An artificial intelligence-based hydropower station fire emergency evacuation guidance system

Abstract

This invention relates to an artificial intelligence-based fire emergency evacuation guidance system for hydropower stations, comprising: a heat detector, a smoke detector, an artificial intelligence module, emergency indicator lights, and a voice broadcasting device. The artificial intelligence module includes: a data acquisition unit that collects fire data and fire addresses; a fire alarm unit that performs fusion analysis on the fire data and fire addresses; an evacuation route generation unit that generates emergency evacuation routes based on the regional fire situation; a route optimization unit that locates and supplements the fire location based on the fire's impact range and updates the routes within the fire's impact range; and an emergency evacuation unit that performs emergency evacuation according to the emergency evacuation routes. This system is beneficial in reminding staff to respond promptly after a fire occurs, shortening evacuation time, and ensuring the safety of personnel and equipment / property.

Description

Technical Field

[0001] This invention relates to the field of fire monitoring technology, and in particular to an artificial intelligence-based fire emergency evacuation guidance system for hydropower stations. Background Technology

[0002] Hydropower stations, as vital national energy infrastructure, typically comprise numerous complex areas including underground powerhouses, cable tunnels, transformer rooms, and main control rooms. Their intricate internal structures and enclosed environments make them extremely difficult to evacuate in the event of a fire, due to rapid smoke dissipation, low visibility, and severe personnel evacuation. Therefore, establishing an efficient and reliable fire emergency evacuation guidance system is of paramount importance for ensuring the safety of personnel working within hydropower stations.

[0003] Chinese Patent Publication No. CN114419816A discloses a method for determining evacuation routes and an intelligent fire protection system. Upon detecting a fire signal, the system identifies the fire area and the area to be evacuated. Based on the fire's spread, it determines a first optimal safety exit corresponding to the area to be evacuated and a second optimal safety exit corresponding to the fire area. The shortest path from the area to be evacuated to the first optimal safety exit is used as the first evacuation route, and the shortest path from the fire area to the second optimal safety exit is used as the second evacuation route. The system controls the direction of the emergency evacuation indicator lights corresponding to the area to be evacuated based on the first evacuation route and the direction of the emergency evacuation indicator lights corresponding to the fire area based on the second evacuation route. However, this Chinese patent suffers from the following problems: a lack of precise three-dimensional positioning capability for the complex structure of hydropower stations, difficulty in quickly determining the fire source location during a fire, a lack of dynamic monitoring of the fire's impact range, a lack of a continuous path optimization and update mechanism, and an inability to dynamically adjust the guiding direction according to the fire's development. Summary of the Invention

[0004] To address these issues, the present invention provides an artificial intelligence-based fire emergency evacuation guidance system for hydropower stations, which overcomes the problems of existing technologies such as lack of three-dimensional precise positioning capability for the complex structure of hydropower stations, difficulty in quickly determining the location of the fire source when a fire occurs, lack of dynamic monitoring of the fire's impact range, lack of a continuous path optimization and update mechanism, and inability to dynamically adjust the guidance direction according to the development of the fire.

[0005] To achieve the above objectives, the present invention provides an artificial intelligence-based hydropower station fire emergency evacuation guidance system, comprising: a heat detector, a smoke detector, an artificial intelligence module, an emergency indicator light, and a voice broadcasting device. The heat detector and the smoke detector are connected to the artificial intelligence module to collect temperature and smoke signals after a fire occurs in the hydropower station and upload them to the artificial intelligence module. The emergency indicator light and the voice broadcasting device are connected to the artificial intelligence module to guide personnel to evacuate in a timely manner according to the prompts when a fire occurs. The artificial intelligence module is used to plan and generate emergency evacuation routes.

[0006] Furthermore, the artificial intelligence module includes: The data acquisition unit is used to collect fire data and fire addresses; The fire alarm unit is used to integrate and analyze fire data and fire addresses, and to accurately locate the fire situation in the area based on the fire situation in the area. Evacuation route generation unit, used to generate emergency evacuation routes based on the fire situation in the area; The path optimization unit is used to obtain the fire impact range, and to locate and supplement the fire location based on the fire impact range; it is also used to update the fire impact range based on the emergency evacuation route information. An emergency evacuation unit is used to conduct emergency evacuations according to emergency evacuation routes.

[0007] Furthermore, when the data acquisition unit collects the fire address, it establishes a three-dimensional model structure that completely corresponds to the geometric dimensions of the hydropower station based on BIM technology, and maps the addresses of each room in the hydropower station to the three-dimensional model structure. The specific method for constructing and mapping the three-dimensional model structure is as follows: Step A01: Collect the physical address of the hydropower station using a 3D laser scanner, obtain the physical address of the hydropower station, and import the physical address of the hydropower station into 3D modeling software to build a BIM model; Step A02: Select an absolute coordinate reference point Q (xq, yq, zq) within the hydropower station area, and establish a three-dimensional rectangular coordinate system with the absolute coordinate reference point Q as the origin; Step A03: Spatialize the physical address of the hydropower station, define the spatial range of each room, and use the bottom center point of the room door as the room positioning anchor point P(xp,yp,zp); Step A04: Mark the installation locations of the heat fire detector, smoke fire detector, emergency indicator light, and voice broadcasting device in the 3D model structure.

[0008] Furthermore, when the fire alarm unit performs fusion analysis on fire data and fire address, the fusion analysis method specifically includes: Step B01: Compare the temperature rise rate Vt and concentration rise rate Vs within the detector's detection range with preset temperature rise rate Vt0 and preset concentration rise rate Vs0. Based on the comparison results, determine the fire situation in the area, and further verify the fire situation based on the determination results. Wherein: When Vt < Vt0 and Vs < Vs0, the fire situation in the area is determined to be safe, and no further verification is required. Otherwise, the fire situation in the area is determined to be abnormal, and further verification is required; Step B02 further verifies the regional fire situation by comparing the temperature T and smoke concentration S within the detector's detection range with preset temperature T0 and preset smoke concentration S0. Based on the comparison results, the regional fire situation is assessed, and precise location is determined according to the assessment results. When T≥T0 and S<S0, the fire situation in the area is determined to be in the open flame stage, and precise location is performed. When T < T0 and S ≥ S0, the fire situation in the area is determined to be in the smoldering stage, and precise location is performed. When T < T0 and S < S0, the fire situation in the area is determined to be normal, and no precise location is performed. When T≥T0 and S≥S0, the fire situation in the area is determined to be a fire, and precise location is performed.

[0009] Furthermore, when the fire alarm unit performs precise location based on the regional fire situation, it queries the ID of the detector that first triggered the alarm and its installation coordinates. The installation location of this detector is used as the initial location coordinates of the fire occurrence point. The detector ID is then associated with its corresponding room location anchor point P(x,y,z), and the room location anchor point coordinates are used as the regional reference coordinates F(xf,yf,zf) of the fire occurrence point. If multiple detectors in the same room respond simultaneously, a weighted centroid positioning algorithm is used to calculate the precise coordinates J(xf,yf,zf) of the fire occurrence point. , , ), , , ,in The number of detectors, The weights of individual detectors are used to obtain the fire situation in the area.

[0010] Furthermore, when the evacuation route generation unit generates emergency evacuation routes based on the regional fire situation, it determines the nodes and edges in the path topology network according to the three-dimensional model structure, obtains the path length L from each room to the main gate, and stores the path length L data in the database. When a fire occurs in a certain room, the path length L data is retrieved from the database and calculated using the formula min{L1, L2…Ln}, where Ln is the path length from a certain room to each main gate exit, thus obtaining the emergency evacuation route.

[0011] Furthermore, when the evacuation route generation unit generates an emergency evacuation route based on the regional fire situation, it obtains the shortest distance d from the precise coordinates J of the fire occurrence point to the emergency evacuation route, compares the shortest distance d with a first preset distance d0, judges the status of the emergency evacuation route based on the comparison result, and outputs the emergency evacuation route based on the judgment result, wherein: When d≥d0, the emergency evacuation route is determined to be safe, and the emergency evacuation route is output. Otherwise, if the emergency evacuation route is deemed dangerous, it will not be output and a new emergency evacuation route will be recalculated.

[0012] Furthermore, when the path optimization unit obtains the fire impact range, it compares the temperature T with the preset danger temperature T1, determines the category of the fire impact range based on the comparison result, and supplements the location of the fire based on the determination result, wherein: When T≥T1, the fire impact area is classified as the fire core area, and the coordinates of the fire core area are used as supplementary coordinates of the fire occurrence point. Otherwise, the smoke concentration S is compared with the preset impact concentration S1. Based on the comparison results, the extent of the fire's impact is assessed, and the location of the fire is supplemented based on the assessment results. If S≥S1, the category of the fire-affected area is determined to be the fire-affected zone. The coordinates of the fire-affected zone are used as the coordinates of the fire-occurrence point to supplement the fire-affected zone, and the second preset distance d1 is output as the first preset distance d0. Otherwise, the fire's impact area will be classified as a safety warning zone, and the location of the fire will not be supplemented.

[0013] Furthermore, when the path optimization unit updates the fire impact range based on the emergency evacuation path information, for each currently used emergency evacuation path, a sliding window prediction algorithm is used to assess the path's safety over a future period. The assessment method specifically includes: Step C01: Divide the emergency evacuation route into several continuous route segments. Calculate the risk index R for each route segment. Risk index R = α × Tp + β × Sp + γ × dp, where Tp is the average temperature of each detector in the route segment, Sp is the average smoke concentration of each detector in the route segment, dp is the reciprocal of the shortest distance between the route segment and the fire point, α is the temperature weight, β is the smoke weight, γ is the distance weight, and α + β + γ = 1. Step C02 involves comparing the risk index R with the preset risk R0, assessing the emergency evacuation routes based on the comparison results, and updating the routes to reflect the fire's impact range. When R≥R0, the emergency evacuation route is determined to be about to fail. The route segment is added to the fire-affected area, and the emergency evacuation route is regenerated. Otherwise, the emergency evacuation route is deemed safe to use, and no route updates are made for the fire-affected area.

[0014] Furthermore, when the emergency evacuation unit conducts emergency evacuation according to the emergency evacuation route, it sends commands to the emergency indicator lights and the voice broadcasting device to activate the flashing of the emergency indicator lights and the voice broadcasting device to broadcast messages. Both work simultaneously to guide personnel in each room of the hydropower station to evacuate along the emergency evacuation route.

[0015] Compared with the prior art, the beneficial effects of the present invention are that it has the ability to accurately locate the complex structure of hydropower stations in three dimensions, quickly determine the location of the fire source when a fire occurs, dynamically monitor the scope of fire impact in real time, ensure continuous optimization of the path and dynamically adjust the guiding direction according to the development of the fire.

[0016] In particular, the data acquisition unit collects fire addresses, which helps to spatialize and digitize physical addresses, establish real-time correlation between equipment and location, and synchronously obtain the precise location of the fire, providing support for dynamic path planning and optimization. Furthermore, the fire alarm unit performs fusion analysis of fire data and fire addresses, which facilitates multi-point fusion, achieves high-precision positioning and spatial correlation, and strengthens fire situation awareness. The evacuation path generation unit generates emergency evacuation paths based on the regional fire situation, which helps to achieve rapid response through pre-stored basic paths and avoid potential risks caused by paths too close to the fire source. The path optimization unit obtains the fire impact range and updates the paths based on the emergency evacuation path information, which helps to classify the fire impact range, make situation awareness clearer, significantly improve early warning capabilities, and adapt to complex fire changes. Finally, the emergency evacuation unit conducts emergency evacuations based on the emergency evacuation paths, which helps to form a dual visual and auditory guidance mechanism to improve the reliability and coverage of information transmission. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the structure of the artificial intelligence-based hydropower station fire emergency evacuation guidance system in this embodiment; Figure 2 This is a schematic diagram of the structure of the artificial intelligence module implemented in this system; Figure 3 This is a schematic diagram of the emergency indicator light in this embodiment. Detailed Implementation

[0018] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0019] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0020] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0021] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0022] Please see Figure 1 The diagram shown is a structural schematic of the artificial intelligence-based fire emergency evacuation guidance system for hydropower stations in this embodiment. The system includes: a heat detector, a smoke detector, an artificial intelligence module, an emergency indicator light, and a voice broadcasting device. The heat detector and smoke detector are connected to the artificial intelligence module to collect temperature and smoke signals after a fire occurs in the hydropower station and upload them to the artificial intelligence module. The emergency indicator light and voice broadcasting device are connected to the artificial intelligence module to guide personnel to evacuate in a timely manner according to the prompts when a fire occurs. The artificial intelligence module is used to plan and generate emergency evacuation routes.

[0023] Specifically, the AI-based hydropower station fire emergency evacuation guidance system is applied to fire emergency evacuation in hydropower stations. It possesses the capability for precise three-dimensional positioning of the complex structure of the hydropower station, quickly determining the location of the fire source during a fire, dynamically monitoring the fire's impact range in real time, ensuring continuous path optimization, and dynamically adjusting guidance directions according to fire development. In particular, by collecting fire addresses through a data acquisition unit, it facilitates the spatialization and digitization of physical addresses, establishing real-time association between equipment and location, and synchronously obtaining the precise location of the fire, providing support for dynamic path planning and optimization. Furthermore, the fire alarm unit performs fusion analysis of fire data and fire addresses, which is beneficial for multi-point... The system integrates various technologies to achieve high-precision positioning and spatial correlation, enhancing fire situation awareness. The evacuation route generation unit generates emergency evacuation routes based on regional fire conditions, facilitating rapid response through pre-stored basic routes and avoiding potential risks caused by routes too close to the fire source. Furthermore, the route optimization unit acquires the fire impact range and updates routes based on emergency evacuation information, enabling fire impact range classification, clearer situation awareness, significantly improved early warning capabilities, and adaptation to complex fire changes. Finally, the emergency evacuation unit conducts emergency evacuations based on the emergency evacuation routes, forming a dual audio-visual guidance mechanism to improve the reliability and coverage of information transmission.

[0024] Please see Figure 2 As shown, this is a structural diagram of the artificial intelligence module in this embodiment. The artificial intelligence module includes: The data acquisition unit is used to collect fire data and fire addresses; The fire alarm unit is used to integrate and analyze fire data and fire addresses, and to accurately locate the fire situation in the area based on the fire situation in the area. Evacuation route generation unit, used to generate emergency evacuation routes based on the fire situation in the area; The path optimization unit is used to obtain the fire impact range, and to locate and supplement the fire location based on the fire impact range; it is also used to update the fire impact range based on the emergency evacuation route information. An emergency evacuation unit is used to conduct emergency evacuations according to emergency evacuation routes.

[0025] Specifically, when the data acquisition unit collects fire data, it obtains temperature data and smoke concentration data from the heat detector and the smoke detector to obtain fire data. When the data acquisition unit collects fire addresses, it establishes a three-dimensional model structure that completely corresponds to the geometric dimensions of the hydropower station based on BIM technology. The addresses of each room within the hydropower station are then mapped to the three-dimensional model structure. The specific method for constructing and mapping the three-dimensional model structure is as follows: Step A01: Collect the physical address of the hydropower station using a 3D laser scanner, obtain the physical address of the hydropower station, and import the physical address of the hydropower station into 3D modeling software to build a BIM model; Step A02: Select an absolute coordinate reference point Q (xq, yq, zq) within the hydropower station area, and establish a three-dimensional rectangular coordinate system with the absolute coordinate reference point Q as the origin; Step A03: Spatialize the physical address of the hydropower station, define the spatial range of each room, and use the bottom center point of the room door as the room positioning anchor point P(xp,yp,zp); Step A04: Mark the installation locations of the heat fire detector, smoke fire detector, emergency indicator light, and voice broadcasting device in the 3D model structure.

[0026] Specifically, the temperature data includes temperature T and the rate of temperature rise Vt; the smoke concentration data includes smoke concentration S and the rate of concentration rise St; the 3D model structure based on BIM technology that completely corresponds to the geometric dimensions of the hydropower station refers to a digital 3D replica constructed using precise measurement technology that maintains strict consistency with the geometric dimensions and spatial position of the actual hydropower station, achieving a 1:1 mapping between the virtual and real worlds. Furthermore, the BIM model's precision meets the relevant standards for digital twin design in the water conservancy industry, such as a precision level of no less than LOD 2.0 for the main building and no less than LOD 3.0 for key electromechanical equipment. The 3D laser scanner refers to a scanner that utilizes the principle of laser ranging. This embodiment does not limit the model of the 3D laser scanner; those skilled in the art can choose according to the actual situation, such as using a ground-based 3D laser scanner like the FARO. The Focus series performs detailed scanning of the indoor structure of hydropower stations. For long, narrow, and complex cable corridors or shafts, a handheld SLAM laser scanner can be used for rapid mobile scanning. The physical address of the hydropower station refers to the actual location of various functional areas, rooms, passages, equipment installation points, etc., within the hydropower station in physical space. The 3D modeling software refers to a computer application capable of importing point cloud data and performing 3D geometric modeling, semantic information input, model integration, and management. This embodiment does not limit the specific brand and model of the 3D modeling software; those skilled in the art can choose according to actual needs, such as Autodesk Revit. The BIM model refers to Building... InformationModeLing, or Building Information Modeling, uses the absolute coordinate reference point Q, a unique and permanent spatial reference point selected within the actual physical space of the hydropower station. This serves as the origin of the entire 3D model's spatial coordinate system, with xq=0, yq=0, and zq=0. Spatial coding refers to converting human-readable, hierarchical address descriptions into computer-recognizable and computationally calculable structured codes, which are then associated with specific spatial areas in the 3D model. These codes include area codes (e.g., main plant area A, switchyard area B, office area C), floor codes (e.g., underground level 3-3, above-ground level 101), and functional zone codes (e.g., generator). Layer G, turbine layer T, cable layer C, room number, such as 01, 02. Example: The spatial code for the electrical distribution room next to generator No. 3 on the main plant floor can be represented as "A-01-G-03-EP01", where A represents the main plant area, 01 represents the first floor, G represents the generator floor functional area, 03 represents generator section No. 3, and EP01 represents electrical distribution room No. 01. The delineation refers to the operation of picking key points, drawing boundary lines, and generating closed surfaces in a 3D modeling software environment based on accurate point cloud data or design drawings. This includes room outline definition and information binding. The outline definition refers to defining a closed spatial volume along the internal surfaces of the room's walls, floors, and ceilings.The information binding refers to the one-to-one binding of the spatial range data of the physical address rooms of the hydropower station. The bottom center point of the room doorway refers to the intersection of the midpoint of the doorway's width and the bottom of the doorway at the opening of each room, serving as the feature positioning point. The positioning anchor point P of the room refers to the reference coordinate point representing the precise location of the room in the 3D model, denoted as P(xp,yp,zp). The annotation refers to the precise identification of the installation locations of all heat detectors, smoke detectors, emergency indicator lights, and voice broadcasting devices in the constructed 3D model, generating a unique ID for each device, and binding and storing it with the coordinate data of that location and the address code of the room to which it belongs.

[0027] Specifically, when the data acquisition unit collects fire data and fire addresses, it helps to realize the spatialization and digitization of physical addresses, establish real-time association between equipment and location, and synchronously obtain the precise location of the fire, providing support for dynamic path planning and optimization.

[0028] Specifically, when the fire alarm unit performs fusion analysis on fire data and fire address, the fusion analysis method is as follows: Step B01: Compare the temperature rise rate Vt and concentration rise rate Vs within the detector's detection range with preset temperature rise rate Vt0 and preset concentration rise rate Vs0. Based on the comparison results, determine the fire situation in the area, and further verify the fire situation based on the determination results. Wherein: When Vt < Vt0 and Vs < Vs0, the fire situation in the area is determined to be safe, and no further verification is required. Otherwise, the fire situation in the area is determined to be abnormal, and further verification is required; Step B02 further verifies the regional fire situation by comparing the temperature T and smoke concentration S within the detector's detection range with preset temperature T0 and preset smoke concentration S0. Based on the comparison results, the regional fire situation is assessed, and precise location is determined according to the assessment results. When T≥T0 and S<S0, the fire situation in the area is determined to be in the open flame stage, and precise location is performed. When T < T0 and S ≥ S0, the fire situation in the area is determined to be in the smoldering stage, and precise location is performed. When T < T0 and S < S0, the fire situation in the area is determined to be normal, and no precise location is performed. When T≥T0 and S≥S0, the fire situation in the area is determined to be a fire, and precise location is performed. When the fire alarm unit performs precise location based on the regional fire situation, it queries the ID of the detector that first triggered the alarm and its installation coordinates. The installation location of that detector is used as the initial location coordinates of the fire point. The detector ID is then associated with its room location anchor point P(x,y,z), and the room location anchor point coordinates are used as the regional reference coordinates F(xf,yf,zf) of the fire point. If multiple detectors in the same room respond simultaneously, a weighted centroid positioning algorithm is used to calculate the precise coordinates J(xf,yf,zf) of the fire point. , , ), , , ,in The number of detectors, The weights of individual detectors are used to obtain the fire situation in the area.

[0029] Specifically, the temperature rise rate Vt and concentration rise rate Vs within the detector's detection range refer to the instantaneous change rate obtained by the fire alarm unit through time difference calculation of continuously sampled temperature data and smoke concentration data. The preset temperature rise rate Vt0 refers to a preset value used to determine the fire situation in the area. This embodiment does not limit the preset temperature rise rate Vt0; those skilled in the art can select it according to actual conditions, such as making reasonable selections and dynamic adjustments based on factors like the environmental characteristics of different areas of the hydropower station, equipment operating conditions, and detector deployment density. For example, for densely cabled corridor areas, the preset temperature rise rate Vt0 can be set to 3℃ / min. In areas with electrical equipment such as transformer rooms and switch stations, the preset temperature rise rate Vt0 is set to 8℃ / min. In areas with high personnel activity such as offices and main control rooms, the preset temperature rise rate Vt0 is set to 2℃ / min. The preset concentration rise rate Vs0 refers to a preset value used to determine the fire situation in the area. This embodiment does not limit the preset concentration rise rate Vs0; those skilled in the art can select it according to actual conditions. For example, for cable corridors, the preset concentration rise rate Vs0 is set to 0.3% / min; for generator floors and transformer rooms, it is set to 2% / min; and for offices and conference rooms, it is set to a different value. 0 is set to 0.5% / min. The temperature T and smoke concentration S within the detector's detection range refer to the measured values ​​obtained by the fire alarm unit from the heat detector and smoke detector. The preset temperature T0 refers to a preset value used to determine the fire situation in the area. This embodiment does not limit the preset temperature T0; those skilled in the art can select it according to actual conditions. For example, for cable corridors, the normal operating temperature of cables is usually below 50℃, so the preset temperature T0 can be set to 70℃. For transformer rooms, the transformer oil temperature can reach 80-90℃ during normal operation, but once an internal fault occurs, the temperature may rapidly exceed 100℃, so the preset temperature T0 can be set to 100℃. In offices and meeting rooms, the normal ambient temperature is generally between 20-30℃. The preset temperature T0 is set to 60℃ to distinguish it from the high temperatures in summer or the normal temperature rise near radiators. For the generator floor, the surface temperature of the generator may reach 60-80℃ during operation. Therefore, the preset temperature T0 needs to be appropriately increased, and can be set to 90℃ to avoid false alarms caused by normal heating of the equipment. The preset smoke concentration S0 refers to the preset value used to judge the fire situation in the area. This embodiment does not limit the preset smoke concentration S0. Those skilled in the art can select it according to the actual situation. For example, according to the national standard point-type smoke detector, the alarm threshold of the smoke detector is usually set between 0.5dB / m and 2.Between 0 dB / m, the detector ID and its installation coordinates that first trigger the alarm refer to the pre-labeled unique identifier of the detector with the earliest response time among all alarm detectors and the precise three-dimensional coordinates of its installation location. The installation location of the detector refers to the installation point of the heat or smoke detector in the actual physical space. The detector ID being associated with its room positioning anchor point P means that during the three-dimensional model construction phase, each detector ID not only records its own installation coordinates but also establishes an association with the positioning anchor point P (x, y, z) of the room it belongs to. The number of detectors. This refers to the total number of detectors that respond simultaneously within the detection range of a single detector during the same fire event, and the weight of each individual detector. In the weighted centroid localization algorithm, this refers to the coefficient assigned to each response detector, reflecting the detector's contribution to the location of the fire. This embodiment does not assign weights to individual detectors. The settings are limited, and those skilled in the art can select according to the actual situation, such as setting it to the current temperature T of the detector based on the difference between the temperature and the ambient temperature. i The difference between the ambient temperature Te and the room temperature, i.e., w i =T i -Te, the larger the difference, the closer the detector is to the core of the fire source, and the higher the weight.

[0030] Specifically, when the fire alarm unit performs fusion analysis on fire data and fire address, and accurately locates the fire based on the regional fire situation, it facilitates multi-point fusion, achieves high-precision positioning and spatial correlation, and enhances fire situation awareness.

[0031] Specifically, when the evacuation route generation unit generates emergency evacuation routes based on the regional fire situation, it determines the nodes and edges in the path topology network according to the three-dimensional model structure, obtains the path length L from each room to the main gate, and stores the path length L data in the database. When a fire occurs in a room, the path length L data is retrieved from the database and calculated using the formula min{L1, L2…Ln}, where Ln is the path length from a room to each main gate exit, thus obtaining the emergency evacuation route. When the evacuation route generation unit generates an emergency evacuation route based on the regional fire situation, it obtains the shortest distance d from the precise coordinates J of the fire occurrence point to the emergency evacuation route. It compares the shortest distance d with a first preset distance d0, judges the status of the emergency evacuation route based on the comparison result, and outputs the emergency evacuation route based on the judgment result. Wherein: When d≥d0, the emergency evacuation route is determined to be safe, and the emergency evacuation route is output. Otherwise, if the emergency evacuation route is deemed dangerous, it will not be output and a new emergency evacuation route will be recalculated.

[0032] Specifically, the path topology network refers to a graph theory path network extracted and constructed based on a 1:1 scale 3D model of a hydropower station for path planning calculations. This graph theory path network transforms a complex 3D spatial structure into an abstract mathematical model that can be recognized and calculated by a computer, where nodes represent key locations in space, edges represent walkable paths between nodes, and each edge is assigned corresponding attribute information. Nodes represent key locations in path planning, edges represent walkable paths between nodes, and path length L refers to the distance from a positioning anchor point P(x, y, z) in a room to a safe exit node E(x, y, z) along the walkable edges in the path topology network. e ,y e , z e The shortest distance d is the sum of the physical distances of all edges traversed by the fire, where the shortest distance d refers to the precise coordinates J(J) from the point where the fire occurred. , , The shortest vertical distance from the calculated emergency evacuation route is d0. The preset first distance d0 is a preset value used to determine the situation of the emergency evacuation route. This embodiment does not limit the specific value of the preset first distance d0. Those skilled in the art can select it according to the spatial layout of different areas of the hydropower station. For example, for the spacious generator floor of the main plant, the preset first distance d0 can be set to 5 meters. For densely populated areas such as offices and main control rooms, in order to ensure personnel safety, the preset first distance d0 can be set to 4 meters.

[0033] Specifically, when the evacuation route generation unit generates emergency evacuation routes based on the regional fire situation, it is beneficial to achieve rapid response by pre-storing basic routes and avoid potential risks caused by routes being too close to the fire source.

[0034] Specifically, when the path optimization unit obtains the fire impact range, it compares the temperature T with the preset danger temperature T1, determines the category of the fire impact range based on the comparison result, and supplements the location of the fire based on the determination result, wherein: When T≥T1, the fire impact area is classified as the fire core area, and the coordinates of the fire core area are used as supplementary coordinates of the fire occurrence point. Otherwise, the smoke concentration S is compared with the preset impact concentration S1. Based on the comparison results, the extent of the fire's impact is assessed, and the location of the fire is supplemented based on the assessment results. If S≥S1, the category of the fire-affected area is determined to be the fire-affected zone. The coordinates of the fire-affected zone are used as the coordinates of the fire-occurrence point to supplement the fire-affected zone, and the second preset distance d1 is output as the first preset distance d0. Otherwise, the fire's impact area will be classified as a safety warning zone, and the location of the fire will not be supplemented. When the path optimization unit updates the fire impact range based on the emergency evacuation route information, for each currently used emergency evacuation route, a sliding window prediction algorithm is used to assess the safety of the route over a future period. The assessment method specifically includes: Step C01: Divide the emergency evacuation route into several continuous route segments. Calculate the risk index R for each route segment. Risk index R = α × Tp + β × Sp + γ × dp, where Tp is the average temperature of each detector in the route segment, Sp is the average smoke concentration of each detector in the route segment, dp is the reciprocal of the shortest distance between the route segment and the fire point, α is the temperature weight, β is the smoke weight, γ is the distance weight, and α + β + γ = 1. Step C02 involves comparing the risk index R with the preset risk R0, assessing the emergency evacuation routes based on the comparison results, and updating the routes to reflect the fire's impact range. When R≥R0, the emergency evacuation route is determined to be about to fail. The route segment is added to the fire-affected area, and the emergency evacuation route is regenerated. Otherwise, the emergency evacuation route is deemed safe to use, and no route updates are made for the fire-affected area.

[0035] Specifically, the preset danger temperature T1 refers to a preset value used to determine the category of fire impact range. This embodiment does not limit the preset danger temperature T1; those skilled in the art can select it according to actual conditions. For example, for areas with dense electrical equipment such as cable corridors and power distribution rooms, the preset danger temperature T1 can be set to 70℃; for transformer rooms and oil tank areas, T1 can be set to 100℃; for offices and main control rooms, T1 can be set to 60℃; and for generator floors, T1 can be set to 90℃. The preset impact concentration S1 refers to a preset value used to determine the category of fire impact range. This embodiment does not limit the preset impact concentration S1; those skilled in the art can select it according to actual conditions, such as according to national standards. The standard point-type smoke detector sets the preset impact concentration S1 to 0.5 dB / m. The second preset distance d1 refers to the adjusted safe distance threshold output by the path optimization unit when the smoke concentration exceeds the standard but the temperature has not yet reached the danger threshold. This embodiment does not limit the second preset distance d1; those skilled in the art can select it according to the actual situation. For example, for an open main factory building, d1 can be set to 3 meters; for a stairwell, d1 can be set to 2.5 meters. The currently used emergency evacuation route refers to the evacuation route that the emergency evacuation unit has issued instructions, the emergency indicator lights are on, the voice broadcasting device has been activated, and it is guiding the actual passage of personnel on site. The use of a sliding window pre- The prediction algorithm refers to a dynamic prediction method based on time series data. By setting a fixed-size time window, such as 30 seconds, and continuously sliding it forward, it uses historical and current data within the window to predict the risk change trend of each path segment over a future period. The average temperature Tp of each detector in the path segment refers to the arithmetic mean of the measured temperatures of all temperature detectors within the defined path segment at the current moment, calculated as the temperature characteristic value of that path segment. Similarly, the average smoke concentration Sp of each detector in the path segment refers to the arithmetic mean of the measured smoke concentration of all smoke detectors within the defined path segment at the current moment. The mean value, used as the characteristic value of smoke concentration for this path segment, is calculated as dp, which is the reciprocal of the shortest distance between the path segment and the fire location. This means calculating the shortest distance d between all points on the path segment and the precise coordinates J of the fire location, and then taking its reciprocal dp = 1 / d. The temperature weight α is the weighting coefficient assigned to the average temperature Tp, reflecting the degree of influence of temperature on path safety. During the open flame stage of a fire, temperature is the primary hazard factor, and α can be set to 0.6. During the smoldering stage, smoke is the primary hazard factor, and α can be appropriately reduced to 0.2. The smoke weight β is the weighting coefficient assigned to the average smoke concentration Sp, reflecting the degree of influence of smoke concentration on path safety. During the smoldering stage, β can be set to 0.The distance weight γ mentioned in section 6 refers to the weighting coefficient assigned to the reciprocal of distance dp, reflecting the impact of the proximity of the path segment to the fire source on safety. During the rapid development phase of a fire, the fire may spread quickly, increasing the importance of the distance factor. γ can be set to 0.3-0.4. The preset risk R0 refers to a preset value used to determine the situation of emergency evacuation routes. This embodiment does not limit the preset risk R0; those skilled in the art can select it according to actual conditions. If the original calculated value is used, a benchmark value needs to be obtained based on actual operational data. For example, 3-5 times the average value of R under normal operating conditions can be taken as the warning threshold.

[0036] Specifically, when the path optimization unit obtains the fire impact range and updates the fire impact range according to the emergency evacuation route, it is beneficial to classify the fire impact range, make the situational awareness clearer, greatly improve the early warning capability, and adapt to complex fire changes.

[0037] Specifically, when the emergency evacuation unit conducts emergency evacuation according to the emergency evacuation route, it sends commands to the emergency indicator lights and the voice broadcasting device to activate the flashing of the emergency indicator lights and the voice broadcasting device to broadcast messages. Both work simultaneously to guide personnel in each room of the hydropower station to evacuate along the emergency evacuation route.

[0038] Specifically, when the emergency evacuation unit conducts emergency evacuation according to the emergency evacuation route, it helps to form a dual audio-visual guidance mechanism to improve the reliability and coverage of information transmission.

[0039] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. An artificial intelligence-based fire emergency evacuation guidance system for a hydropower station, characterized in that, include: The system includes a heat detector, a smoke detector, an artificial intelligence module, an emergency indicator light, and a voice broadcasting device. The heat detector and smoke detector are connected to the artificial intelligence module to collect temperature and smoke signals after a fire occurs in the hydropower station and upload them to the artificial intelligence module. The emergency indicator light and voice broadcasting device are connected to the artificial intelligence module to guide personnel to evacuate in a timely manner according to the prompts when a fire occurs. The artificial intelligence module is used to plan and generate emergency evacuation routes.

2. The artificial intelligence based fire emergency evacuation guidance system for hydroelectric power plants as claimed in claim 1 wherein, The artificial intelligence module includes: The data acquisition unit is used to collect fire data and fire addresses; The fire alarm unit is used to integrate and analyze fire data and fire addresses, and to accurately locate the fire situation in the area based on the fire situation in the area. Evacuation route generation unit, used to generate emergency evacuation routes based on the fire situation in the area; The path optimization unit is used to obtain the fire impact range, and to locate and supplement the fire location based on the fire impact range; it is also used to update the fire impact range based on the emergency evacuation route information. An emergency evacuation unit is used to conduct emergency evacuations according to emergency evacuation routes.

3. The artificial intelligence based fire emergency evacuation guidance system for hydropower station as claimed in claim 2 wherein, When the data acquisition unit collects fire addresses, it establishes a three-dimensional model structure that completely corresponds to the geometric dimensions of the hydropower station based on BIM technology. The addresses of each room within the hydropower station are then mapped to the three-dimensional model structure. The specific method for constructing and mapping the three-dimensional model structure is as follows: Step A01: Collect the physical address of the hydropower station using a 3D laser scanner, obtain the physical address of the hydropower station, and import the physical address of the hydropower station into 3D modeling software to build a BIM model; Step A02: Select an absolute coordinate reference point Q (xq, yq, zq) within the hydropower station area, and establish a three-dimensional rectangular coordinate system with the absolute coordinate reference point Q as the origin; Step A03: Spatialize the physical address of the hydropower station, define the spatial range of each room, and use the bottom center point of the room door as the room positioning anchor point P(xp,yp,zp); Step A04: Mark the installation locations of the heat fire detector, smoke fire detector, emergency indicator light, and voice broadcasting device in the 3D model structure.

4. The artificial intelligence based fire emergency evacuation guidance system for hydropower station of claim 2, wherein, When the fire alarm unit performs fusion analysis on fire data and fire address, the fusion analysis method is specifically as follows: Step B01: Compare the temperature rise rate Vt and concentration rise rate Vs within the detector's detection range with preset temperature rise rate Vt0 and preset concentration rise rate Vs0. Based on the comparison results, determine the fire situation in the area, and further verify the fire situation based on the determination results. Wherein: When Vt < Vt0 and Vs < Vs0, the fire situation in the area is determined to be safe, and no further verification is required. Otherwise, the fire situation in the area is determined to be abnormal, and further verification is required; Step B02 further verifies the regional fire situation by comparing the temperature T and smoke concentration S within the detector's detection range with preset temperature T0 and preset smoke concentration S0. Based on the comparison results, the regional fire situation is assessed, and precise location is determined according to the assessment results. When T≥T0 and S<S0, the fire situation in the area is determined to be in the open flame stage, and precise location is performed. When T < T0 and S ≥ S0, the fire situation in the area is determined to be in the smoldering stage, and precise location is performed. When T < T0 and S < S0, the fire situation in the area is determined to be normal, and no precise location is performed. When T≥T0 and S≥S0, the fire situation in the area is determined to be a fire, and precise location is performed.

5. The artificial intelligence-based hydropower station fire emergency evacuation guidance system according to claim 4, characterized in that, When the fire alarm unit performs precise location based on the regional fire situation, it queries the ID of the detector that first triggered the alarm and its installation coordinates. The installation location of that detector is used as the initial location coordinates of the fire point. The detector ID is then associated with its corresponding room location anchor point P(xp,yp,zp), and the room location anchor point coordinates are used as the regional reference coordinates F(xf,yf,zf) of the fire point. If multiple detectors in the same room respond simultaneously, a weighted centroid positioning algorithm is used to calculate the precise coordinates J(xf,yf,zf) of the fire point. , , ), , , ,in The number of detectors, The weights of individual detectors are used to obtain the fire situation in the area.

6. The artificial intelligence-based hydropower station fire emergency evacuation guidance system according to claim 2, characterized in that, When the evacuation route generation unit generates emergency evacuation routes based on the regional fire situation, it determines the nodes and edges in the path topology network according to the three-dimensional model structure, obtains the path length L from each room to the main gate, and stores the path length L data in the database. When a fire occurs in a certain room, the path length L data is retrieved from the database and calculated using the formula min{L1, L2…Ln}, where Ln is the path length from a certain room to each main gate exit, thus obtaining the emergency evacuation route.

7. The artificial intelligence-based hydropower station fire emergency evacuation guidance system according to claim 6, characterized in that, When the evacuation route generation unit generates an emergency evacuation route based on the regional fire situation, it obtains the shortest distance d from the precise coordinates J of the fire occurrence point to the emergency evacuation route. It compares the shortest distance d with a first preset distance d0, judges the status of the emergency evacuation route based on the comparison result, and outputs the emergency evacuation route based on the judgment result. Wherein: When d≥d0, the emergency evacuation route is determined to be safe, and the emergency evacuation route is output. Otherwise, if the emergency evacuation route is deemed dangerous, it will not be output and a new emergency evacuation route will be recalculated.

8. The artificial intelligence-based hydropower station fire emergency evacuation guidance system according to claim 2, characterized in that, When the path optimization unit obtains the fire impact range, it compares the temperature T with the preset danger temperature T1, determines the category of the fire impact range based on the comparison result, and supplements the location of the fire based on the determination result, wherein: When T≥T1, the fire impact area is classified as the fire core area, and the coordinates of the fire core area are used as supplementary coordinates of the fire occurrence point. Otherwise, the smoke concentration S is compared with the preset impact concentration S1. Based on the comparison results, the extent of the fire's impact is assessed, and the location of the fire is supplemented based on the assessment results. If S≥S1, the category of the fire-affected area is determined to be the fire-affected zone. The coordinates of the fire-affected zone are used as the coordinates of the fire-occurrence point to supplement the fire-affected zone, and the second preset distance d1 is output as the first preset distance d0. Otherwise, the fire's impact area will be classified as a safety warning zone, and the location of the fire will not be supplemented.

9. The artificial intelligence-based hydropower station fire emergency evacuation guidance system according to claim 2, characterized in that, When the path optimization unit updates the fire impact range based on the emergency evacuation route information, for each currently used emergency evacuation route, a sliding window prediction algorithm is used to assess the safety of the route over a future period. The assessment method specifically includes: Step C01: Divide the emergency evacuation route into several continuous route segments. Calculate the risk index R for each route segment. Risk index R = α × Tp + β × Sp + γ × dp, where Tp is the average temperature of each detector in the route segment, Sp is the average smoke concentration of each detector in the route segment, dp is the reciprocal of the shortest distance between the route segment and the fire point, α is the temperature weight, β is the smoke weight, γ is the distance weight, and α + β + γ = 1. Step C02 involves comparing the risk index R with the preset risk R0, assessing the emergency evacuation routes based on the comparison results, and updating the routes to reflect the fire's impact range. When R≥R0, the emergency evacuation route is determined to be about to fail. The route segment is added to the fire-affected area, and the emergency evacuation route is regenerated. Otherwise, the emergency evacuation route is deemed safe to use, and no route updates are made for the fire-affected area.

10. The artificial intelligence-based hydropower station fire emergency evacuation guidance system according to claim 2, characterized in that, When the emergency evacuation unit conducts an emergency evacuation according to the emergency evacuation route, it sends commands to the emergency indicator lights and the voice broadcasting device to activate the flashing of the emergency indicator lights and the voice broadcasting device to broadcast messages. Both work simultaneously to guide personnel in each room of the hydropower station to evacuate along the emergency evacuation route.

Images