoring effectiveness index, and automatically allocate the remaining weights to the reverse index of the data deviation degree to form a real-time weight combination. Then, normalize the monitoring effectiveness to quantify the ability of the system to detect fire anomalies. At the same time, calculate the data deviation degree to reflect the difference degree between the current data and the dynamic composite anomaly threshold. Finally, fuse the reverse value of the monitoring effectiveness and the data deviation degree according to the dynamic weight to generate the joint evaluation value of each node;

[0156] Generate the monitoring priority of the fire area nodes according to the joint evaluation value of each node in the fire area, specifically including:

[0157] Sort the joint evaluation values of each node in the fire area from high to low to generate a monitoring priority queue. When the joint evaluation value ≥ 0.85, it is determined as a high-priority node and real-time monitoring is started. When 0.5 ≤ joint evaluation value < 0.85, it is determined as a medium-priority node and the normal monitoring cycle is executed. When the joint evaluation value < 0.5, it is determined as a low-priority node and enters the standby monitoring mode;

[0158] Dynamically adjust the weight of the three-dimensional evaluation matrix according to the risk level of the fire area;

[0159] By integrating three key indicators: the remaining battery life, monitoring effectiveness, and energy consumption cost, a three-dimensional evaluation matrix is formed, which realizes the comprehensive quantification of battery-related performance. Through the normalization processing of the three indicators, the dimension difference is eliminated, enabling data of different natures to be compared and comprehensively analyzed on the same scale. Next, time series integration processing further integrates the dynamic changes of each indicator in the time dimension into a value that can reflect the overall state, providing a more stable and representative data basis for subsequent clustering analysis. Based on the spatio-temporal position relationship of these data points in the three-dimensional space, the fuzzy clustering algorithm can identify the correlation patterns of each indicator with time and space changes, and the generated clustering results accurately reflect the comprehensive state of the battery under different conditions, enabling precise management of the battery in fire services;

[0160] The dynamic weighted coupling algorithm is used to jointly evaluate the monitoring efficiency and data deviation degree, which can adjust the weight combination in real time and dynamically. The entropy weight method calculates the weight dynamically according to the distribution characteristics of historical monitoring data, ensuring the objectivity and adaptability of the evaluation process, making the fusion of monitoring efficiency and data deviation degree more scientific and reasonable. The generated joint evaluation value accurately reflects the monitoring capabilities and data quality of each node. The monitoring priority queue of the fire area nodes is generated according to the joint evaluation value, realizing the reasonable allocation of monitoring resources. High-priority nodes start real-time monitoring, which can detect fire hazards in time. Medium-priority nodes execute normal monitoring cycles to ensure regular monitoring requirements. Low-priority nodes enter the standby monitoring mode to save resources. At the same time, the weights of the three-dimensional evaluation matrix are dynamically adjusted according to the risk level of the fire area, enabling the evaluation system to flexibly adapt to different risk environments, further optimizing the fire management strategy, and effectively ensuring the safety of the fire area.

[0161] In one case of this embodiment, the dynamic adjustment of the weights of the three-dimensional evaluation matrix according to the risk level of the fire area includes:

[0162] When the fire area is a high-risk area, increase the weight of monitoring efficiency and decrease the weight of energy consumption cost;

[0163] When the fire area is a medium-risk area, maintain the default weights;

[0164] When the fire area is a low-risk area, increase the weight of energy consumption cost and decrease the weight of monitoring efficiency;

[0165] In a high-risk fire area:

[0166] 1. In a high-risk area, the fire extinguishers may accelerate battery aging due to high ambient temperature, and need to be frequently triggered in case of emergency (such as real-time uploading data to the main node). The battery performance directly determines whether the fire extinguisher can remain stable during multiple uses or long-term operation. Therefore, the weight of battery data (γ) can be set to 0.4. Although the environmental parameters (such as temperature and humidity) in high-risk areas affect battery performance, such areas are usually equipped with constant temperature and humidity equipment (such as precision air conditioners), the environmental fluctuation range is small, and the environmental risks have been controlled by physical measures (such as fireproof materials and ventilation systems) during the previous infrastructure construction. Therefore, the importance of real-time monitoring of environmental data is relatively lower than the immediate status of the battery. Therefore, the weight of environmental data (α) is appropriately reduced, and the weight of environmental data (α) can be set to 0.3. The pressure index in the working condition data is the direct basis for whether the fire extinguisher can spray the extinguishing agent (such as insufficient pressure will cause it to fail to work). Even if the battery is in good condition, abnormal working conditions will still lead to fire extinguishing failure. Therefore, the basic monitoring weight needs to be maintained. So the weight of working condition data (β) can be set to 0.3, which can maintain the basic monitoring of the working state of the fire extinguisher;

[0167] 2. In high-risk scenarios, whether the battery can be triggered multiple times is crucial. The health state (HS) of the battery reflects the long-term aging degree of the battery (such as internal resistance and capacity attenuation). Even if the state of charge (SOC) of the battery is sufficient (such as 0.8), if the health state (HS) of the battery < 0.5 (severe aging), the battery may fail due to excessive internal resistance during the first large-current discharge. Therefore, w HS can take the value of 0.8 to ensure the reliability of the battery under extreme working conditions. w SOC can take the value of 0.2;

[0168] 3. The working condition data (pressure) directly determines whether fire extinguishing can be successful (such as spraying failure if the pressure is insufficient), and the battery data (SOC / HS) determines whether the state can be continuously fed back. Therefore, w B can take the value of 0.5. For example, when the battery is abnormal (such as SOC < 0.2), even if the environment and working conditions are normal, it is necessary to wake up immediately and increase the sampling frequency to avoid "the device is available but the sampling is interrupted". w E can take the value of 0.2. Since the environment in high-risk areas has been controlled by physical measures, abnormalities are mostly short-term emergencies (such as a sudden increase in smoke concentration), and the dynamic adjustment coefficient δ in the dynamic composite anomaly threshold can be used for rapid response, without excessive weighting in the deviation calculation. w R can take the value of 0.3;

[0169] 4. The wake-up strategy for high-risk fire areas needs to synchronously respond to the overall risk of the area and the health of individual batteries. For example, the comprehensive risk score of a certain computer room area = 85 (high risk), but the health score HF of a certain fire extinguisher battery = 0.9 (healthy). At this time, the wake-up priority is dominated by the comprehensive risk score of the area to ensure that all devices in the area enter high-frequency monitoring. If the comprehensive risk score of the area = 80 and the health score HF of the battery = 0.2 (severe aging), then the weight of the health score HF of the battery is increased, and the device with the abnormal battery is woken up first. Therefore, wa1 can take the value of 0.5, and wa2 can take the value of 0.5 to avoid over-waking or missed waking caused by a single factor (such as high area risk but healthy battery);

[0170] In medium-risk fire areas:

[0171] 1. In the medium-risk area, since the environmental risks (such as human smoking and equipment overheating) and operating conditions risks (such as the fire extinguisher being moved and pressure naturally decaying) in the general office area need to be monitored synchronously. For example, an increase in humidity may cause corrosion of the metal parts of the fire extinguisher (affecting the operating conditions), and at the same time accelerate the corrosion of the battery electrodes (affecting the battery data). The two are related, and balancing the weights can avoid missing detections due to a single factor. Therefore, the environmental data weight (α) can be set to 0.35, the operating conditions data weight (β) can be set to 0.35. In the medium-risk area, the fire extinguisher is used infrequently, and the battery is in a long-term standby state. The immediate availability of the battery SOC (current battery level) has little impact (even if the battery level is 50%, it can meet the single-trigger requirement), and the slow decline of the battery's health state HS can be captured through regular sampling monitoring (such as secondary sampling frequency). Therefore, there is no extreme dependence on battery data, and the battery data weight (γ) can be set to 0.3;

[0172] 2. In the medium-risk fire area, the fire extinguisher in the office area needs to balance the battery SOC (current battery level) and the battery's health state HS. When the battery SOC < 0.2, it may cause the sensor to fail to sample on time, affecting data integrity. When the battery's health state HS < 0.6, it means that the remaining battery life < 1 year, and replacement needs to be planned in advance. Therefore, w HS can be set to 0.7, w SOC can be set to 0.3;

[0173] 3. In the medium-risk fire area, the environmental (such as smoking causing an increase in smoke concentration), operating conditions (such as the fire extinguisher being moved resulting in a position deviation), and battery (such as long-term standby causing a decrease in the battery's health state HS) risks in the office area are evenly distributed. Therefore, w R can be set to 0.3 to ensure that abnormal pressure is captured in a timely manner, w E can be set to 0.3. At the same time, attention needs to be paid to the gradual impact of temperature and humidity fluctuations on the battery and operating conditions (such as for every 10% increase in humidity, the battery self-discharge rate increases by 15%). w B can be set to 0.4 to balance the battery health and the immediate battery level, and avoid missing detections of "normal operating conditions / environment but the battery is about to fail";

[0174] 4. The overall risk of the medium-risk area mainly reflects the comprehensive impact of the fire extinguisher deployment density and the distance between the master and slave nodes. Therefore, wa1 is slightly higher than wa2 because the management goal of the medium-risk area is efficiency optimization - to give priority to waking up the fire-fighting equipment with "higher regional risk but medium battery health" (such as the regional comprehensive risk score = 60, and the battery health score HF = 0.6), and avoid energy consumption waste caused by excessive attention to the health of individual batteries (such as frequent waking up of high-health batteries in low-risk areas). Therefore, wa1 can be set to 0.55 and wa2 can be set to 0.45;

[0175] In the low-risk fire area:

[0176] 1. Although the environmental parameters seem relatively stable in the low-risk area, long-term high humidity (e.g., humidity > 80% in a storage room due to poor ventilation) or extreme temperature (e.g., temperature < 0°C in a stairwell in winter) will slowly damage the battery and operating conditions. For example, the internal resistance of the battery increases in a low-temperature environment, resulting in insufficient starting current for the sensor (it may not be able to wake up even when the battery SOC = 0.5). High humidity will cause the internal circuit board of the fire extinguisher to be damp and short-circuit (abnormal operating conditions). Therefore, the system needs to give priority to capturing environmental parameters by increasing the α weight. Therefore, the weight of environmental data (α) can be set to 0.4. The fire extinguishers in the low-risk area are rarely moved or triggered, the natural fluctuation range of operating condition data (pressure, position) is small (e.g., the pressure drops by 0.05 MPa per year), and the degradation rate of the health state (HS) of the battery is also lower than that in the high- and medium-risk areas. Therefore, the β and γ weights are set to medium, which not only ensures the monitoring of the basic state of the operating conditions and the remaining battery life but also avoids wasting resources due to excessive attention to stable parameters. Therefore, the weight of operating condition data (β) can be set to 0.3, and the weight of battery data (γ) can be set to 0.3;

[0177] 2. The battery in the low-risk fire area is in a low-power standby state for a long time, and the battery SOC fluctuates little (0.3 - 0.7). However, the degradation of the health state (HS) of the battery is hidden (e.g., it drops by 0.05 per year), and it may fail due to a sudden drop in capacity when triggered for the first time. Therefore, w HS can be set to 0.6, w SOC can be set to 0.4, thereby increasing the sensitivity to "long-term aging" through the weight of the health state (HS) of the battery and avoiding hidden failures such as "enough power but the battery has failed";

[0178] 3. In the low-risk fire area, the hidden erosion in the long-term stable environment is the main risk (e.g., the humidity in the storage room is > 85% for a long time, resulting in battery corrosion). Therefore, w E can be set to 0.5, the weight of operating condition data is the lowest, w R can be set to 0.2. Since the position of the fire extinguisher in the low-risk area is fixed and the pressure naturally decays slowly, high-frequency monitoring is not required, and w B can be set to 0.3;

[0179] 4. The comprehensive risk score of the low-risk area is usually < 40 points, indicating a high deployment density, short node distance, and stable overall environment. At this time, the wake-up priority depends more on the regional distribution characteristics. Therefore, wa1 can be set to 0.6, emphasizing "regional stability first", and wa2 can be set to 0.4;

[0180] Through the differential weight configuration of risk levels, the pertinence and reliability of fire monitoring have been significantly improved. In high-risk areas, it focuses on the immediate status of the battery and the core indicators of working conditions, and gives priority to ensuring the stable operation of fire-fighting equipment in extreme scenarios, avoiding fire extinguishing failures caused by battery aging or abnormal pressure. In medium-risk areas, it balances the weights of environmental, working condition, and battery data, and synchronously captures associated risks (such as the gradual impact of humidity on equipment components), preventing missed detections due to single factors. In low-risk areas, it strengthens the long-term erosion monitoring of the environment and gives early warnings of hidden faults (such as circuit boards getting damp due to high humidity), making up for potential risk blind spots in a stable environment. This hierarchical strategy enables the monitoring resources to precisely match the risk characteristics, avoiding monitoring blind spots in high-risk scenarios and preventing waste of resources in low-risk scenarios;

[0181] Through the dynamic adjustment of weights, the monitoring efficiency and energy consumption cost have been effectively balanced, and the system management efficiency has been improved. In high-risk areas, by increasing the weight of the battery health status and the regional risk dominant mechanism in the wake-up strategy, the timeliness and comprehensiveness of emergency response are ensured. In medium-risk areas, under the goal of efficiency optimization, the wake-up priorities are reasonably allocated to avoid energy consumption waste caused by over-wake-up. In low-risk areas, by reducing the weights of working conditions and immediate battery data, while ensuring basic monitoring, the energy consumption is significantly reduced, and low-value high-frequency wake-ups are reduced through the wake-up strategy, achieving the collaborative optimization of the fire management system in management.

[0182] In one case of this embodiment, based on a three-dimensional evaluation matrix, a fire management strategy is generated, including:

[0183] Analyze the three-dimensional evaluation matrix to obtain the comprehensive score of the fire area node. Specifically, by weighted fusion of the battery remaining life index, monitoring efficiency index, and energy consumption cost index, the comprehensive score of the fire area node can be obtained, and the value range of the comprehensive score is [0, 1];

[0184] Based on the comprehensive score, a fire management strategy is generated, and the fire management strategy includes: battery health management strategy, battery dynamic power consumption regulation strategy;

[0185] Among them, the battery health management strategy is as follows:

[0186] When the node is of high priority and the comprehensive score ≥ 0.85, the sampling frequency is increased to 1 minute / time at this time, and the battery is marked for replacement within 30 days;

[0187] When the node is of medium priority and 0.5 ≤ comprehensive score < 0.85, maintain the sampling at 10 minutes / time;

[0188] When the node is of low priority and the comprehensive score < 0.5, the sampling frequency is reduced to 6 hours / time, and only the basic monitoring of the battery SOC is retained;

[0189] The specific battery dynamic power consumption regulation strategy is as follows:

[0190] For high-risk fire areas, the master node is always active, and the slave node wakes up every 30 minutes;

[0191] For medium-risk fire areas, the master node works intermittently (wakes up every 10 minutes), and the slave node dynamically adjusts sampling according to the battery health score, as follows:

[0192] When the battery health score HF≥0.7 (battery is healthy), the slave node collects data at the secondary sampling frequency, and only uploads operating condition data (pressure, position) and environmental data (temperature, humidity). Battery data (SOC / HS) is collected at the tertiary sampling frequency (once every 30 minutes);

[0193] When the battery health score HF<0.7 (battery health is declining), the slave node collects all three types of data (environment, operating condition, battery) at the secondary sampling frequency, and gives priority to uploading the battery health status abnormal signal (such as SOC<0.3 or HS<0.6) when the master node is activated;

[0194] For low-risk fire areas, the master node is in deep sleep (activated 2 times a day), and the slave node wakes up emergently only when the temperature in the fire area >60°C;

[0195] By comprehensively scoring the fire area nodes through a three-dimensional evaluation matrix, the precision and scientific nature of fire management are realized. The core indicators such as battery remaining life, monitoring efficiency, and energy consumption cost are weighted and integrated, changing the one-sidedness of single-index evaluation, and being able to comprehensively reflect the actual state of the nodes. The battery health management strategy formulated based on the comprehensive score adopts differentiated management measures for high-, medium-, and low-priority nodes: high-priority nodes increase the sampling frequency and mark for replacement to ensure real-time monitoring and timely maintenance of key nodes, medium-priority nodes maintain regular monitoring to avoid excessive intervention, and low-priority nodes reduce the sampling frequency to save energy consumption, which not only guarantees the fire protection efficiency of key areas but also avoids resource waste, realizes the dynamic balance between monitoring accuracy and management cost, and improves the overall reliability of the fire protection system;

[0196] By implementing hierarchical management according to the risk levels of fire areas, the matching degree between system energy consumption and monitoring efficiency is effectively optimized. The master nodes in high-risk areas are fully active at all times, and the slave nodes are regularly awakened to ensure real-time response to abnormal situations and maximize fire safety protection. In medium-risk areas, the sampling frequency and data upload strategy of slave nodes are dynamically adjusted according to battery health. While ensuring the timely transmission of key data (such as battery anomaly signals), unnecessary energy consumption is reduced through differential sampling. In low-risk areas, the master nodes are deeply dormant, and the slave nodes are only awakened at high temperatures, significantly reducing system power consumption and extending the device's battery life. This hierarchical control mechanism not only meets the monitoring requirements of different risk areas but also reduces the overall energy consumption through intelligent sleep and wake-up strategies, realizing a positive interaction between the fire monitoring system in terms of efficiency guarantee and energy conservation and consumption reduction, providing technical support for long-term stable operation.

[0197] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent fire protection service management system, characterized in that, Including: A data acquisition unit, which is used to obtain three basic parameters of the environmental data, operating conditions data, and battery data of the fire protection objects in the office building, where the fire protection objects are fire extinguishers; A fusion unit, which is used to fuse the environmental data, operating conditions data, and battery data to generate a dynamic composite anomaly threshold for battery status perception; An adjustment unit, which is used to compare the operating conditions data, environmental data, and battery data with the dynamic composite anomaly threshold and execute a three-level sampling frequency adjustment mechanism based on the battery health; A management unit, which is used to divide the office building into fire protection areas managed by master and slave nodes, and execute a battery priority hierarchical wake-up strategy in combination with the distribution characteristics of the fire protection areas and the dynamic composite anomaly threshold; A balance unit, which is used to construct a three-dimensional evaluation matrix of battery remaining life - monitoring efficiency - energy consumption cost according to the battery data and the three-level sampling frequency adjustment mechanism, and generate a fire protection management strategy based on the three-dimensional evaluation matrix.

2. The intelligent fire protection service management system according to claim 1, wherein, Fusing the environmental data, operating conditions data, and battery data to generate a dynamic composite anomaly threshold for battery status perception, including: Performing fusion weighting on the normalized environmental data, operating conditions data, and battery data to obtain fusion data; Analyzing the fusion data to generate a dynamic composite anomaly threshold for battery status perception, specifically including: Calculating the average value of the fusion data in the historical period to obtain the historical average value; Calculating the standard deviation of the fusion data in the historical period to obtain the historical standard deviation; By scaling the product of the historical average value and the historical standard deviation, the scaling ratio is controlled by a dynamic adjustment coefficient, and finally the scaled result is superimposed on the historical mean to form a dynamic composite anomaly threshold; The dynamic adjustment coefficient is composed of the superposition of three parts, and the three parts respectively reflect the dynamic influence of the battery health status, environmental fluctuation characteristics, and operating conditions stability on the threshold.

3. The intelligent fire protection service management system according to claim 2, characterized in that, Comparing the operating conditions data, environmental data, and battery data with the dynamic composite anomaly threshold and executing a three-level sampling frequency adjustment mechanism based on the battery health, including: Calculating the battery data to obtain a battery health score; Calculating the data deviation degree of the operating conditions data, environmental data, and battery data from the dynamic composite anomaly threshold, specifically as follows: Calculating the absolute difference between the operating condition parameter and the dynamic composite anomaly threshold, and then dividing it by the dynamic composite anomaly threshold to obtain the relative deviation degree of the operating condition parameter; Calculating the absolute difference between the environmental parameter and the dynamic composite anomaly threshold, and then dividing it by the dynamic composite anomaly threshold to obtain the relative deviation degree of the environmental parameter; Calculating the absolute difference between the battery parameter and the dynamic composite anomaly threshold, and then dividing it by the dynamic composite anomaly threshold to obtain the relative deviation degree of the battery parameter; Among them, the weight value range of the relative deviation degree of the operating condition parameter is between 0.3 - 0.4, the weight value range of the relative deviation degree of the environmental parameter is between 0.2 - 0.3, and the weight value range of the relative deviation degree of the battery parameter is between 0.3 - 0.5, and the sum of the three weights is equal to 1; Adding the relative deviation degree of the operating condition parameter, the relative deviation degree of the environmental parameter, and the relative deviation degree of the battery parameter respectively according to their respective weights to obtain the data deviation degree; According to the battery health score and data deviation, the sampling frequency is divided into three levels, namely the first-level sampling frequency, the second-level sampling frequency and the third-level sampling frequency.

4. The intelligent fire protection service management system according to claim 1, characterized in that The office building is divided into fire protection zones managed by master and slave nodes, including: Divide the office building into multiple fire protection zones according to the floors, with each floor being a fire protection zone; Each fire protection area contains all the fire extinguishers on this floor, and each fire protection area is set up with a master node, and each master node is set up with multiple slave nodes; The master node is used to manage multiple slave nodes on the floor; The slave node is associated with multiple fire extinguishers, and collects environmental data, operating data, and battery data of the fire extinguishers in the area, and transmits the three basic parameters to the master node.

5. The intelligent fire protection service management system according to claim 4, wherein Combining the fire zone distribution characteristics with the dynamic composite abnormality threshold, a battery priority-level wake-up strategy is implemented, including: According to the fire protection zones divided by floors, the distribution characteristic parameters of each fire protection zone are extracted. The distribution characteristic parameters are composed of the fire extinguisher deployment density and the master-slave node distance; The fire extinguisher deployment density can be obtained by dividing the number of fire extinguishers in the fire protection area of each floor by the area of the fire protection area of that floor. The distance between each fire extinguisher and its master node is calculated using the Euclidean distance formula to obtain the master-slave node distance. The dynamic composite anomaly thresholds are associated with each region. The master node aggregates the dynamic composite anomaly thresholds reported by all slave nodes on the floor and calculates the regional average dynamic threshold, including: The master node collects the dynamic composite anomaly thresholds reported by all slave nodes on the floor. The interquartile range method is used to identify and eliminate outlier dynamic composite anomaly thresholds. The effective dynamic composite anomaly thresholds are spatially weighted averaged according to the fire extinguisher deployment density and distance weight in the area associated with each slave node. Finally, the regional average dynamic threshold that reflects the overall status of the fire protection system on the floor is generated.

6. The intelligent fire protection service management system according to claim 5, characterized in that, Combining the fire zone distribution characteristics with the dynamic composite abnormality threshold, a battery priority wake-up strategy is implemented, which also includes: The regional average dynamic threshold is combined with distribution characteristic parameters to generate a comprehensive regional risk score. Specifically, the following steps are performed: normalizing the fire extinguisher deployment density and the master-slave node distance. A multi-parameter coupling model is then established, using the regional average dynamic threshold as the core risk input, the deployment density as the inverse adjustment factor, and the master-slave node distance as the efficiency compensation term. The analytic hierarchy process is then used to determine the weights of each parameter. The regional average dynamic threshold and the distribution characteristic parameters are nonlinearly coupled using a weighted fusion algorithm to generate a quantitative risk score on a scale of 0-100. When the comprehensive risk score of a region is between 0 and 40 points, the fire protection area is a low-risk area; When the comprehensive risk score of a region is between 41 and 70 points, the fire protection area is a medium-risk area; When the comprehensive risk score of a region is between 71 and 100 points, the fire protection area is a high-risk area; Execute battery priority wake-up strategy based on comprehensive regional risk score.

7. The intelligent fire protection service management system according to claim 6, characterized in that, Execute battery priority wake-up strategy based on comprehensive regional risk scores, including: Analyze the region's comprehensive risk score and battery health score to generate a wake-up priority index; Execute the battery priority hierarchical wake-up strategy based on the wake-up priority index and the regional comprehensive risk score, specifically including: The battery priority hierarchical wake-up strategy includes primary wake-up, secondary wake-up, and tertiary wake-up.

8. An intelligent fire protection service management system according to claim 6, characterized in that, Construct a three-dimensional evaluation matrix of battery remaining life - monitoring efficiency - energy consumption cost according to the battery data and the three-level sampling frequency adjustment mechanism: Based on the battery data and the three-level sampling frequency adjustment mechanism, obtain the normalized battery remaining life index, monitoring efficiency index, and energy consumption cost index. Based on these three indexes, construct a three-dimensional evaluation matrix of battery remaining life - monitoring efficiency - energy consumption cost; Dynamically evaluate the monitoring efficiency and data deviation degree to generate the joint evaluation value of each node; Generate the monitoring priority of the fire area nodes according to the joint evaluation value of each node in the fire area, specifically including: When the joint evaluation value ≥ 0.85, it is determined as a high-priority node and real-time monitoring is started. When 0.5 ≤ joint evaluation value < 0.85, it is determined as a medium-priority node and the normal monitoring cycle is executed. When the joint evaluation value < 0.5, it is determined as a low-priority node and enters the standby monitoring mode; Dynamically adjust the weight of the three-dimensional evaluation matrix according to the risk level of the fire area.

9. A smart fire protection service management system according to claim 8, characterized in that, Dynamically adjust the weight of the three-dimensional evaluation matrix according to the risk level of the fire area, including: When the fire area is a high-risk area, increase the monitoring efficiency weight and reduce the energy consumption cost weight; When the fire area is a medium-risk area, maintain the default weight; When the fire area is a low-risk area, increase the energy consumption cost weight and reduce the monitoring efficiency weight.

10. A smart fire protection service management system according to claim 8, characterized in that, Generate a fire management strategy based on the three-dimensional evaluation matrix, including: Analyze the three-dimensional evaluation matrix to obtain the comprehensive score of the fire area nodes; Generate a fire management strategy based on the comprehensive score. The fire management strategy includes: battery health management strategy and battery dynamic power consumption regulation strategy; The battery health management strategy is as follows: When the node is of high priority and the comprehensive score ≥ 0.85, the sampling frequency is increased to 1 minute / time at this time, and the battery is marked for replacement within 30 days; When the node is of medium priority and 0.5 ≤ comprehensive score < 0.85, maintain the sampling at 10 minutes / time; When the node is of low priority and the comprehensive score < 0.5, the sampling frequency is reduced to 6 hours / time, and only the basic monitoring of the battery SOC is retained; The battery dynamic power consumption regulation strategy is as follows: For high-risk fire areas, the master node is active all the time, and the slave node wakes up every 30 minutes; For medium-risk fire areas, the master node works intermittently, and the slave node adjusts the sampling dynamically according to the battery health score; For low-risk fire areas, the master node is in deep sleep, and the slave node is only emergently woken up when the temperature in the fire area > 60°C.

Images