Intelligent fire-fighting linkage management system for data center in multi-station fusion transformer substation

Abstract

The invention provides a data center intelligent fire-fighting linkage management system in a multi-station fusion transformer substation, and relates to the technical field of multi-station fusion, and the system comprises a fire detection module which is used for collecting the smoke concentration, environment temperature and humidity and electrical parameters in data in real time, and obtaining original fire data; the control module is used for receiving the original fire behavior data and analyzing the smoke concentration, the environment temperature and humidity and the electrical parameters by using a spotted serow optimization algorithm so as to generate a fire behavior grade judgment result; and the automatic fire extinguishing module is used for triggering a collaborative fire extinguishing program of the heptafluoropropane gas fire extinguishing device and the water mist fire extinguishing device according to the fire behavior grade judgment result. According to the invention, through efficient and accurate fire detection, cooperative fire extinguishing and emergency linkage, the fire response efficiency and safety are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-station integration, and particularly to an intelligent fire linkage management system for a data center in a multi-station integrated substation. Background Art

[0002] In the event of a fire, the traditional fire linkage system has deficiencies in fire monitoring. For example, in the data center of a certain multi-station integrated substation, when a fire is caused by aging of the lines near the transformer, the traditional smoke detector may not be able to detect the slow increase in smoke concentration in a timely manner due to the installation position or sensitivity, resulting in a delay in discovering the fire. This is because traditional detectors mostly use fixed threshold judgments, making it difficult to adapt to the complex and changeable operating environment of electrical equipment and unable to accurately perceive the initial weak fire changes.

[0003] When dealing with fires threatening core equipment, traditional technologies lack pertinence in formulating fire extinguishing strategies. Suppose a fire breaks out in the switch cabinet in the data center due to overload. The traditional fire extinguishing system may uniformly initiate full-area fire extinguishing measures, such as spraying a large amount of fire extinguishing agent throughout the computer room at the same time. However, although this method can extinguish the fire, a large amount of agent will cause unnecessary corrosion and damage to the surrounding normally operating equipment. Especially for the UPS battery pack, the traditional fire extinguishing method cannot adopt appropriate fire extinguishing means according to the special chemical properties and protection requirements, and cannot protect the safety of these core equipment to the greatest extent while extinguishing the fire. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide an intelligent fire linkage management system for a data center in a multi-station integrated substation, which improves the fire safety guarantee ability and promotes the collaborative management of multi-station integration.

[0005] To solve the above technical problems, the technical solution of the present invention is as follows:

[0006] In a first aspect, an intelligent fire linkage management system for a data center in a multi-station integrated substation includes:

[0007] A fire detection module, which is used to collect the smoke concentration, ambient temperature and humidity, and electrical parameters in the data in real time, set monitoring parameters for the transformer, switch cabinet, and uninterruptible power supply battery pack, and obtain the original fire data;

[0008] A control module, which is used to determine the fire level by screening out the final combination of smoke concentration, temperature and humidity, and electrical parameters based on the spotted hyena optimization algorithm by setting population parameters and iteratively updating the individual fitness values, and automatically trigger the protection protocol when the evaluation result threatens the core equipment;

[0009] An automatic fire extinguishing module, which is used to trigger the coordinated fire extinguishing procedures of the heptafluoropropane gas fire extinguishing device and the fine water mist fire extinguishing device according to the fire level determination result;

[0010] A cloud platform module, which is used to upload the start signal of the collaborative fire extinguishing program and real-time fire scene data to the cloud, generate a three-dimensional heat map, and analyze the three-dimensional heat map to generate a fire alarm instruction;

[0011] An emergency linkage module, which is used to execute non-fire power cut-off, fire shutter gradient speed reduction and closing, emergency lighting path planning and evacuation broadcast linkage operations according to the fire alarm instruction.

[0012] Furthermore, based on the spotted hyena optimization algorithm, by setting population parameters and iteratively updating individual fitness values, the final combination of smoke concentration, temperature and humidity, and electrical parameters is selected to determine the fire level. When the evaluation result threatens the core equipment, the protection protocol is automatically triggered, including:

[0013] Set the initial parameters of the spotted hyena optimization algorithm, including population size, number of iterations, and search range;

[0014] For each individual in the population, that is, the combination of smoke concentration, oil temperature and resistance value parameters, by calculating the actual-to-reference ratio of smoke concentration, oil temperature, and resistance value and the historical fire frequency, and fusing the contribution values of multi-source parameters, calculate the fitness value corresponding to each individual, and update the individuals in the population according to the fitness value, including selection, crossover, and mutation operations;

[0015] Repeat calculating the fitness value of the individual, and perform selection and update operations according to the fitness value, continuously optimize the population until the preset number of iterations is reached, and determine the final solution from the population according to the fitness value of the individual, including the combination of smoke concentration, environmental temperature and humidity, and electrical parameters; generate a fire level determination result according to the parameter combination of the final solution;

[0016] Evaluate the fire level determination result. If it shows a threat to the transformer, switchgear, and uninterruptible power supply battery pack, the core equipment priority protection protocol is automatically triggered and relevant protection measures are started.

[0017] Furthermore, for each individual in the population, that is, the combination of smoke concentration, oil temperature and resistance value parameters, by calculating the actual-to-reference ratio of smoke concentration, oil temperature, and resistance value and the historical fire frequency, and fusing the contribution values of multi-source parameters, calculate the fitness value corresponding to each individual, including:

[0018] Obtain the actual smoke concentration value, calculate the proportional relationship according to the actual smoke concentration value and the preset maximum smoke concentration value, and calculate the smoke concentration contribution value according to the proportional relationship;

[0019] Obtain the current oil temperature value, calculate the oil temperature ratio of the current oil temperature value to the maximum allowable change range of the oil temperature of the transformer under normal operating conditions, and determine the oil temperature contribution value according to the oil temperature ratio;

[0020] Obtain the resistance value of the current UPS battery pack, and calculate the resistance ratio of the resistance value to the reference resistance value of the battery under normal conditions;

[0021] Extract relevant data from the historical fire record database of the data center and conduct analysis to calculate the frequency value of past fires in the area;

[0022] Fuse the smoke concentration contribution value, oil temperature contribution value, resistance ratio, and frequency value to obtain the fitness value.

[0023] Furthermore, the fire situation level determination results include no fire, minor fire, moderate fire, and severe fire.

[0024] Furthermore, according to the fire situation level determination result, trigger the coordinated fire extinguishing program of the heptafluoropropane gas fire extinguishing device and the water mist fire extinguishing device, including:

[0025] Match the corresponding fire extinguishing strategy from the preset fire extinguishing strategy library, including starting the water mist fire extinguishing device for minor fires and starting both the heptafluoropropane gas fire extinguishing device and the water mist fire extinguishing device for severe fires;

[0026] According to the fire extinguishing strategy, send a start command to release the heptafluoropropane gas into the fire area. At the same time, the water mist fire extinguishing device starts the water pump to spray water into the fire area;

[0027] During the fire extinguishing process, adjust the fire extinguishing strategy and the coordinated fire extinguishing program in real time according to the on-site temperature and smoke concentration.

[0028] Furthermore, upload the start signal of the coordinated fire extinguishing program and the real-time data of the fire scene to the cloud, generate a three-dimensional heat map, and analyze the three-dimensional heat map to generate a fire alarm instruction, including:

[0029] Sort out the start signal and the real-time data of the fire scene to form a data set;

[0030] Upload the data set to the cloud platform and extract the temperature and smoke concentration information of the fire scene;

[0031] According to the temperature of the fire scene, combined with the geographical information and building layout of the fire scene, construct a three-dimensional model of the fire scene to generate a heat map;

[0032] According to the heat map, analyze the real-time data of the fire scene, including predicting the trend of fire spread and evaluating the fire extinguishing effect, to obtain a comprehensive analysis report on the current situation of the fire scene;

[0033] According to the comprehensive analysis report and the data analysis results, generate a fire alarm instruction, including adjusting the fire extinguishing strategy, dispatching additional rescue forces, and evacuating the surrounding personnel.

[0034] Further, based on the heat map, real-time fire data analysis is performed, including prediction of the fire spread trend and evaluation of the fire extinguishing effect, to obtain a comprehensive analysis report on the current situation of the fire scene, including:

[0035] According to the change of temperature over time in different regions of the heat map, identify and mark the regions where the temperature continues to rise;

[0036] Conduct an analysis of the temperature gradient distribution in the regions where the temperature continues to rise to determine the direction of heat propagation, that is, the fire spread trend;

[0037] According to the fire spread trend, analyze the temperature change in the high-temperature regions before and after fire extinguishing, evaluate the degree of temperature drop in the high-temperature regions after the implementation of fire extinguishing measures, and obtain a preliminary evaluation result;

[0038] Integrate the fire spread trend, preliminary evaluation results, and abnormal values of smoke concentration and electrical parameters in the real-time data to obtain a comprehensive analysis report on the current situation of the fire scene.

[0039] Further, it also includes an emergency linkage module for verifying fire alarm instructions, including the location of the fire and the danger level; determining the scope of non-fire-fighting power sources to be cut off according to the location of the fire, and sequentially cutting off the power of the corresponding circuits through controlling circuit breakers and contactor devices; obtaining the current position and operating status of the fire shutter according to the fire alarm instruction, formulating a closing strategy, and sending an instruction to the controller to reduce the closing speed according to the closing strategy; planning multiple safe evacuation routes according to the location of the fire, and adjusting the brightness and flashing frequency of emergency lighting fixtures; formulating evacuation broadcast content according to the fire alarm instruction and multiple safe evacuation routes, including evacuation instructions and safety tips.

[0040] In a second aspect, a computing device includes:

[0041] One or more processors;

[0042] A storage device for storing one or more programs, which when executed by the one or more processors cause the one or more processors to implement the system described above.

[0043] In a third aspect, a computer-readable storage medium stores a program that, when executed by a processor, implements the system.

[0044] The above solution of the present invention has at least the following beneficial effects:

[0045] Based on multi-source data fusion and intelligent analysis technology, it can keenly capture minor anomalies in the data center of the substation, issue accurate fire warnings in advance, strive for the golden time for personnel evacuation and initial fire extinguishing, and significantly reduce the risk of fire occurrence. When a fire occurs, it can quickly link multiple fire extinguishing devices, take precise measures according to the fire situation, quickly control the spread of the fire, and effectively reduce the damage of the fire to the equipment and data security of the data center.

[0046] It supports operation and maintenance personnel to remotely monitor the status of fire-fighting equipment and fire information in real time, eliminating the need for frequent on-site inspections, reducing labor costs, and improving the response speed and handling efficiency for sudden fire situations. It deeply analyzes the operation data of fire-fighting equipment, warns of equipment failures in advance, and generates a scientific maintenance plan.

[0047] In the multi-station integration scenario, it realizes the sharing and linkage response of fire-fighting information among various stations, and constructs an all-round fire safety network. Through precise early warning and efficient fire extinguishing, it reduces the damage of the fire to the equipment in the data center of the substation, reduces the risk of service interruption and data loss caused by the fire, and ensures the stable operation of the substation and the data center. According to the fire-fighting needs and resource conditions of each station, it intelligently allocates fire-fighting equipment and personnel, realizes the efficient utilization of resources, and improves the overall fire management efficiency. Brief Description of the Drawings

[0048] Figure 1 It is a schematic diagram of an intelligent fire-fighting linkage management system for the data center in a multi-station integrated substation provided by an embodiment of the present invention.

[0049] Figure 2 It is a schematic diagram of the process of uploading the start signal of the collaborative fire extinguishing program and the real-time fire data in an intelligent fire-fighting linkage management system for the data center in a multi-station integrated substation provided by an embodiment of the present invention to the cloud, generating a three-dimensional heat map, and analyzing the three-dimensional heat map to generate a fire alarm instruction. Detailed Embodiment

[0050] Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art.

[0051] As Figure 1 shown, an embodiment of the present invention proposes an intelligent fire-fighting linkage management system for the data center in a multi-station integrated substation, including:

[0052] The fire detection module 1 is used to collect the smoke concentration, ambient temperature and humidity, and electrical parameters in the data in real time, set the monitoring parameters for transformers, switch cabinets, and uninterruptible power supply battery packs, and obtain the original fire data;

[0053] The control module 2 is used to optimize the algorithm based on the spotted hyena, set the population parameters and iteratively update the individual fitness values, screen out the final combination of smoke concentration, temperature and humidity, and electrical parameters to determine the fire level, and automatically trigger the protection protocol when the evaluation result threatens the core equipment;

[0054] The automatic fire extinguishing module 3 is used to trigger the coordinated fire extinguishing procedures of the heptafluoropropane gas fire extinguishing device and the fine water mist fire extinguishing device according to the fire level determination result;

[0055] The cloud platform module 4 is used to upload the start signal of the coordinated fire extinguishing procedure and the real-time fire data to the cloud, generate a three-dimensional thermal map, analyze the three-dimensional thermal map, and generate a fire alarm instruction;

[0056] The emergency linkage module 5 is used to execute the operations of cutting off non-fire power supplies, gradually reducing the speed of fireproof rolling shutters, planning emergency lighting paths, and linking evacuation broadcasts according to the fire alarm instructions.

[0057] In the embodiment of the present invention, the fire detection module 1 collects multi-dimensional data of smoke concentration, ambient temperature and humidity, and electrical parameters in real time, obtains the original fire data, can comprehensively and accurately capture the subtle signs before the fire occurs and the key information during the fire, improves the early detection ability of the fire, reduces the fire hazards, and effectively reduces the probability of fire occurrence.

[0058] The control module 2 uses the spotted hyena optimization algorithm to deeply analyze the smoke concentration, ambient temperature and humidity, and electrical parameters, and generates a scientific and reasonable fire level determination result. The spotted hyena optimization algorithm has self-adaptability and optimization ability, and can make more accurate judgments according to different scenarios and fire characteristics. Quickly and accurately determining the fire level helps to shorten the time interval from discovering the fire to taking corresponding measures, improves the efficiency of fire response, and reduces the losses caused by the fire.

[0059] The automatic fire extinguishing module 3, the heptafluoropropane gas fire extinguishing device has the advantages of fast fire extinguishing speed and no damage to equipment, while the fine water mist fire extinguishing device has a comprehensive fire extinguishing effect of cooling, suffocation, and heat insulation. The two work together to give full play to their respective advantages, achieve efficient fire extinguishing for different types and scales of fires, and improve the fire extinguishing success rate. Automatically triggering the fire extinguishing procedure without manual intervention can quickly respond at the moment of the fire occurrence, accurately carry out fire extinguishing operations on the fire source, and reduce the harm of the fire to people and property.

[0060] The cloud platform module 4 uploads the start signal of the collaborative fire extinguishing program and the real-time fire data to the cloud, enabling managers and relevant personnel to remotely monitor the fire situation at any time and anywhere, and promptly grasp the fire dynamics. By generating a three-dimensional heat map, it intuitively displays the temperature distribution of the fire scene, providing more comprehensive and intuitive information for decision-makers and helping to make more scientific and reasonable decisions.

[0061] The emergency linkage module 5 forms an all-round emergency guarantee system through the mutual cooperation of emergency measures. Cutting off non-fire-fighting power sources can prevent the spread of electrical fires, and gradually reducing the speed to close the fireproof rolling shutters can effectively prevent the spread of fire and smoke. The emergency lighting path planning provides a safe guide for personnel evacuation, and the evacuation broadcast linkage operation promptly notifies personnel to evacuate, ensuring the safety of personnel lives.

[0062] In a preferred embodiment of the present invention, based on the spotted hyena optimization algorithm, by setting population parameters and iteratively updating the individual fitness values, the final combination of smoke concentration, temperature, humidity, and electrical parameters is selected to determine the fire level. When the evaluation result threatens the core equipment, the protection protocol can be automatically triggered, which may include:

[0063] Set the initial parameters of the spotted hyena optimization algorithm, including population size, number of iterations, and search range;

[0064] For each individual in the population, that is, the combination of smoke concentration, oil temperature, and resistance value parameters, by calculating the actual and reference ratios of smoke concentration, oil temperature, and resistance value and the historical fire frequency, the contribution values of multi-source parameters are fused to calculate the fitness value corresponding to each individual. And according to the fitness value, the individuals in the population are updated, including selection, crossover, and mutation operations;

[0065] Repeat the calculation of the fitness value of the individual, and perform selection and update operations according to the fitness value, continuously optimizing the population until the preset number of iterations is reached. Determine the final solution from the population according to the fitness value of the individual, including the combination of smoke concentration, environmental temperature, humidity, and electrical parameters; generate a fire level determination result according to the parameter combination of the final solution;

[0066] Evaluate the fire level determination result. If it shows a threat to transformers, switchgear, and uninterruptible power supply battery packs, the core equipment priority protection protocol will be automatically triggered and relevant protection measures will be started.

[0067] In the embodiments of the present invention, the population size is determined by comprehensively considering the complexity of the data and the computing resources. The data complexity is reflected in the diversity of the data, the correlation between parameters, and the degree of dispersion of the data distribution. If the data is complex, it means that there may be multiple local final solutions, and more individuals are needed to comprehensively explore the parameter space. At this time, the population size can be appropriately increased. For example, when historical data shows that there is a complex non-linear relationship between the smoke concentration, ambient temperature and humidity, and electrical parameters, and the data distribution is relatively scattered, the population size can be set to 100-150 individuals. If the computing resources are limited, the memory is small, and the CPU processing power is weak, an overly large population size will significantly increase the computing time and may even lead to crashes. At this time, the population size needs to be reduced and set to 30-50 individuals.

[0068] The setting of the number of iterations is crucial, directly affecting whether the algorithm can converge to the final solution and the utilization efficiency of computing resources. If the number of iterations is too small, the algorithm may not fully converge to the final solution, resulting in inaccurate results; if the number of iterations is too large, a large amount of computing resources will be wasted. For the problem of fire risk level determination, the number of iterations can be set according to the complexity of the problem and historical experience. Generally, the number of iterations can be set to 100-200 times. If the problem is relatively complex, for example, the relationship between parameters is very complex, and historical data shows that more iterations are required to find the final solution, the number of iterations can be appropriately increased to 200-300 times; if the problem is relatively simple, the number of iterations can be reduced to 50-100 times.

[0069] The search range determines the value range of the smoke concentration, ambient temperature and humidity, and electrical parameters. For the smoke concentration, the minimum and maximum values can be determined based on the smoke concentration data during past fires and normal operations. Historical data shows that the minimum smoke concentration is 0.1 ppm under normal conditions and the maximum reaches 10 ppm during a fire. Then, the search range can be set to [0.1, 10] ppm. For the ambient temperature and humidity, considering the actual operating environment of the substation data center, the normal temperature range may be 15-35 °C, and the humidity range is 30%-70%. Then, the temperature search range can be set to [15, 35] °C, and the humidity search range can be set to [30%, 70%]. For electrical parameters, including current and voltage, the search range can be determined based on the rated parameters of the equipment and the fluctuation range during actual operation. The rated current of the electrical equipment is 200 A, and the current fluctuation range during actual operation is 100-250 A. Then, the current search range can be set to [100, 250] A.

[0070] For each individual in the population, a fitness function is constructed to evaluate its quality. The fitness function is constructed based on historical fire data. By analyzing the fire occurrence probabilities corresponding to different parameter combinations (smoke concentration, environmental temperature and humidity, and electrical parameters), the fitness values are obtained. A large amount of historical fire data is collected, including records of smoke concentration, environmental temperature and humidity, electrical parameters, and whether a fire occurred at different time points. The historical data is preprocessed, including removing outliers and normalizing, to improve the data quality and comparability. The historical data is classified according to different parameter combinations, and the number of fire occurrences and the total number of records under each parameter combination are counted. According to the statistical results, the fire occurrence probability corresponding to each parameter combination is calculated, which will be used as the fitness value of the parameter combination.

[0071] Based on the fitness values, selection operations are performed on the individuals in the population. The purpose is to determine the individual with the maximum fitness as the parent and pass on the excellent genes to the next generation. The roulette wheel selection method is used to calculate the proportion of the fitness value of each individual in the total fitness value as the probability of the individual being selected, and a random number between 0 and 1 is generated. According to the interval where the random number falls, the corresponding individual is determined. If the proportion of the fitness value of individual A in the total fitness value is 0.2, the probability of being selected is 0.2. If the random number falls within the interval of 0 - 0.2, then individual A is selected as the parent. Two individuals are randomly determined from the selected parent individuals as the parents for the crossover operation. The purpose of the crossover operation is to exchange part of the genes of the parent individuals to generate new offspring individuals and increase the diversity of the population. According to the single-point crossover method, a crossover point is randomly determined, and the genes of the two parent individuals after the crossover point are exchanged to generate two new offspring individuals. The gene sequence of parent individual A is [1, 2, 3, 4, 5], and the gene sequence of parent individual B is [6, 7, 8, 9, 10]. If the crossover point is 3, then the genes after the crossover point are exchanged, and the gene sequence of offspring individual C is [1, 2, 3, 9, 10], and the gene sequence of offspring individual D is [6, 7, 8, 4, 5].

[0072] For the generated offspring individuals, perform mutation operations with a certain mutation probability. The purpose of the mutation operation is to obtain new genes and prevent the algorithm from falling into a local optimal solution. The mutation probability is generally set to a small value, including 0.01 - 0.1. For each gene of each offspring individual, generate a random number between 0 and 1. If the random number is less than the mutation probability, perform a mutation operation on the gene. The mutation operation can be to randomly change the value of the gene to randomly take a value within the search range. If the gene of the offspring individual represents the smoke concentration and the value is 2 ppm, and the search range is [0.1, 10] ppm, if the gene needs to mutate, a value can be randomly determined within the range of [0.1, 10] ppm as the new smoke concentration value. During each iteration process, record the individual with the maximum fitness value and the fitness value in the current population. As the number of iterations increases, the overall fitness of the population will gradually increase and finally converge to a relatively stable state.

[0073] When the preset number of iterations is reached, the algorithm stops. Determine the final solution from the population according to the individual fitness values, that is, the parameter combination (smoke concentration, environmental temperature and humidity, and electrical parameters) corresponding to the individual with the maximum fitness value. Compare the parameter combination of the final solution with the preset fire level standard to generate a fire level determination result.

[0074] Suppose in the data center of a multi-station integrated substation, the original fire data collected by the fire detection module is as follows: the smoke concentration fluctuates between 30% - 80%, the environmental temperature is between 25°C - 60°C, the environmental humidity is between 40% - 70%, and the current value in the electrical parameters is between 10 A - 50 A. Set the population size to 80 individuals and the number of iterations to 150 times. The search range of the smoke concentration is set to 0 - 100%, the search range of the environmental temperature is set to -20°C - 80°C, the search range of the environmental humidity is set to 0% - 100%, and the search range of the current value is set to 0 A - 100 A.

[0075] Construct a fitness function, calculate the fitness value of each individual, and perform single-point crossover and mutation operations to continuously update the population. After 150 iterations, the individual with the maximum fitness value is obtained, and the parameter combination is smoke concentration 75%, environmental temperature 55°C, environmental humidity 50%, and current value 45 A. According to the preset fire level standard, it is determined as a high-level fire and corresponding fire extinguishing measures need to be immediately activated.

[0076] Through continuous iteration and optimization, the Spotted Hyena Optimization Algorithm can find the parameter combination of the final solution in a complex parameter space. By comprehensively considering multiple factors such as smoke concentration, environmental temperature and humidity, and electrical parameters, it can more accurately determine the fire level, avoiding the limitations of single-parameter judgment. According to different data center environments and historical data, flexibly adjusting the initial parameters and fitness function can adapt to various complex situations, improving generality and adaptability. Accurate determination of the fire level can rationally allocate fire-fighting resources, avoid waste of resources, discover potential fire hazards through real-time analysis and optimization of the original fire data, and issue early warnings in advance, striving for more time to take preventive measures and reducing the risk and losses of fires.

[0077] In a preferred embodiment of the present invention, for each individual in the population, by calculating the actual-to-reference ratios of smoke concentration, oil temperature, and resistance value, as well as the historical fire frequency, and fusing the contribution values of multi-source parameters to calculate the fitness value corresponding to each individual, and according to the fitness value, updating the individuals in the population, including selection, crossover, and mutation operations, may include:

[0078] Obtain the actual smoke concentration value, calculate the proportional relationship according to the actual smoke concentration value and the preset maximum smoke concentration value, and calculate the smoke concentration contribution value according to the proportional relationship;

[0079] Obtain the current oil temperature value, calculate the oil temperature ratio of the current oil temperature value to the maximum allowable change range of the oil temperature of the transformer under normal operating conditions, and determine the oil temperature contribution value according to the oil temperature ratio;

[0080] Obtain the resistance value of the current UPS battery pack, and calculate the resistance ratio of the resistance value to the reference resistance value of the battery under normal conditions;

[0081] Extract relevant data from the historical fire record database of the data center and perform analysis to calculate the frequency value of fires that occurred in the area in the past;

[0082] Fuse the smoke concentration contribution value, oil temperature contribution value, resistance ratio, and frequency value to obtain the fitness value.

[0083] In an embodiment of the present invention, for the convenience of calculation, it is necessary to normalize each value participating in the calculation, and then perform the operation. Real-time monitor the environmental smoke concentration, obtain the actual smoke concentration value, denoted as According to the data center environment, preset the maximum smoke concentration value in the current environment, that is By calculating Obtain the proportional relationship between the current smoke concentration and the maximum value. Then subtract the proportional relationship from 1, that is , this result reflects the degree of deviation of the current smoke concentration from the maximum value. The greater the degree of deviation, the lower the current smoke concentration and the safer the environment. Finally, multiply the degree of deviation by a preset coefficient = 0.3, that is , determine the contribution value of the smoke concentration. Part of the calculation is based on the real-time smoke concentration, reasonably evaluate the environmental safety status, and flexibly adjust the importance of the smoke concentration to the overall evaluation through the coefficient . Continuously monitor the oil temperature of the transformer, obtain the current oil temperature value, denoted as . According to the performance parameters of the transformer and the requirements for safe operation, set a threshold for the change in oil temperature, that is . It represents the maximum allowable range of change in the oil temperature of the transformer under normal operating conditions. Calculate . The ratio reflects the current oil temperature of the transformer compared to the safety threshold. The larger the ratio, the higher the oil temperature relative to the threshold, and the greater the risk the transformer may face. Multiply the ratio by a preset coefficient = 0.25, that is , determine the contribution value of the oil temperature.

[0084] Measure the resistance of the UPS battery pack in real time, obtain the current resistance value, denoted as . According to the factory parameters of the battery and the normal working standards, determine the reference resistance value of the battery under normal conditions, denoted as . Calculate . The ratio reflects the change in the current battery resistance relative to the normal reference resistance. The farther the ratio deviates from 1, the more abnormal the battery state. Extract relevant data from the historical fire record database of the data center and analyze it to calculate the frequency value of past fires in the area, denoted as . Multiply the value by a preset coefficient = 0.15, that is , determine the resistance ratio of the UPS battery pack. Add the resistance ratio of the UPS battery pack to the result of multiplying the historical fire factor by the coefficient , that is , to obtain a value that comprehensively reflects the battery state and the historical fire situation. Then multiply this comprehensive value by the coefficient = 0.3, that is , determine the frequency value of past fires in the area. Finally, add the smoke concentration contribution value , the oil temperature contribution value , the resistance ratio, and the frequency value . Add these three values together to obtain the fitness value, that is .

[0085] Through smoke concentration, transformer oil temperature, UPS battery pack resistance, and historical fire factors, a comprehensive assessment of the fire safety status of the data center in a multi-station integrated substation is achieved. When judging whether a region is safe, it no longer relies on single smoke concentration detection, but combines the operating status of equipment (transformer oil temperature, battery resistance) and historical fire information, making the assessment results more accurate and reliable. By real-time monitoring and calculating various factors, potential risks can be captured at the early stage of a fire. For example, through continuous monitoring and calculation of the transformer oil temperature, when the oil temperature gradually approaches the threshold, the fitness value will decrease accordingly, and an early warning can be issued in a timely manner to prompt the operation and maintenance personnel to take measures to avoid the occurrence and spread of the fire, effectively improving the fire prevention ability of the data center.

[0086] In a preferred embodiment of the present invention, the fire situation level determination results include no fire, minor fire, moderate fire, and severe fire.

[0087] In the embodiment of the present invention, features related to the fire situation level are determined from the cleaned and normalized data, including smoke concentration, the change rate of ambient temperature, and the degree of abnormality of electrical parameters. Analyze the features to understand the distribution rules and change trends under different fire situation levels.

[0088] Determine deep learning models for fire situation level determination, including decision trees, random forests, support vector machines, and neural networks. Use historical fire situation data as the training set to train the selected models. During the training process, use the known fire situation levels as labels to enable the models to learn the mapping relationship between features and fire situation levels, so as to obtain trained models.

[0089] Evaluate the trained models, calculate the accuracy rate, recall rate, and F1 value metrics of the models. Input the real-time collected and preprocessed fire situation data into the trained determination models, and perform calculations and judgments based on the input data to obtain the corresponding fire situation levels, including no fire, minor fire, moderate fire, and severe fire.

[0090] In a preferred embodiment of the present invention, according to the fire situation level determination results, triggering the coordinated fire extinguishing procedures of the heptafluoropropane gas fire extinguishing device and the water mist fire extinguishing device may include:

[0091] Match the corresponding fire extinguishing strategies from the preset fire extinguishing strategy library, including starting the water mist fire extinguishing device for minor fires and starting both the heptafluoropropane gas fire extinguishing device and the water mist fire extinguishing device for severe fires;

[0092] According to the fire extinguishing strategy, send a start command to release the heptafluoropropane gas into the fire area. At the same time, the water mist fire extinguishing device starts the water pump to spray water into the fire area;

[0093] During the fire extinguishing process, the fire extinguishing strategy and the coordinated fire extinguishing procedure are adjusted in real time according to the on-site temperature and smoke concentration.

[0094] In the embodiment of the present invention, according to the fire level, a specific fire extinguishing strategy is searched in the preset fire extinguishing strategy library. For a minor fire, a fine water mist fire extinguishing device is activated. For a severe fire, a heptafluoropropane gas fire extinguishing device and a fine water mist fire extinguishing device are activated simultaneously. A start command is sent and the fire extinguishing operation is executed, including signals for valve opening and commands for pump startup. The start command is sent to the heptafluoropropane gas fire extinguishing device to trigger the corresponding valve to open, enabling the heptafluoropropane gas to be quickly released into the fire area. For the fine water mist fire extinguishing device, the command makes the pump start working, and water is sprayed into the fire area through pipelines and nozzles.

[0095] During the fire extinguishing process, sensors distributed in the data center continuously collect on-site temperature and smoke concentration data and transmit the data back to the intelligent fire-fighting linkage management system. Based on the received real-time data, the temperature and smoke concentration are analyzed. If the temperature continues to rise or the smoke concentration does not decrease significantly, it indicates that the current fire extinguishing strategy is not effective. At this time, according to the preset rules and algorithms, combined with the real-time data, it is judged whether it is necessary to adjust the fire extinguishing strategy and the coordinated fire extinguishing procedure. If adjustment is needed, a more suitable fire extinguishing strategy is rematched from the fire extinguishing strategy library.

[0096] Suppose a fire breaks out in the data center of a multi-station integrated substation. The intelligent fire-fighting linkage management system analyzes the smoke concentration, ambient temperature and humidity, and electrical parameters, determines it as a severe fire, and quickly matches the strategy for severe fires in the fire extinguishing strategy library, that is, simultaneously activating the heptafluoropropane gas fire extinguishing device and the fine water mist fire extinguishing device. The intelligent fire-fighting linkage management system generates start commands and sends them to the heptafluoropropane gas fire extinguishing device and the fine water mist fire extinguishing device respectively. The valve of the heptafluoropropane gas fire extinguishing device quickly opens, and a large amount of heptafluoropropane gas is released into the fire area; the pump of the fine water mist fire extinguishing device starts, and water is sprayed into the fire area. The two fire extinguishing agents act simultaneously to try to control the fire.

[0097] During the fire extinguishing process, the data collected by the sensors shows that although the two fire extinguishing agents are working simultaneously, the temperature in the fire area is still rising slowly and the smoke concentration is not decreasing significantly. After analyzing the data, the intelligent fire-fighting linkage management system determines that the current fire extinguishing strategy is not ideal. Therefore, a strategy is rematched from the fire extinguishing strategy library, and it is decided to increase the injection volume of the heptafluoropropane gas and adjust the injection angle and flow rate of the fine water mist. The intelligent fire-fighting linkage management system sends new commands to adjust the two fire extinguishing devices. The adjusted fire extinguishing devices continue to work, and the fire is gradually controlled, the temperature begins to drop, and the smoke concentration also decreases significantly.

[0098] By matching corresponding fire extinguishing strategies according to different fire levels, precise strikes against fires are achieved. For minor fires, activating the fine water mist fire extinguishing device can quickly extinguish the fire when the fire is small, avoiding waste of fire extinguishing agents. For severe fires, two fire extinguishing devices are activated simultaneously. Leveraging the different fire extinguishing advantages of heptafluoropropane gas and fine water mist, they cooperate to improve the fire extinguishing efficiency, more effectively control the spread of the fire, and protect the equipment and personnel safety in the data center. During the fire extinguishing process, the fire extinguishing strategy and the cooperative fire extinguishing procedure are adjusted in real time according to the on-site temperature and smoke concentration, enabling the fire extinguishing work to adapt to the dynamic changes of the fire. This dynamic adjustment mechanism improves the fire extinguishing success rate and reduces the losses caused by the fire. The coordinated use of the heptafluoropropane gas fire extinguishing device and the fine water mist fire extinguishing device can protect the equipment in the data center while extinguishing the fire. After the heptafluoropropane gas extinguishes the fire, no residue is left, causing less damage to the equipment. The fine water mist fire extinguishing has the functions of cooling, temperature reduction, and smoke suppression, which can reduce the thermal damage and smoke erosion of the equipment caused by the fire.

[0099] In a preferred embodiment of the present invention, the start signal of the cooperative fire extinguishing procedure and the real-time fire data are uploaded to the cloud to generate a three-dimensional heat map, and the three-dimensional heat map is analyzed to generate a fire alarm instruction, which may include:

[0100] Sort out the start signal and the real-time fire data to form a data set;

[0101] Upload the data set to the cloud platform to extract the temperature and smoke concentration information of the fire scene;

[0102] According to the temperature of the fire scene, combined with the geographical information and building layout of the fire scene, construct a three-dimensional model of the fire scene to generate a heat map;

[0103] Based on the heat map, analyze the real-time fire data, including predicting the trend of fire spread and evaluating the fire extinguishing effect, to obtain a comprehensive analysis report on the current situation of the fire scene;

[0104] According to the comprehensive analysis report and the data analysis results, generate a fire alarm instruction, including adjusting the fire extinguishing strategy, dispatching additional rescue forces, and evacuating the surrounding personnel.

[0105] In the embodiments of the present invention, the collected startup signals and real-time fire data are classified, cleaned, and formatted to remove invalid data and noise, forming a complete data set, and the key information of the temperature and smoke concentration in the fire scene is extracted and uploaded to the cloud platform. The geographical information and building layout information of the area where the fire scene is located are obtained from a pre-stored database or a relevant Geographic Information System (GIS), including the shape, size, floor distribution, and passage location of the building. The extracted temperature data of the fire scene is combined with the geographical information and building layout information, and a 3D model of the fire scene is constructed using 3D modeling software (Blender, 3dsMax) or a dedicated geographic information processing library (GeoTools). In the model, different temperature regions can be represented by different colors or heights to visually display the temperature distribution. Based on the constructed 3D model, a 3D heat map is generated using data visualization techniques (Matplotlib, Seaborn libraries in Python, or D3.js library in JavaScript).

[0106] By analyzing the temperature change of different regions in the heat map over time, combined with the principles of fire dynamics and meteorological conditions, the spreading direction and speed of the fire are predicted. According to the temperature change of the area where fire extinguishing measures have been taken in the heat map, the temperature data before and after fire extinguishing are compared and analyzed to evaluate the fire extinguishing effect. The prediction results of the fire spreading trend, the evaluation results of the fire extinguishing effect, the smoke concentration, and the personnel distribution are integrated to generate a comprehensive analysis report of the fire scene situation. The generated comprehensive analysis report and data analysis results are analyzed, and combined with preset rules and emergency plans, the measures to be taken are determined, and a fire alarm instruction is generated, including adjusting the fire extinguishing strategy, dispatching additional rescue forces, and evacuating the surrounding personnel.

[0107] Suppose a fire breaks out in a large shopping mall, and the intelligent fire linkage management system starts the collaborative fire extinguishing program, collecting the startup signal of the collaborative fire extinguishing program and the real-time data transmitted from temperature sensors and smoke sensors at various positions in the shopping mall. The key information of the temperature and smoke concentration is extracted, and after forming a data set, it is uploaded to the cloud platform. The cloud platform obtains the geographical information and building layout of the shopping mall from the database, and combines the real-time temperature data to construct a 3D model of the shopping mall. Through data visualization techniques, a 3D heat map of the shopping mall is generated, from which the temperature distribution of different regions can be clearly seen. For example, the area where the fire occurs has a higher temperature and a darker color.

[0108] Through the analysis of the heat map, it is found that the fire is spreading towards the east side of the shopping mall and at a relatively fast speed. At the same time, evaluating the fire extinguishing effect, it is found that the current fire extinguishing measures are effective in some areas, but ineffective in the direction of the fire spread. Combining these information, a detailed comprehensive analysis report is generated, pointing out the development trend of the fire, the fire extinguishing effect, and the possible risks.

[0109] Based on the comprehensive analysis report and data analysis results, a fire alarm instruction is generated. The instruction includes adjusting the fire extinguishing strategy, increasing the deployment of dry powder fire extinguishers in the direction of fire spread; dispatching additional firefighters and fire trucks to the scene; at the same time, organizing the staff in the shopping mall to guide the evacuation of people in the surrounding areas. By uploading the start signal of the collaborative fire extinguishing procedure and real-time fire data to the cloud for analysis, more comprehensive and accurate fire information can be obtained. The three-dimensional heat map and comprehensive analysis report provide intuitive and detailed fire situation for decision-makers, enabling them to better understand the development trend of the fire and the fire extinguishing effect, and thus make more scientific and accurate decisions. Real-time data analysis and rapid generation of fire alarm instructions enable the fire department and relevant rescue personnel to timely understand the fire situation and quickly take corresponding measures. Instructions for adjusting the fire extinguishing strategy, dispatching additional rescue forces, and evacuating surrounding people can be executed within the shortest time, effectively improving the emergency response ability and reducing the losses caused by the fire. The use of the cloud platform enables decision-makers to remotely monitor the fire situation in real time, issue effective fire alarm instructions without being restricted by geographical location, realize remote command and dispatch, and improve the efficiency and coordination of fire extinguishing and rescue.

[0110] In a preferred embodiment of the present invention, based on the heat map, the real-time fire data is analyzed, including predicting the fire spread trend and evaluating the fire extinguishing effect, and a comprehensive analysis report on the current fire situation can be obtained, which may include:

[0111] According to the change of temperature in different regions of the heat map over time, identify and mark the regions where the temperature continues to rise;

[0112] Analyze the temperature gradient distribution in the regions where the temperature continues to rise to determine the direction of heat propagation, that is, the fire spread trend;

[0113] According to the fire spread trend, analyze the temperature change in the high-temperature regions before and after fire extinguishing, evaluate the degree of temperature drop in the high-temperature regions after the implementation of fire extinguishing measures, and obtain a preliminary evaluation result;

[0114] Integrate the fire spread trend, preliminary evaluation result, and abnormal values of smoke concentration and electrical parameters in the real-time data to obtain a comprehensive analysis report on the current fire situation.

[0115] In the embodiments of the present invention, the heat map of the fire scene is divided into grids according to scientific and reasonable rules, and each grid represents a specific area. Through the research and analysis of a large amount of historical fire data and in combination with the actual fire scene situation, appropriate temperature rise thresholds and heating rate thresholds are set. The temperature rise threshold is used to measure whether the temperature in the area has increased, and the heating rate threshold further considers the speed of temperature rise. These two thresholds are the key indicators for judging the development trend of the fire. For each grid area, the temperature changes at different time points are carefully calculated and tracked. When the temperature in a certain area shows an upward trend and the heating rate exceeds the pre-set threshold, the area is marked as a potential fire spread area. During the actual monitoring process, an automated data analysis program is used to scan the temperature data of each grid area in real time. Once a qualified area is found, the marking program is immediately started and marked in a prominent way on the heat map.

[0116] For the marked potential fire spread areas, calculate the temperature gradient between adjacent areas. The temperature gradient represents the direction and intensity of heat transfer between different areas. Through the comprehensive analysis of the temperature gradients of multiple adjacent areas and according to the distribution of the temperature gradients, determine the direction of heat propagation, that is, the fire spread trend. When calculating the temperature gradient, mathematical algorithms are used, considering the spatial position relationship, temperature difference and time factors between areas to ensure the accuracy and reliability of the calculation results. Use visualization technology to intuitively display the temperature gradient and the direction of heat propagation on the heat map.

[0117] According to the markings or relevant records in the heat map, determine the areas where fire extinguishing measures have been taken, and extract the temperature data of the areas before and after fire extinguishing at specific time points from the heat map data. Calculate the temperature difference of the area before and after fire extinguishing to evaluate the degree of temperature drop in the high-temperature area. Set multiple evaluation indicators, including the temperature drop range, the cooling rate, and the change in the area of the high-temperature area. According to the comprehensive performance of the indicators, judge the effectiveness of the fire extinguishing measures and obtain a preliminary evaluation result. During the evaluation process, considering the complexity of the fire scene environment, the temperature data is checked and corrected multiple times to ensure the credibility of the evaluation result.

[0118] Integrate the marked temperature rise areas, the determined heat propagation direction, the calculated temperature drop degree, the judgment of the fire extinguishing effectiveness, and the key information of smoke concentration and humidity in the real-time data. During the integration process, data fusion technology is used to standardize different types of data and generate a comprehensive analysis report on the current situation of the fire scene, including an overview of the basic situation of the fire scene, an analysis of the fire spread trend, an evaluation of the fire extinguishing effect, existing problems and risk warnings, and corresponding countermeasures.

[0119] Suppose a fire breaks out in the warehouse of a large factory, and the real-time data of the fire scene is analyzed through a heat map. Through data collection, the heat map of the warehouse is divided into multiple grid areas. When analyzing the temperature change, it is found that the temperature in the northwest corner area of the warehouse continues to rise, and the rising speed is relatively fast, exceeding the set threshold. So this area is marked out. By calculating the temperature gradient, it is found that the heat spreads from the northwest corner area to the southeast direction, and it is speculated that the fire may spread towards the southeast direction. The temperature data before and after the fire extinguishing in the area are extracted, and it is found that the average temperature before the fire extinguishing is 100°C, and the average temperature drops to 40°C after the fire extinguishing, with an average temperature drop of 60°C, indicating that the fire extinguishing effect in the area is good.

[0120] Integrate the results of the prediction of the fire spread trend and the evaluation of the fire extinguishing effect, and write a comprehensive analysis report. The report points out that the main situation of the current fire scene is that the fire in the northwest corner area has a spreading trend and may develop towards the southeast direction; the fire extinguishing measures in the northeast corner area have achieved good results. At the same time, it is recommended to strengthen the monitoring and fire extinguishing force deployment in the northwest corner area to prevent the fire from spreading further.

[0121] Through the detailed analysis of the temperature data in the heat map, the fire spread trend can be accurately predicted and the fire extinguishing effect can be evaluated. This enables the fire commanders to formulate more reasonable fire extinguishing strategies and rescue plans based on the scientific analysis results, improving the accuracy and effectiveness of fire response. Accurate prediction of the fire spread trend and evaluation of the fire extinguishing effect help to reasonably allocate fire resources. The limited fire forces can be concentrated in the key areas where the fire spreads, avoiding waste of resources and improving the fire extinguishing efficiency. Timely understanding of the fire spread trend and the fire extinguishing effect can provide guarantee for the safety of firefighters and surrounding people. Firefighters can take corresponding protective measures according to the accurate information to avoid getting into dangerous areas; they can also organize the evacuation of surrounding people in time to reduce the risk of casualties.

[0122] In a preferred embodiment of the present invention, there is also an emergency linkage module for verifying the fire alarm instructions, including the fire occurrence location and the danger level; determining the range of non-fire power supplies to be cut off according to the fire occurrence location, and sequentially cutting off the power of the corresponding circuits through controlling circuit breakers and contactor devices; obtaining the current position and operating state of the fire shutter according to the fire alarm instructions, formulating a closing strategy, and sending an instruction to the controller to reduce the closing speed according to the closing strategy; planning multiple safe evacuation paths according to the fire occurrence location, and adjusting the brightness and flashing frequency of emergency lighting fixtures; formulating evacuation broadcast content according to the fire alarm instructions and multiple safe evacuation paths, including evacuation instructions and safety tips.

[0123] In an embodiment of the present invention, it is checked whether the format of the extracted information conforms to a preset specification, and the logical relationship between the fire occurrence location and the danger level is verified for reasonableness. According to the fire occurrence location, the pre-stored electrical system layout diagram and power distribution information are queried to determine the non-fire-fighting power supply circuits corresponding to the location and the areas with potential safety hazards affected by the fire around it, and the circuit breaker and contactor devices are controlled in sequence. A cut-off signal is sent to the relevant circuit breaker through the control system. After receiving the signal, the circuit breaker disconnects the corresponding circuit connection and cuts off the power supply.

[0124] Through communication with the fire shutter controller, the current position and operating status information of the fire shutter are obtained.

[0125] And a closing strategy is formulated based on the danger level in the fire alarm instruction and the current state of the fire shutter. If the danger level is high and the fire shutter is currently in the open state, a rapid deceleration closing strategy is formulated, but it is also necessary to ensure the stable operation of the shutter during the deceleration process to avoid damage to the shutter due to excessive speed; if the danger level is medium and the shutter has already descended to a certain extent, a relatively gentle deceleration closing strategy is formulated.

[0126] According to the fire occurrence location, combined with the layout diagram of the building and the safety exit location information, multiple safe evacuation paths are planned using a path planning algorithm. Considering factors such as the personnel density, passage width, and fire spread direction in different areas, it is ensured that the planned paths can guarantee the safe evacuation of personnel. According to the planned evacuation paths, the brightness and blinking frequency of the emergency lighting fixtures are adjusted. The evacuation broadcast content is compiled based on the collected information, including reminders such as asking people to cover their mouths and noses with wet towels, bend down to move forward, and not crowd. The compiled evacuation broadcast content is played through the building's broadcast system to ensure that people in each area can clearly hear the broadcast information.

[0127] Suppose a fire breaks out in a large shopping mall, and the fire alarm instruction shows that the fire occurs in the clothing area on the second floor of the mall, and the danger level is medium. After receiving the instruction, the fire occurrence location and danger level information are extracted, the format is checked, and it is found that it conforms to the preset specification, and the logical relationship between the location and the danger level is reasonable and passes the verification. According to the electrical system layout of the mall, the non-fire-fighting power supply circuits in the clothing area on the second floor and adjacent areas are determined, and cut-off signals are sent to the relevant circuit breakers and contactors in sequence. First, the non-fire-fighting power supplies of the lighting and advertising light boxes in the clothing area are cut off, and then some non-essential power supplies in the adjacent areas are cut off to ensure the electrical safety in the fire area and its surroundings.

[0128] The current position of the fire shutter near the second - floor clothing area is obtained as fully open, and the operating state is stopped. According to the medium - risk level and the current state, a strategy is formulated to first quickly descend a certain distance and then gradually slow down to close. An instruction is sent to the fire - shutter controller, and the controller adjusts the descending speed of the shutter according to the strategy to achieve a gradient - speed - reduction closing. Combining the mall layout and the location of the safety exits, three evacuation paths are planned, leading to different safety exits respectively. The brightness of the emergency lighting fixtures on these three paths is increased and set to flash 4 times per second. Through the monitoring system, it is known that the personnel distribution in the mall is relatively uniform, and the fire has not shown an obvious spread at present. The evacuation broadcast content is written as "Dear customers and employees, there is a fire in the second - floor clothing area of the mall. Please do not panic. Please evacuate from the nearest safety exit according to the emergency lighting instructions, cover your nose and mouth with a wet towel, bend down and move forward in a low - posture manner, and do not crowd." And it is played in a loop in the mall through the broadcast system.

[0129] By implementing the non - fire - related power cut - off, the electrical safety risks at the fire scene are reduced, and the risk of electric shock to personnel is avoided; the gradient - speed - reduction closing of the fire shutter effectively blocks the spread of fire and smoke, buying time for personnel evacuation; reasonably planning the emergency lighting paths and adjusting the fixtures guides personnel to quickly find the safety exits; the evacuation broadcast provides clear evacuation instructions and safety tips to help personnel evacuate in an orderly manner. These operations comprehensively guarantee the life safety of personnel in case of a fire. Executing a series of operations according to the fire alarm instructions reduces the manual intervention links and improves the response speed. Quickly cutting off the non - fire - related power supply, orderly closing the fire shutter, timely planning the evacuation paths and playing the broadcast can quickly control the situation at the fire scene, create favorable conditions for fire - fighting and rescue work, and improve the overall fire - response efficiency. The non - fire - related power cut - off and the closing of the fire shutter can limit the scope of the fire's impact and reduce the damage to the equipment and commodity property in the building; reasonable evacuation operations can avoid crowded stampede accidents caused by personnel panic, indirectly reducing the property losses caused by the accidents and protecting the property safety of the mall or the venue. The automated operation of the entire process reflects the intelligent characteristics of the intelligent fire - fighting linkage management system, which can quickly and accurately respond according to the fire alarm instructions, complete complex operation tasks, improve the intelligent level of fire - fighting management, and meet the high requirements of modern buildings for fire safety.

[0130] An embodiment of the present invention also provides a computing device, including: a processor and a memory storing a computer program. When the computer program is run by the processor, it executes the system as described above. All the implementation manners in the above - mentioned system embodiment are applicable to this embodiment and can also achieve the same technical effects.

[0131] An embodiment of the present invention further provides a computer-readable storage medium storing instructions that, when run on a computer, cause the computer to execute the system described above. All implementation manners in the above system embodiments are applicable to this embodiment and can also achieve the same technical effects.

[0132] The above is the preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. An intelligent fire linkage management system for data centers in multi-station fusion substations, characterized in that: include: The fire detection module is used to collect smoke concentration, ambient temperature and humidity, and electrical parameters in real time, and set monitoring parameters for transformers, switch cabinets, and uninterruptible power supply battery packs to obtain original fire data; The control module is used to set population parameters and iteratively update individual fitness values ​​based on the spotted hyena optimization algorithm to screen out the final combination of smoke concentration, temperature, humidity and electrical parameters to determine the fire level, and automatically trigger the protection protocol when the evaluation result threatens the core equipment; Automatic fire extinguishing module, used to trigger the coordinated fire extinguishing procedure of the heptafluoropropane gas fire extinguishing device and the fine water mist fire extinguishing device according to the fire level determination result; The cloud platform module is used to upload the start signal of the coordinated fire-fighting program and the real-time data of the fire scene to the cloud, generate a three-dimensional thermal map, analyze the three-dimensional thermal map, and generate a fire alarm instruction.

2. According to claim 1, the intelligent fire linkage management system for data centers in multi-station fusion substations is characterized in that: Based on the spotted hyena optimization algorithm, by setting population parameters and iteratively updating individual fitness values, the final combination of smoke concentration, temperature, humidity and electrical parameters is screened to determine the fire level. When the evaluation result threatens the core equipment, the protection protocol is automatically triggered, including: Set the initial parameters of the spotted hyena optimization algorithm, including population size, number of iterations, and search range; According to each individual in the population, i.e., the combination of smoke concentration, oil temperature and resistance value parameters, the fitness value corresponding to each individual is calculated by calculating the actual and benchmark ratios of smoke concentration, oil temperature, resistance value and historical fire frequency, and fusing the multi-source parameter contribution values. According to the fitness value, the individuals in the population are updated, including selection, crossover and mutation operations; Repeatedly calculate the fitness value of the individual, and select and update operations based on the fitness value, continuously optimize the population until the preset number of iterations is reached, and determine the final solution from the population based on the fitness value of the individual, including the combination of smoke concentration, ambient temperature and humidity, and electrical parameters; generate the fire level determination result based on the parameter combination of the final solution; The fire level assessment results show that if the transformer, switchgear and uninterruptible power supply battery pack are threatened, the core equipment priority protection protocol will be automatically triggered and relevant protection measures will be initiated.

3. The intelligent fire linkage management system for data centers in multi-station fusion substations according to claim 2 is characterized in that: According to each individual in the population, that is, the combination of smoke concentration, oil temperature and resistance value parameters, by calculating the actual and benchmark ratios of smoke concentration, oil temperature, resistance value and historical fire frequency, the fitness value corresponding to each individual is calculated by integrating the multi-source parameter contribution values, including: Obtaining an actual smoke concentration value, calculating a proportional relationship based on the actual smoke concentration value and a preset maximum smoke concentration value, and calculating a smoke concentration contribution value based on the proportional relationship; Obtain the current oil temperature value, calculate the oil temperature ratio between the current oil temperature value and the maximum allowable oil temperature variation range of the transformer under normal operating conditions, and determine the oil temperature contribution value based on the oil temperature ratio; Obtain the resistance value of the current UPS battery pack, and calculate the resistance ratio between the resistance value and the reference resistance value of the battery in a normal state; Extract relevant data from the historical fire record database of the data center and analyze it to calculate the frequency of fires that occurred in the area in the past; The smoke concentration contribution value, oil temperature contribution value, resistance ratio value and frequency value are integrated to obtain the fitness value.

4. The intelligent fire linkage management system for data centers in multi-station fusion substations according to claim 3 is characterized in that: The fire level determination results include no fire, slight fire, moderate fire and severe fire.

5. The intelligent fire linkage management system for data centers in multi-station integrated substations according to claim 4 is characterized in that: According to the fire level determination result, the coordinated fire extinguishing procedure of the heptafluoropropane gas fire extinguishing device and the water mist fire extinguishing device is triggered, including: Match the corresponding fire extinguishing strategy from the preset fire extinguishing strategy library, including activating the fine water mist fire extinguishing device for minor fires and activating the heptafluoropropane gas fire extinguishing device and the fine water mist fire extinguishing device at the same time for serious fires; According to the fire extinguishing strategy, a start command is sent to release HFC-227ea gas into the fire area. At the same time, the water mist fire extinguishing device starts the water pump to spray water into the fire area. During the fire-fighting process, the fire-fighting strategy and coordinated fire-fighting procedures are adjusted in real time according to the on-site temperature and smoke concentration.

6. The intelligent fire linkage management system for data centers in multi-station integrated substations according to claim 5 is characterized in that: The start signal of the coordinated fire-fighting program and the real-time data of the fire scene are uploaded to the cloud to generate a three-dimensional thermal map, which is then analyzed to generate fire alarm instructions, including: Organize the start signal and real-time fire scene data to form a data set; Upload the data set to the cloud platform to extract the temperature and smoke density information of the fire scene; According to the temperature of the fire scene, combined with the geographical information and building layout of the fire scene, a three-dimensional model of the fire scene is constructed to generate a thermal map; According to the heat map, the real-time data of the fire scene is analyzed, including the prediction of the fire spread trend and the evaluation of the fire extinguishing effect, to obtain a comprehensive analysis report on the current situation of the fire scene; Based on the comprehensive analysis report and data analysis results, fire alarm instructions are generated, including adjusting fire extinguishing strategies, increasing rescue forces and evacuating surrounding personnel.

7. The intelligent fire linkage management system for data centers in multi-station integrated substations according to claim 6 is characterized in that: According to the heat map, the real-time data of the fire scene is analyzed, including the prediction of the fire spread trend and the evaluation of the fire extinguishing effect, and a comprehensive analysis report of the current situation of the fire scene is obtained, including: Based on the temperature changes over time in different areas of the thermal map, identify and mark areas where the temperature continues to rise; Conduct temperature gradient distribution analysis in areas where the temperature continues to rise to determine the direction of heat propagation, i.e. the trend of fire spread; According to the fire spread trend, the temperature changes in the high-temperature area before and after the fire was extinguished were analyzed, and the degree of temperature drop in the high-temperature area after the fire extinguishing measures were implemented was evaluated to obtain preliminary evaluation results; The fire spread trend, preliminary assessment results, and abnormal values ​​of smoke concentration and electrical parameters in real-time data are integrated to obtain a comprehensive analysis report on the current situation of the fire scene.

8. The intelligent fire linkage management system for data centers in multi-station integrated substations according to claim 1 is characterized in that: It also includes an emergency linkage module, which is used to verify the fire alarm instructions, including the location of the fire and the danger level; determine the scope of non-fire power supply to be cut off according to the location of the fire, and cut off the power supply of the corresponding circuits in sequence by controlling the circuit breaker and contactor equipment; obtain the current position and operating status of the fireproof roller shutter according to the fire alarm instruction, formulate a closing strategy, and send instructions to the controller to reduce the closing speed according to the closing strategy; plan multiple safe evacuation routes according to the location of the fire, and adjust the brightness and flashing frequency of emergency lighting fixtures; formulate evacuation broadcast content according to the fire alarm instructions and multiple safe evacuation routes, including evacuation instructions and safety tips.

9. A computing device, characterized in that include: one or more processors; A storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the system according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program, and when the program is executed by a processor, the system according to any one of claims 1 to 8 is implemented.

Images