An intelligent control system for fire safety management

Abstract

The present invention discloses an intelligent control system for fire safety management, which relates to the field of safety management technology, including a multimodal sensor module and a digital twin modeling module, wherein the multimodal sensor module and the digital twin modeling module are connected and communicated through a data transmission network, and the digital twin modeling module includes an edge computing node in the region and a central data processing server, and constructs a three-dimensional digital twin model of a building and its environment by receiving and fusing data from the multimodal sensor module and utilizing real-time data processing and model updating algorithms. The present invention realizes the function of constructing and continuously updating a digital twin model of a building and its environment by designing a multimodal sensor module and a digital twin modeling module, and reflects the internal structure of the building, environmental parameters and personnel flow status in real time, solves the problems of data dispersion, lack of real-time performance and static model in traditional systems, and ensures the accuracy and timeliness of fire risk assessment.

Description

Technical Field

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

[0002] In fire safety management, traditional systems mainly rely on static sensors and fixed threshold judgments, which are unable to reflect the fire conditions inside the building in real time, resulting in the inability to understand the fire conditions in real time and delayed early warning responses, making it difficult to achieve accurate fire risk prediction and efficient emergency dispatch.

[0003] To this end, this design introduces digital twins, virtual simulation, intelligent collaborative decision-making and blockchain data sharing technologies to build a real-time dynamic virtual replica of the building, realize full-link simulation and collaborative response of fire scenes, and thus comprehensively improve the level of fire safety management.

[0004] Patent document CN118446411B discloses a fire safety management system and its processing method based on the application of digital technology. The above patent realizes the information processing and information storage process that effectively distinguishes between fire incidents and fire tasks, and improves the system reliability of the fire safety management link.

[0005] In summary, the above patents improve the performance of PaaS and SaaS services in the process of implementing differentiated management set configurations, build a new model of online and offline technical services and management of fire safety based on digital information systems, but there is still room for improvement in real-time reflection of fire conditions inside buildings; To this end, the present application proposes an intelligent control system for fire safety management that can build and continuously update a digital twin model of a building and its environment. Summary of the invention

[0006] The purpose of the present invention is to provide an intelligent control system for fire safety management to solve the technical problems of being unable to understand the fire status in real time and delayed early warning response proposed in the above background technology.

[0007] To achieve the above-mentioned object, the present invention provides the following technical solutions: an intelligent control system for fire safety management, comprising a multimodal sensor module and a digital twin modeling module, wherein the multimodal sensor module and the digital twin modeling module are connected and communicated via a data transmission network; The multimodal sensor module is equipped with a temperature sensor, a smoke sensor, a combustible gas sensor, a structural vibration sensor, a video image sensor, an infrared sensor and a personnel flow detection unit. Each sensor is connected to the data transmission network through a data acquisition interface. The digital twin modeling module includes an edge computing node and a central data processing server in the region. By receiving and fusing data from the multimodal sensor module, a three-dimensional digital twin model of the building and its environment is constructed using real-time data processing and model update algorithms.

[0008] Preferably, each sensor in the multimodal sensor module is equipped with a self-diagnosis unit for real-time detection of sensor status and data anomalies. Each sensor is equipped with an adaptive calibration unit for automatically adjusting the acquisition sensitivity according to ambient temperature and humidity parameters, and feeding back the calibration parameters to the central data processing server through a data interface.

[0009] Preferably, the edge computing node is responsible for preprocessing, noise filtering and preliminary fusion of the data collected in the area, and uses local algorithms to achieve data compression and delay optimization. After the central data processing server receives the data transmitted by the edge node, it adopts a real-time model update algorithm to use the preprocessed data to update the three-dimensional model of the building, and uses a distributed computing architecture to achieve parallel processing of multi-region data.

[0010] Preferably, the real-time model update algorithm adopts an adaptive parameter correction mechanism, specifically including: a model parameter initialization module and a dynamic feedback adjustment module. The model parameter initialization module sets the initial model parameters according to the building structure, material properties and initial environmental parameters. The dynamic feedback adjustment module collects sensor feedback data in real time, compares the model prediction with the actual measurement value, and uses the least squares method and adaptive filtering algorithm to automatically correct the temperature, smoke diffusion and structural vibration parameters to ensure that the digital twin model is consistent with the on-site environment.

[0011] Preferably, the control system further comprises a virtual simulation module, the virtual simulation module is configured with a fire physics simulation engine, and the fire physics simulation engine is implemented using a computational fluid dynamics (CFD) algorithm: The fire spread simulation unit meshes the spread of fire sources in the building and calculates the combustion reaction, heat release rate and pressure distribution; The smoke diffusion simulation unit calculates smoke concentration distribution and visibility changes through a physical diffusion model; The heat transfer simulation unit calculates temperature gradients in real time based on convection, radiation and conduction models; The personnel evacuation simulation unit calculates personnel flow and optimal evacuation path using discrete elements or based on social mechanics models, and outputs the simulation results to the decision-making module.

[0012] Preferably, the virtual simulation module is further configured with a simulation parameter configuration unit, which supports inputting or automatically obtaining the following parameters: Fire source location and initial fire parameters; Combustion characteristic parameters of food materials in buildings, including calorific value and burning speed; Initial ambient temperature, humidity and ventilation condition parameters; Information on population density, distribution and evacuation routes; The simulation time step and resolution settings ensure that the simulation data is transmitted to the intelligent collaborative decision-making module in the management and control system in real time.

[0013] Preferably, the intelligent collaborative decision-making module is embedded with a fire risk prediction model based on a deep neural network, and the fire risk prediction model based on the deep neural network realizes the calculation of the fire risk index through the following steps: The data input unit receives real-time data from the multimodal sensor module, the data twin module, and the virtual simulation module; The multi-layer neural network structure uses convolutional layers, recurrent layers, and fully connected layers to extract features and perform spatiotemporal correlation analysis on input data; The risk prediction output unit outputs the fire risk index and regional risk distribution map according to the training results of the fire risk prediction model based on the deep neural network.

[0014] Preferably, the intelligent collaborative decision-making module is further configured with a data fusion and risk assessment unit, specifically including: The data fusion unit uses weighted average, Kalman filtering or Bayesian estimation algorithm to fuse the data information from the multimodal sensor and the virtual simulation module; The risk assessment unit performs statistical analysis and normalization on the fused data to form a regional fire risk distribution, and determines the fire warning level through preset thresholds and dynamic adjustment algorithms; The emergency response suggestion generator automatically generates evacuation routes and fire resource dispatch plans based on risk levels, building layout, and personnel distribution data.

[0015] Preferably, the control system further includes a cross-department linkage platform, which is based on a data sharing unit, a consensus mechanism module and a real-time information synchronization interface of the blockchain to achieve data interaction and command linkage between the control system and the joint defense departments, specifically including: The data sharing unit uses distributed ledger technology to store real-time data of each module and implements data access permission control through smart contracts; The consensus mechanism module uses proof of work or proof of stake algorithm to ensure that data cannot be tampered with; The real-time information synchronization interface supports synchronizing encrypted and verified data to the joint defense department terminal through the RESTful API or MQTT protocol.

[0016] Preferably, the management and control system further includes an integrated user interface module, which includes a desktop terminal and a mobile terminal, specifically including: The graphical user interface GUI uses a high-resolution display to present fire alarms, building digital twin models, simulation dynamics and risk distribution maps in real time; The mobile terminal interface supports data transmission based on wireless communication and real-time reception of emergency instructions; The voice recognition input unit has built-in noise suppression and natural language processing algorithms, which supports users to query alarm information, adjust simulation parameters and start emergency plans through voice commands; The data interaction interface is compatible with HTTPS and WebSocket protocols, realizing real-time information interaction and command issuance with other system modules and external emergency platforms.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention realizes the function of building and continuously updating the digital twin model of the building and its environment by designing a multimodal sensor module and a digital twin modeling module, which reflects the internal structure of the building, environmental parameters and personnel flow status in real time, solves the problems of data dispersion, lack of real-time performance and static model of the traditional system, provides high-quality real-time basic data, and ensures the accuracy and timeliness of fire risk assessment; 2. The present invention realizes the function of dynamically simulating fire drill scenes by designing a virtual simulation module, solves the limitations of the traditional system in predicting single fire scenes and incomplete emergency plans, optimizes evacuation and rescue plans, and improves emergency response efficiency; 3. The present invention realizes the functions of automatically calculating the fire risk index, warning level and generating the optimal emergency response plan by designing an intelligent collaborative decision-making module, which solves the problems of false alarms and missed alarms caused by the single and fixed threshold of the traditional early warning mechanism, improves the accuracy of fire risk prediction and enhances the prevention and control capabilities; 4. The present invention realizes the function of multi-department data sharing and collaborative generation of emergency command responses by designing a cross-department linkage platform, solves the problems of information islands and data delays in traditional management and control systems, enhances collaborative combat capabilities, and improves overall emergency handling efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 Schematic diagram of the workflow of the management and control system of the present invention. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0020] See also Figure 1 , an embodiment provided by the present invention: an intelligent control system for fire safety management, comprising a multimodal sensor module and a digital twin modeling module, wherein the multimodal sensor module and the digital twin modeling module are connected and communicated via a data transmission network; The multimodal sensor module is equipped with a temperature sensor, a smoke sensor, a combustible gas sensor, a structural vibration sensor, a video image sensor, an infrared sensor, and a personnel flow detection unit. Each sensor is connected to a data transmission network through a data acquisition interface. The digital twin modeling module includes an edge computing node in the region and a central data processing server. By receiving and fusing data from the multimodal sensor module, a three-dimensional digital twin model of the building and its environment is constructed using real-time data processing and model update algorithms; Each sensor in the multimodal sensor module is equipped with a self-diagnosis unit for real-time detection of sensor status and data anomalies. Each sensor is equipped with an adaptive calibration unit to automatically adjust the acquisition sensitivity according to ambient temperature and humidity parameters, and feed back the calibration parameters to the central data processing server through a data interface. The edge computing node is responsible for preprocessing, noise filtering and preliminary fusion of the data collected in the area, and uses local algorithms to achieve data compression and delay optimization. After receiving the data transmitted by the edge node, the central data processing server adopts a real-time model update algorithm to use the preprocessed data to update the three-dimensional model of the building, and adopts a distributed computing architecture to achieve multi-region data parallel processing; The real-time model update algorithm adopts an adaptive parameter correction mechanism, which specifically includes: a model parameter initialization module and a dynamic feedback adjustment module. The model parameter initialization module sets the initial model parameters according to the building structure, material properties and initial environmental parameters. The dynamic feedback adjustment module collects sensor feedback data in real time, compares the model prediction with the actual measurement value, and uses the least squares method and adaptive filtering algorithm to automatically correct the temperature, smoke diffusion and structural vibration parameters to ensure that the digital twin model is consistent with the on-site environment. Furthermore, the temperature sensor is used to monitor the temperature changes in various areas of the building in real time, the smoke sensor monitors the smoke concentration in the air, the combustible gas sensor detects the leakage of flammable gas, the structural vibration sensor monitors the structural health of the building and promptly detects changes in the building structure caused by fire or other external factors, the video image sensor monitors the situation in the building in real time through the camera and provides the scene picture when the fire occurs, the infrared sensor is based on thermal imaging technology and assists the smoke sensor in capturing the fire source, and the personnel flow detection unit monitors the personnel flow in real time through the access control system, infrared sensing or video analysis; The edge computing node is responsible for the preliminary preprocessing of the collected sensor data locally, including filtering, denoising and data fusion. In the edge computing stage, local algorithms are used to compress the data, optimize transmission efficiency, reduce latency, and ensure that the data can be quickly sent to the central data processing server. The central data processing server receives the data transmitted by the edge computing node and uses a real-time model update algorithm to use the preprocessed data to update the three-dimensional mathematical twin model of the building in real time. The model dynamically reflects information such as fire risks, personnel distribution and environmental changes inside the building; When the system is initialized, the model parameter initialization module sets the initial model according to the building's structure, material properties and initial environmental parameters. These parameters include the building's size, the heat conduction system of the structural material, the initial temperature, the air flow conditions, etc. These initial parameters lay the foundation for subsequent real-time updates. The core of the dynamic feedback module is to compare the data fed back by the multimodal sensor with the model prediction results. Whenever the system receives sensor data, the dynamic feedback adjustment module will use the least squares method and adaptive filtering algorithm to correct the model parameters. For example, when the temperature value detected by the temperature sensor does not match the model prediction value, the system will automatically adjust the model's temperature parameters to ensure that the temperature change is suppressed in accordance with the actual situation; similarly, data such as smoke concentration and structural vibration will also be corrected through the algorithm.

[0021] See also Figure 1 The present invention provides an embodiment: an intelligent control system for fire safety management, the control system further comprising a virtual simulation module, the virtual simulation module is configured with a fire physics simulation engine, and the fire physics simulation engine is implemented using a computational fluid dynamics (CFD) algorithm: The fire spread simulation unit meshes the spread of fire sources in the building and calculates the combustion reaction, heat release rate and pressure distribution; The smoke diffusion simulation unit calculates smoke concentration distribution and visibility changes through a physical diffusion model; The heat transfer simulation unit calculates temperature gradients in real time based on convection, radiation and conduction models; The personnel evacuation simulation unit calculates personnel flow and optimal evacuation path using discrete elements or based on social mechanics models, and outputs the simulation results to the decision-making module; The virtual simulation module is also configured with a simulation parameter configuration unit, which supports inputting or automatically obtaining the following parameters: Fire source location and initial fire parameters; Combustion characteristic parameters of food materials in buildings, including calorific value and burning speed; Initial ambient temperature, humidity and ventilation condition parameters; Information on population density, distribution and evacuation routes; The simulation time step and resolution settings ensure that the simulation data is transmitted to the intelligent collaborative decision-making module in the control system in real time; Furthermore, before the simulation starts, the system will obtain necessary parameter data from the multimodal sensor module, the digital twin model, and the external environment, and set them through the simulation parameter configuration unit: including the fire source location and initial fire parameters, fuel combustion characteristic parameters in the building, initial environmental parameters, occupant density and distribution, and simulation time step and resolution settings; The physical simulation engine divides the internal space of the building into grids, each grid representing a calculation unit. The grid size is adjusted according to the building structure, fire risk area and simulation accuracy requirements. The fire spread simulation unit uses CFD algorithm to simulate the combustion reaction, including fuel consumption, oxygen consumption rate, heat release rate, etc. This information helps to predict the energy release of the fire source and the impact on the surrounding environment. By calculating the heat and pressure changes generated by the combustion reaction, the simulation unit calculates the temperature distribution and airflow changes during the fire spread process in the building, providing basic data for subsequent smoke diffusion and temperature gradient calculations; The smoke diffusion simulation unit uses a physical diffusion model to simulate how smoke and harmful gases diffuse after a fire occurs. Based on the CFD algorithm, the physical diffusion model takes into account the airflow, temperature differences and resistance effects within the building and calculates the concentration distribution of smoke. The smoke diffusion simulation unit calculates the visibility changes within the building in real time and gives the line of sight of the fire area, helping decision makers assess the fire risk area and providing support for personnel evacuation. The heat transfer simulation unit calculates the temperature gradients in the building in real time based on three heat transfer mechanisms: Convection: Calculates heat transfer caused by air movement within a building; Radiation: simulates the radiation heat transfer between the fire source and surrounding objects; Conduction: Simulates the temperature changes of building materials caused by fire and the spread of heat to other areas; The personnel evacuation simulation unit calculates the flow of personnel and the best evacuation path according to the personnel distribution in the building, the evacuation channel information, the environmental conditions when the fire occurs, and the personnel behavior model such as the social mechanics model or the flow model based on discrete elements. It simulates how the personnel in the building choose the evacuation path according to the fire source location, smoke diffusion and temperature distribution, and determines the flow speed of the personnel. According to the evacuation channel information of the building, it calculates the evacuation time and evaluates the evacuation efficiency when the fire occurs, ensuring that the personnel can escape from the fire area in the shortest time. The physical simulation engine generates information such as fire spread, smoke diffusion, temperature gradient, and personnel evacuation through simulation calculations to form dynamic fire scene data. This data will be output to the intelligent collaborative decision-making module of the management and control system for subsequent decision support. Through the simulation data interface, the system will transmit this information to the intelligent collaborative decision-making module in real time, evaluate the fire risk, personnel evacuation efficiency, and fire resource allocation needs based on the simulation results, thereby generating the optimal emergency response plan and adjusting emergency decisions based on real-time changes.

[0022] See also Figure 1 , an embodiment provided by the present invention: an intelligent control system for fire safety management, wherein the intelligent collaborative decision-making module is embedded with a fire risk prediction model based on a deep neural network, and the fire risk prediction model based on the deep neural network realizes the calculation of the fire risk index through the following steps: The data input unit receives real-time data from the multimodal sensor module, the data twin module, and the virtual simulation module; The multi-layer neural network structure uses convolutional layers, recurrent layers, and fully connected layers to extract features and perform spatiotemporal correlation analysis on input data; The risk prediction output unit outputs the fire risk index and regional risk distribution map according to the training results of the fire risk prediction model based on the deep neural network; The intelligent collaborative decision-making module is also equipped with a data fusion and risk assessment unit, which specifically includes: The data fusion unit uses weighted average, Kalman filtering or Bayesian estimation algorithm to fuse the data information from the multimodal sensor and the virtual simulation module; The risk assessment unit performs statistical analysis and normalization on the fused data to form a regional fire risk distribution, and determines the fire warning level through preset thresholds and dynamic adjustment algorithms; The emergency response suggestion generator automatically generates evacuation routes and fire resource dispatch plans based on risk levels, building layout, and personnel distribution data; Furthermore, the first step of the intelligent collaborative decision-making module is to receive real-time data from multiple data sources, including sensor data provided by the multimodal sensor module, the three-dimensional model of the building and its environmental data provided by the digital twin module, and the fire simulation data provided by the virtual simulation module. The convolution layer of the deep neural network DNN fire risk prediction model performs a convolution operation on the output sensor data through a sliding window to extract spatial features. The convolution layer helps the model identify the fire risk characteristics of each area, such as abnormally high temperature areas or areas with high smoke concentrations; the circulation layer processes time series data, analyzes the temporal dynamics of the fire development process, and uses the fire data at previous moments to predict the fire risk at the current moment, helping the model capture the time series correlation. After the convolution layer and the circulation layer, the output of all layers will pass through the fully connected layer for comprehensive integration of features. This layer fuses features of different dimensions through the multi-layer perceptron MPL structure to output accurate fire risk predictions; After the neural network calculation, the model will output a comprehensive fire risk index, which represents the probability and severity of fire in the building area. This index is based on the model's learning results of real-time data and historical fire data, as well as the actual conditions inside the current building. According to the predicted fire risk index, the model generates a regional risk distribution map to show the fire risk level of each area. The color depth of the risk map indicates the degree of fire danger in different areas, helping decision makers quickly identify high-risk areas. The data fusion unit uses the weighted average method to fuse the data from the multimodal sensors, virtual simulation modules and digital twin modules. The weighted average method assigns different weights to each data source according to the importance of different data sources, and integrates the output of each data source according to the weighted rules. For sensor data with high noise or uncertainty, the fusion unit uses Kalman filtering or Bayesian estimation methods to smooth and correct the data, eliminate the impact of inaccurate or abnormal data, and ensure the reliability of the final data; the results of data fusion will be transmitted to the risk assessment unit for statistical analysis first. The unit will normalize the risk index of multiple areas so that the risk data of each area are comparable and standardized to a unified range. Based on the statistical analysis and normalized data, the system uses preset thresholds and dynamic adjustment algorithms to calculate the fire warning level. According to different fire warning levels, the system will output different levels of response suggestions: Low risk: If the fire risk index is below a certain threshold, the system indicates that the fire risk is low and routine monitoring can be maintained; Medium risk: If the fire risk index is close to the medium level, the system prompts that further monitoring needs to be strengthened and corresponding emergency resources need to be prepared; High risk: If the fire risk index exceeds the high risk threshold, the system will immediately trigger a high risk warning, activate the emergency response plan, and pass it through relevant departments.

[0023] See also Figure 1 , an embodiment provided by the present invention: an intelligent control system for fire safety management, the control system further includes a cross-department linkage platform, the cross-department linkage platform is based on a blockchain data sharing unit, a consensus mechanism module and a real-time information synchronization interface, to achieve data interaction and command linkage between the control system and the joint defense department, specifically including: The data sharing unit uses distributed ledger technology to store real-time data of each module and implements data access permission control through smart contracts; The consensus mechanism module uses proof of work or proof of stake algorithm to ensure that data cannot be tampered with; The real-time information synchronization interface supports synchronizing encrypted and verified data to the joint defense department terminal through the RESTful API or MQTT protocol; Furthermore, when the system identifies that the fire risk level reaches a preset threshold, such as a high risk level, the intelligent collaborative decision-making module will automatically trigger the cross-departmental linkage platform and send fire information and emergency instructions to relevant departments. Through the data sharing unit, all relevant data will be immediately synchronized to the fire, medical and emergency management departments. After receiving the latest fire information, the fire department obtains the temperature distribution, smoke concentration and other data in the fire area through the real-time synchronization interface, and formulates the optimal fire extinguishing strategy based on these data. The medical department receives information such as the number and severity of the wounded through the interface, and dispatches emergency resources according to demand. The fire information obtained by the emergency management department helps it coordinate resources, evacuate safety personnel, control traffic, etc. All departments achieve collaborative operations through the data sharing platform and real-time synchronization interface.

[0024] See also Figure 1 The present invention provides an embodiment: an intelligent control system for fire safety management, the control system also includes an integrated user interface module, the integrated user interface module includes a desktop terminal and a mobile terminal, specifically including: The graphical user interface GUI uses a high-resolution display to present fire alarms, building digital twin models, simulation dynamics and risk distribution maps in real time; The mobile terminal interface supports data transmission based on wireless communication and real-time reception of emergency instructions; The voice recognition input unit has built-in noise suppression and natural language processing algorithms, which supports users to query alarm information, adjust simulation parameters and start emergency plans through voice commands; The data interaction interface is compatible with HTTPS and WebSocket protocols, realizing real-time information interaction and command issuance with other system modules and external emergency platforms; Furthermore, when the system detects a fire risk, the alarm information will be displayed immediately on the GUI interface, including the fire type, location and risk level, etc., and the three-dimensional digital twin model of the building will be presented in real time, showing information such as the temperature and smoke concentration distribution in the building, providing users with an intuitive fire situation and updating it dynamically in real time; based on the data provided by the virtual simulation module, the GUI interface will dynamically present the fire risk distribution of each area of ​​the fire simulation process machine, and the risk distribution map will distinguish the fire risk level through color changes, helping users quickly understand the high-risk areas; The mobile terminal interface supports data transmission based on wireless communication technologies such as 4G, 5G and Wi-Fi, ensuring that users can receive real-time information related to fire anytime and anywhere. All data sent through the mobile terminal, such as feedback on fire conditions and on-site environmental data, will be synchronized to the cloud platform and system modules to ensure information synchronization between the commander and on-site personnel. After receiving the command, the mobile terminal can perform related operations such as starting the fire extinguisher and selecting the evacuation route through the touch screen, buttons, etc., and feedback the execution results to the system. The voice recognition unit has a built-in noise suppression algorithm, which can clearly recognize voice commands in noisy environments. This is particularly important for applications in high-noise environments such as fire scenes. After the voice input unit converts the voice command into text, it uses the natural language processing algorithm to parse the user's intention. The NLP algorithm identifies the keywords in the command and performs semantic analysis to make a correct response; users can query various information in the system in real time through voice commands, such as fire alarms, building models, smoke concentrations, temperature and other data; users can also control the system through voice commands, such as adjusting simulation parameters, starting emergency plans, adjusting evacuation routes, etc. The combination of voice recognition and natural language processing technology allows users to interact with the system more flexibly and conveniently.

[0025] Working principle: The control system collects real-time environmental data inside the building by deploying multimodal sensors, and pre-processes and fuses the data through edge computing nodes and central data processing servers to create and update the three-dimensional digital twin model of the building. The model reflects the actual status of the building, fire development and risk areas in real time; After receiving the sensor data, the intelligent collaborative decision-making module uses a deep neural network to perform spatiotemporal correlation analysis on the data, calculate the fire risk index and generate a regional risk distribution map. Based on the fire risk prediction results, the system automatically adjusts the fire warning level and generates an emergency response plan; The virtual simulation module embedded in the system simulates the spread of fire, smoke diffusion and personnel evacuation routes through computational fluid dynamics. The simulation results can be used to optimize emergency plans and provide dynamic fire scenarios to the intelligent collaborative decision-making module in real time. The emergency response plan generator automatically generates evacuation routes and fire resource scheduling plans based on the fire risk level, building layout and personnel distribution.

[0026] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. An intelligent control system for fire safety management, comprising a multimodal sensor module and a digital twin modeling module, characterized in that: The multimodal sensor module and the digital twin modeling module are connected and communicated via a data transmission network; The multimodal sensor module is equipped with a temperature sensor, a smoke sensor, a combustible gas sensor, a structural vibration sensor, a video image sensor, an infrared sensor and a personnel flow detection unit. Each sensor is connected to the data transmission network through a data acquisition interface. The digital twin modeling module includes an edge computing node and a central data processing server in the region. By receiving and fusing data from the multimodal sensor module, a three-dimensional digital twin model of the building and its environment is constructed using real-time data processing and model update algorithms.

2. The intelligent control system for fire safety management according to claim 1, characterized in that: Each sensor in the multimodal sensor module is equipped with a self-diagnosis unit for real-time detection of sensor status and data anomalies. Each sensor is equipped with an adaptive calibration unit for automatically adjusting the acquisition sensitivity according to ambient temperature and humidity parameters, and feeding back the calibration parameters to the central data processing server through a data interface.

3. The intelligent control system for fire safety management according to claim 1 is characterized in that: The edge computing node is responsible for preprocessing, noise filtering and preliminary fusion of the data collected in the area, and uses local algorithms to achieve data compression and delay optimization. After the central data processing server receives the data transmitted by the edge node, it adopts a real-time model update algorithm to use the preprocessed data to update the three-dimensional model of the building, and uses a distributed computing architecture to achieve parallel processing of multi-region data.

4. The intelligent control system for fire safety management according to claim 3 is characterized by: The real-time model update algorithm adopts an adaptive parameter correction mechanism, which specifically includes: a model parameter initialization module and a dynamic feedback adjustment module. The model parameter initialization module sets the initial model parameters according to the building structure, material properties and initial environmental parameters. The dynamic feedback adjustment module collects sensor feedback data in real time, compares the model prediction with the actual measurement value, and uses the least squares method and adaptive filtering algorithm to automatically correct the temperature, smoke diffusion and structural vibration parameters to ensure that the digital twin model is consistent with the on-site environment.

5. The intelligent control system for fire safety management according to claim 1 is characterized in that: The control system also includes a virtual simulation module, which is configured with a fire physics simulation engine, and the fire physics simulation engine is implemented using a computational fluid dynamics (CFD) algorithm: The fire spread simulation unit meshes the spread of fire sources in the building and calculates the combustion reaction, heat release rate and pressure distribution; The smoke diffusion simulation unit calculates smoke concentration distribution and visibility changes through a physical diffusion model; The heat transfer simulation unit calculates temperature gradients in real time based on convection, radiation and conduction models; The personnel evacuation simulation unit calculates personnel flow and optimal evacuation path using discrete elements or based on social mechanics models, and outputs the simulation results to the decision-making module.

6. The intelligent control system for fire safety management according to claim 5, characterized in that: The virtual simulation module is also configured with a simulation parameter configuration unit, which supports inputting or automatically obtaining the following parameters: Fire source location and initial fire parameters; Combustion characteristic parameters of food materials in buildings, including calorific value and burning speed; Initial ambient temperature, humidity and ventilation condition parameters; Information on population density, distribution and evacuation routes; The simulation time step and resolution settings ensure that the simulation data is transmitted to the intelligent collaborative decision-making module in the management and control system in real time.

7. The intelligent control system for fire safety management according to claim 6, characterized in that: The intelligent collaborative decision-making module is embedded with a fire risk prediction model based on a deep neural network. The fire risk prediction model based on the deep neural network calculates the fire risk index through the following steps: The data input unit receives real-time data from the multimodal sensor module, the data twin module, and the virtual simulation module; The multi-layer neural network structure uses convolutional layers, recurrent layers, and fully connected layers to extract features and perform spatiotemporal correlation analysis on input data; The risk prediction output unit outputs the fire risk index and regional risk distribution map according to the training results of the fire risk prediction model based on the deep neural network.

8. The intelligent control system for fire safety management according to claim 7, characterized in that: The intelligent collaborative decision-making module is also equipped with a data fusion and risk assessment unit, which specifically includes: The data fusion unit uses weighted average, Kalman filtering or Bayesian estimation algorithm to fuse the data information from the multimodal sensor and the virtual simulation module; The risk assessment unit performs statistical analysis and normalization on the fused data to form a regional fire risk distribution, and determines the fire warning level through preset thresholds and dynamic adjustment algorithms; The emergency response suggestion generator automatically generates evacuation routes and fire resource dispatch plans based on risk levels, building layout, and personnel distribution data.

9. The intelligent control system for fire safety management according to claim 1, characterized in that: The control system further includes a cross-department linkage platform, which is based on a blockchain data sharing unit, a consensus mechanism module, and a real-time information synchronization interface to achieve data interaction and command linkage between the control system and the joint defense departments, specifically including: The data sharing unit uses distributed ledger technology to store real-time data of each module and implements data access permission control through smart contracts; The consensus mechanism module uses proof of work or proof of stake algorithm to ensure that data cannot be tampered with; The real-time information synchronization interface supports synchronizing encrypted and verified data to the joint defense department terminal through the RESTful API or MQTT protocol.

10. The intelligent control system for fire safety management according to claim 1, characterized in that: The management and control system also includes a comprehensive user interface module, which includes a desktop terminal and a mobile terminal, specifically including: The graphical user interface GUI uses a high-resolution display to present fire alarms, building digital twin models, simulation dynamics and risk distribution maps in real time; The mobile terminal interface supports data transmission based on wireless communication and real-time reception of emergency instructions; The voice recognition input unit has built-in noise suppression and natural language processing algorithms, which supports users to query alarm information, adjust simulation parameters and start emergency plans through voice commands; The data interaction interface is compatible with HTTPS and WebSocket protocols, realizing real-time information interaction and command issuance with other system modules and external emergency platforms.

Images