Intelligent decision-making method, system, equipment and medium for space emergency evacuation of underground track

Abstract

The invention relates to the technical field of emergency evacuation, and discloses an underground track space emergency evacuation intelligent decision-making method, system and device and a medium. The method comprises the following steps: acquiring real-time sensor data and historical evacuation data, and setting a nerve activation threshold and an energy consumption constraint; processing the real-time sensor data and the historical evacuation data based on the edge computing node to obtain an optimal evacuation path; obtaining an optical parameter adjustment instruction according to the real-time sensor data, the nerve activation threshold and the energy consumption constraint; and generating an LED driving signal and a voice synthesis instruction according to the optimal evacuation path and the optical parameter adjustment instruction, so as to drive a light source module of an optical guide identifier of an evacuation channel and carry out evacuation voice prompt. According to the method, the optical parameters and the evacuation path are dynamically adjusted through multi-mode sensing fusion, so that the emergency evacuation efficiency of the underground track space is improved, the evacuation time is shortened, and the path deviation rate is reduced.

Description

Technical Field

[0001] This invention relates to the field of emergency evacuation technology, and in particular to an intelligent decision-making method, system, equipment and medium for emergency evacuation in underground rail spaces. Background Technology

[0002] In the field of emergency evacuation, optical guidance signs are a commonly used method of evacuation guidance. However, there are still many shortcomings in traditional optical guidance signs. Traditional optical guidance only achieves physical-level indication and does not explore the emergency response enhancement mechanisms at the neurocognitive level.

[0003] For example, traditional static signage, with its fixed direction, can easily mislead people into dangerous areas and is unsuitable for scenarios involving the dynamic spread of fire. While dynamic signage and intelligent systems can adjust paths via a central controller, they rely on a single visual information channel, ignoring the impact of cognitive load and stress levels on decision-making. Optogenetics has limitations; current research focuses on neuronal manipulation in laboratory environments, failing to address the engineering challenges of large-scale, non-invasive light stimulation. Hardware design flaws include: dynamic signage is susceptible to reduced visibility due to glare or ambient light interference, and lacks spectral-frequency co-optimization for neural activation. Summary of the Invention

[0004] The main objective of this invention is to provide an intelligent decision-making method, system, device, and medium for emergency evacuation in underground rail spaces, aiming to solve at least one of the aforementioned technical problems.

[0005] To achieve the above objectives, the present invention provides an intelligent decision-making method for emergency evacuation in underground rail spaces, comprising:

[0006] Acquire real-time sensor data and historical evacuation data, and set neural activation thresholds and energy consumption constraints;

[0007] The optimal evacuation path is obtained by processing the real-time sensor data and historical evacuation data using edge computing nodes.

[0008] The optical parameter adjustment command is obtained based on the real-time sensor data, neural activation threshold, and energy consumption constraint.

[0009] Based on the optimal evacuation path and light parameter adjustment instructions, LED driving signals and voice synthesis instructions are generated to drive the light source module of the optical guide sign of the evacuation channel and provide evacuation voice prompts.

[0010] In some embodiments, the wavelength range of the optical guide mark is 450±5nm, and the pulse frequency range is 8±0.5Hz;

[0011] The light source module includes a high-density LED array, and the light source module has a built-in optical diffusion film.

[0012] In some embodiments, acquiring real-time sensor data and historical evacuation data includes:

[0013] Real-time sensor data is acquired based on sensor networks; wherein, the real-time sensor data includes millimeter-wave radar population density data, CO concentration gradient data, and building structure risk coefficient;

[0014] Historical evacuation data is obtained based on building information modeling; wherein, the historical evacuation data includes turbulence model parameters from evacuation drills;

[0015] The real-time sensor data is transmitted to the edge computing node via a cognitive packet network.

[0016] In some embodiments, the step of processing the real-time sensor data and historical evacuation data based on edge computing nodes to obtain the optimal evacuation path includes:

[0017] Kalman filtering is performed on the real-time sensor data based on edge computing nodes, and a real-time dynamic environment map is generated by fusing the filtered real-time sensor data.

[0018] The real-time dynamic environment map and historical evacuation data are input into the neural symbolic model to obtain the optimal evacuation route.

[0019] In some embodiments, inputting the real-time dynamic environment map and historical evacuation data into the neural symbolic model to obtain the optimal evacuation route includes:

[0020] The real-time dynamic environment map and historical evacuation data are input into the symbolic logic layer to obtain the shortest path candidate set;

[0021] The path safety coefficient of the shortest path candidate set is evaluated based on the DQN neural network to obtain the optimal evacuation path; wherein, the optimal evacuation path includes dynamic arrow direction and voice command.

[0022] In some embodiments, obtaining the optical parameter adjustment command based on the real-time sensor data, neural activation threshold, and energy consumption constraints includes:

[0023] The ambient illuminance and remaining battery capacity are obtained based on the real-time sensor data.

[0024] The ambient illuminance, remaining battery capacity, neural activation threshold, and energy consumption constraints are input into the DNN prediction model to obtain light parameter adjustment instructions.

[0025] In some embodiments, inputting the ambient illuminance, remaining battery capacity, neural activation threshold, and energy consumption constraint into the DNN prediction model to obtain light parameter adjustment instructions includes:

[0026] The ambient illuminance, remaining battery capacity, and neural activation threshold are normalized to obtain normalized data.

[0027] The normalized data is input into the multilayer fully connected network of the DNN prediction model and adjusted based on the energy consumption constraint to output the frequency, duty cycle and light intensity.

[0028] The light parameter adjustment command is generated based on the frequency, duty cycle, and light intensity.

[0029] Furthermore, to achieve the above objectives, this invention also proposes an intelligent decision-making system for emergency evacuation in underground rail spaces, comprising:

[0030] The data acquisition module is used to acquire real-time sensor data and historical evacuation data, and to set neural activation thresholds and energy consumption constraints.

[0031] The path calculation module is used to process the real-time sensor data and historical evacuation data based on edge computing nodes to obtain the optimal evacuation path.

[0032] The instruction calculation module is used to obtain optical parameter adjustment instructions based on the real-time sensor data, neural activation threshold, and energy consumption constraints.

[0033] The control execution module is used to generate LED driving signals and voice synthesis commands based on the optimal evacuation path and light parameter adjustment instructions, so as to drive the light source module of the optical guidance sign of the evacuation channel and provide evacuation voice prompts.

[0034] Furthermore, to achieve the above objectives, the present invention also proposes an electronic device, which includes: a memory, a processor, and an intelligent decision-making program for emergency evacuation of underground rail spaces stored in the memory and executable on the processor, wherein the intelligent decision-making program for emergency evacuation of underground rail spaces is configured to implement the intelligent decision-making method for emergency evacuation of underground rail spaces as described above.

[0035] Furthermore, to achieve the above objectives, the present invention also proposes a storage medium storing an intelligent decision-making program for emergency evacuation in underground rail spaces. This intelligent decision-making program is used to enable the processor to implement the intelligent decision-making method for emergency evacuation in underground rail spaces as described above.

[0036] This invention provides an intelligent decision-making method for emergency evacuation in underground rail spaces, comprising: acquiring real-time sensor data and historical evacuation data, setting neural activation thresholds and energy consumption constraints; processing the real-time sensor data and historical evacuation data based on edge computing nodes to obtain the optimal evacuation path; obtaining optical parameter adjustment instructions based on the real-time sensor data, neural activation thresholds, and energy consumption constraints; and generating LED driving signals and voice synthesis instructions based on the optimal evacuation path and optical parameter adjustment instructions to drive the light source modules of optical guidance signs in the evacuation channel and provide evacuation voice prompts. In this invention, multimodal perception fusion is used to dynamically adjust optical parameters and evacuation paths, thereby improving the efficiency of emergency evacuation in underground rail spaces, reducing evacuation time, and lowering path deviation rates. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the structure of an electronic device in the hardware operating environment involved in the embodiments of the present invention;

[0038] Figure 2 This is a flowchart illustrating an embodiment of the intelligent decision-making method for emergency evacuation in underground rail spaces according to the present invention.

[0039] Figure 3 This is a schematic diagram of the modules involved in the emergency evacuation scenario according to the embodiments of the present invention;

[0040] Figure 4 This is a schematic diagram of the neural activation mechanism involved in the embodiments of the present invention;

[0041] Figure 5 This is a schematic diagram of the LED array arrangement involved in the embodiment of the present invention;

[0042] Figure 6 This is a schematic diagram of the technical route involved in the embodiments of the present invention;

[0043] Figure 7 This is a structural block diagram of an embodiment of the intelligent decision-making system for emergency evacuation in underground rail space according to the present invention.

[0044] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0046] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0047] Furthermore, the use of terms such as "first" and "second" in this invention is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention.

[0048] Reference Figure 1 , Figure 1 This is a schematic diagram of the electronic device structure of the hardware operating environment involved in the embodiments of the present invention.

[0049] like Figure 1 As shown, the electronic device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen or an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wireless-Fidelity (Wi-Fi) interface). The memory 1005 may be high-speed random access memory (RAM) or stable non-volatile memory (NVM), such as a disk drive. The memory 1005 may also optionally be a storage device independent of the aforementioned processor 1001.

[0050] Those skilled in the art will understand that Figure 1 The structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0051] like Figure 1 As shown, the memory 1005, which serves as a storage medium, may include an operating system, a network communication module, a user interface module, and an intelligent decision-making program for emergency evacuation in underground rail spaces.

[0052] exist Figure 1 In the electronic device shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the electronic device of the present invention can be set in the electronic device, and the electronic device calls the intelligent decision-making program for emergency evacuation of underground rail space stored in the memory 1005 through the processor 1001, and executes the intelligent decision-making method for emergency evacuation of underground rail space provided in the embodiment of the present invention.

[0053] This invention proposes an intelligent decision-making method, system, equipment, and medium for emergency evacuation in underground rail spaces.

[0054] This invention provides an intelligent decision-making method for emergency evacuation in underground rail spaces, referring to... Figure 2 , Figure 2 This is a flowchart illustrating an embodiment of the intelligent decision-making method for emergency evacuation in underground rail spaces according to the present invention.

[0055] like Figure 2 As shown, the intelligent decision-making method for emergency evacuation in underground rail spaces includes:

[0056] Step S100: Acquire real-time sensor data and historical evacuation data, and set neural activation thresholds and energy consumption constraints;

[0057] Step S200: Process the real-time sensor data and historical evacuation data based on edge computing nodes to obtain the optimal evacuation path;

[0058] Step S300: Obtain the optical parameter adjustment command based on the real-time sensor data, neural activation threshold, and energy consumption constraint;

[0059] Step S400: Generate LED driving signals and voice synthesis commands according to the optimal evacuation path and light parameter adjustment instructions to drive the light source module of the optical guide sign of the evacuation channel and provide evacuation voice prompts.

[0060] It should be noted that this embodiment uses the implementation of an intelligent decision-making method for emergency evacuation in underground rail space through an intelligent evacuation system as an example for illustration. (Reference) Figure 3The emergency evacuation scenario illustrated in the diagram integrates various modules of a smart evacuation system. This system incorporates dynamic path planning algorithms (e.g., deep reinforcement learning and fuzzy logic), crowd density monitoring (cascaded CNN models), and real-time environmental perception (wireless sensor networks). The hardware architecture of the smart evacuation system can adopt a layered design, consisting of a sensor network, edge computing nodes, an FPGA control module, and a high-density LED light source module. For example, the sensor network (e.g., millimeter-wave radar and CO sensors) transmits real-time data to the edge computing nodes (typically deployed in building electrical rooms or nearby low-voltage wiring shafts) via a cognitive packet network (CPN). The edge computing nodes execute deep reinforcement learning algorithms to generate evacuation path parameters and send control commands to the FPGA control module via a high-speed bus (e.g., PCIe). The FPGA control module dynamically adjusts the pulse frequency (5-10Hz), duty cycle (10%-90%), and light intensity (50-200 lux) of the LED array according to the commands. The architecture of the smart evacuation system can conform to the GB 51309-2018 standard and supports integration with BIM (Building Information Modeling) systems. The following describes the specific steps.

[0061] In one embodiment, the wavelength range of the optical guide mark is 450±5nm, and the pulse frequency range is 8±0.5Hz;

[0062] The light source module includes a high-density LED array, and the light source module has a built-in optical diffusion film.

[0063] It is understandable that optical guidance signs are a commonly used method in emergency evacuation. However, current optical guidance only provides physical indication and does not explore the neurocognitive mechanisms that enhance emergency response. To overcome the shortcomings of traditional optical guidance signs, this embodiment proposes an innovative neuroscience-based emergency evacuation optical guidance sign. Supported by biological experimental data and verified by optogenetic principles, it uses optical guidance signs of specific wavelengths and frequencies (e.g., 450nm, 8Hz blue light pulses) to activate the brain's emergency response areas, thereby enhancing the crisis cognition and emergency response capabilities of evacuees. This is of great significance in optimizing the guidance efficiency of optical signs, shortening evacuation decision-making time, improving evacuation success rates, and ensuring the safety of evacuees, and it is technically feasible.

[0064] Specifically, this embodiment proposes a mapping relationship between the optical parameters of the optical guidance sign and the neural response: An activation threshold model is established for the amygdala (fear response) and prefrontal cortex (decision-making) based on a combination of 450±5nm blue light and 8±0.5Hz pulses. By utilizing 450±5nm blue light to activate the melanopsin pathway in the retina, crisis cognition is enhanced through regions such as the thalamus and hippocampus. Combined with the simultaneous activation of the visual cortex (V1 area) and locus coeruleus at an 8±0.5Hz pulse frequency, alertness is improved.

[0065] It is understandable that, such as Figure 4 As shown, the improved neural activation mechanism of the optically guided markers in this embodiment is as follows: For the thalamus-cortex pathway, 8Hz pulsed light activates the V1 area through the retina-lateral geniculate body pathway, enhancing spatial orientation ability. The optimal frequency is 8±0.5Hz (7.5-8.5Hz), matching the theta wave resonance frequency (4-8Hz) of the thalamus-cortex pathway, with a limiting range of 5-10Hz. The low-frequency band (5-7Hz) is used for calming emotions, while the high-frequency band (8-10Hz) enhances alertness. For the locus coeruleus-norepinephrine system: 450nm blue light stimulates melanopsin, triggering the release of norepinephrine from the locus coeruleus, enhancing alertness and reaction speed. The core wavelength is 450±5nm (445-455nm), covering the maximum absorption peak (λ) of melanopsin. max =450nm), expandable to a compatible range of 430-470nm, requiring optical filters to suppress wavelength shift.

[0066] This embodiment presents an innovative neuroscience-based optical guidance signage for emergency evacuation. This technical solution utilizes optical guidance signs of specific wavelengths and frequencies to activate the brain's emergency response areas, enhancing the crisis awareness and emergency response capabilities of evacuees. It is significant in optimizing the guidance efficiency of optical signs, shortening evacuation decision-making time, improving evacuation success rates, and ensuring the safety of evacuees. This solution demonstrates significant technological advancement and strong practicality, innovating research on evacuation guidance signs in the field of emergency evacuation. It aligns with the current development trend of smart city infrastructure and possesses broad application value and prospects.

[0067] For example, the light source module includes a high-density LED array, and the light source module has a built-in optical diffusion film. For example... Figure 5As shown, the LED array is arranged using flip-chip COB (Chip-on-Board) technology to achieve high-density integration. The LED chips are arranged in a matrix with alternating red and blue LEDs (spacing ≤1mm), and the row-column staggered layout improves the uniformity of photon flux density (PPFD). The LED density is ≥10,000 points per square meter, and the light field distribution angle is optimized to 120° using a microlens array. The optical diffusion film is mounted close to the LED array surface and consists of a PET substrate and stacked optical microstructures. The microstructure height is ≥1μm, and the platform width varies gradually (decreasing from the center to the periphery). Multiple refractions weaken local hotspots, resulting in a light intensity distribution standard deviation ≤5%.

[0068] In one embodiment, acquiring real-time sensor data and historical evacuation data includes: acquiring real-time sensor data based on a sensor network; wherein the real-time sensor data includes millimeter-wave radar population density data, CO concentration gradient data, and building structure risk coefficients; acquiring historical evacuation data based on a building information model; wherein the historical evacuation data includes turbulence model parameters from evacuation drills; and transmitting the real-time sensor data to an edge computing node via a cognitive packet network.

[0069] Specifically, the sensor (sensor network) deployment requirements are as follows: millimeter-wave radar installation height ≥ 6m (preferably signal poles), horizontal field of view ≥ 75°, and detection density error ≤ 2 people / m. 2 The CO (carbon monoxide) sensor can be deployed at key nodes in the fire source diffusion path (such as ventilation openings and corridor corners), with a sampling period of ≤1 second. After installation, it can collect data such as crowd turbulence index, CO concentration gradient, and thermal imaging fire source location coordinates.

[0070] For example, such as Figure 6 As shown, the input data layer collects data such as crowd density, movement data, and turbulence index through millimeter-wave radar in the deployed sensor network; it collects data such as concentration gradient and fire source diffusion direction through CO sensors; and it obtains building structure risk coefficients (e.g., exit width, load-bearing wall location) through the Building Information Modeling (BIM) system. Preset parameters include: neural activation threshold (e.g., BOLD signal Δ≥1.8%) and energy consumption constraints (battery capacity, light intensity-power consumption mapping table, etc.).

[0071] In one embodiment, the optimal evacuation path is obtained by processing the real-time sensor data and historical evacuation data based on edge computing nodes, including: performing Kalman filtering on the real-time sensor data based on edge computing nodes, and generating a real-time dynamic environment map by fusing the filtered real-time sensor data; and inputting the real-time dynamic environment map and historical evacuation data into a neural symbolic model to obtain the optimal evacuation path.

[0072] In one embodiment, inputting the real-time dynamic environment map and historical evacuation data into a neural symbolic model to obtain the optimal evacuation path includes: inputting the real-time dynamic environment map and historical evacuation data into a symbolic logic layer to obtain a shortest path candidate set; evaluating the path safety coefficient of the shortest path candidate set based on a DQN neural network to obtain the optimal evacuation path; wherein, the optimal evacuation path includes dynamic arrow directions and voice commands.

[0073] Understandably, edge computing nodes are typically deployed in the electrical shafts or fire control rooms on each floor of a building. This embodiment uses Kalman filtering to eliminate sensor noise, inputs real-time data into a deep reinforcement learning model, outputs dynamic evacuation paths, and combines a DNN model to calculate optical parameter adjustment strategies, balancing neural activation and system power consumption.

[0074] Specifically, intelligent decision-making algorithms - path optimization:

[0075] ① Data Input. Real-time data: such as millimeter-wave radar crowd density, CO concentration gradient, and building structural risk coefficient (from the BIM system); historical data: such as turbulence model parameters from evacuation drills.

[0076] ② Model processing. Neural symbolic model: Combining symbolic logic (e.g., shortest path algorithm) with deep Q-network (DQN), the reward function can be designed as follows:

[0077]

[0078] in, , These are the weighting coefficients.

[0079] Experience replay: Store the quadruple (current state s) t Action a t Reward r t The next state s t+1 Batch sampling training.

[0080] ③ Path display. Dynamic LED arrows: Indicate directional changes through light intensity gradients (e.g., 200→50 lux); Voice prompts: 85dB@1 meter, including directional instructions and safety distance reminders. Here, 85dB@1 meter is an exemplary sound level indicator used to describe the intensity of sound under specific conditions. "85dB" refers to the loudness level of the sound, measured in decibels (dB), and "@1 meter" indicates that this loudness was measured at a distance of 1 meter from the sound source.

[0081] For example, such as Figure 6As shown, the data processing layer performs Kalman filtering on the input data through edge computing nodes to eliminate noise, followed by data fusion to generate a dynamic environment map (including crowd heatmaps, fire source locations, etc.). A deep reinforcement learning model (e.g., a neural symbolic model) processes the input dynamic environment map and historical evacuation data to output the optimal evacuation path (including dynamic arrow directions and voice commands). The data processing of the neural symbolic model includes a symbolic logic layer (compiling the shortest path candidate set) and a DQN network (evaluating path safety factors).

[0082] In one example, an environmental model is built based on dynamic environmental maps such as crowd heatmaps, exit locations, and fire source locations. A state space is defined, including the current location of the crowd, exit locations, crowd density, and turbulence indices from historical evacuation data. An action space, i.e., possible evacuation paths, is defined, where each action represents a possible evacuation path, and the agent can choose one of these paths to evacuate. A reward function is designed to evaluate the quality of the evacuation paths. A deep reinforcement learning model (e.g., a neural symbolic model) is trained using historical evacuation data and real-time environmental maps. During training, the neural symbolic model continuously learns how to select the optimal evacuation path based on the current state, while also generalizing using historical data to improve its ability to generate optimal paths under different conditions. In a real-time environment, the trained neural symbolic model can use real-time environmental maps and crowd density data to predict the current turbulence indices and generate the optimal evacuation path based on this information. It can update its state at each time step and select the next best action until the evacuation is complete. This embodiment, by combining deep reinforcement learning, real-time environmental maps, and historical evacuation data, enables this neural symbolic model to generate more efficient and safer evacuation paths, providing strong support for crowd evacuation.

[0083] In one embodiment, obtaining a light parameter adjustment instruction based on the real-time sensor data, neural activation threshold, and energy consumption constraint includes: acquiring ambient illuminance and remaining battery capacity based on the real-time sensor data; and inputting the ambient illuminance, remaining battery capacity, neural activation threshold, and energy consumption constraint into a DNN prediction model to obtain the light parameter adjustment instruction.

[0084] In one embodiment, the ambient illuminance, remaining battery capacity, neural activation threshold, and energy consumption constraint are input into a DNN prediction model to obtain light parameter adjustment instructions. This includes: normalizing the ambient illuminance, remaining battery capacity, and neural activation threshold to obtain normalized data; inputting the normalized data into a multilayer fully connected network of the DNN prediction model and adjusting it based on the energy consumption constraint to output frequency, duty cycle, and light intensity; and generating light parameter adjustment instructions based on the frequency, duty cycle, and light intensity.

[0085] Specifically, the intelligent decision-making algorithm - adaptive light parameters:

[0086] ① Input parameters. Real-time data: such as population density, CO concentration, ambient illuminance, and remaining battery capacity; physiological parameters: such as preset neural activation thresholds (e.g., V1 region BOLD signal Δ≥1.8%).

[0087] ② Processing flow. Data preprocessing: Normalize to the [0, 1] interval to eliminate dimensional differences; Model architecture: 5-layer fully connected network (input layer → 3 hidden layers → output layer), activation function ReLU, loss function MAE; Output parameters: Light intensity: 50-200 lux (linear mapping); Frequency: 5-10Hz (Sigmoid function output); Adjustment basis: Energy consumption constraint: priority order is safety > neural activation > battery life; Adaptive learning rate: Adam optimizer can be used, initial lr (initial learning rate) = 0.001.

[0088] For example, such as Figure 6 As shown, the data processing layer processes the input ambient illuminance, battery capacity (remaining battery capacity), and neural activation threshold based on the DNN prediction model, outputting light parameter adjustment instructions (including frequency, duty cycle, and light intensity). The data processing of the DNN prediction model includes a fully connected network (predicting light intensity of 50-200 lux and frequency of 5-10Hz) and adaptive optimization (adjusting weights using the Adam algorithm).

[0089] In one embodiment, an LED driving signal and a voice synthesis command are generated based on the optimal evacuation path and the light parameter adjustment command to drive the light source module of the optical guide sign of the evacuation channel and provide evacuation voice prompts.

[0090] Specifically, the edge computing node executes a deep reinforcement learning algorithm to generate evacuation path parameters (including the dynamic arrow direction of the optimal evacuation path and voice commands), and sends the control commands of the evacuation path parameters to the FPGA control module via a high-speed bus (such as PCIe); the FPGA control module dynamically adjusts the pulse frequency (5-10Hz), duty cycle (10%-90%) and light intensity (50-200 lux) of the LED array according to the control commands.

[0091] For example, such as Figure 6As shown, this is the control execution layer. The FPGA control module processes the input optical parameter commands and path direction, outputting LED drive signals (PWM waveforms) and speech synthesis commands. The FPGA control module's processing includes generating the pulse base frequency (5-10Hz) using a phase-locked loop (PLL) and dynamically adjusting the duty cycle (10%-90% adjustment achieved by a 32-bit counter). The FPGA control module can employ a dual-closed-loop control strategy. Based on the PLL module, with an input clock of 50MHz, the frequency is dynamically divided by a factor N (N=2). 32 ×T sysclk / T pwm T sysclk T is the time of the system clock cycle. pwm A 5-10Hz base frequency is generated for the period of the PWM pulse width modulation signal. A 32-bit counter is used; when the count value is less than the threshold duty, a low level is output, and vice versa. The dynamic adjustment formula duty=2 is used. 32 The duty cycle is dynamically adjusted by multiplying the input voltage by (1-DC) (where DC is the duty cycle: 0.1≤DC≤0.9). The PLL parameters are dynamically reconfigured via the APB interface, with an adjustment delay of <50ms, supporting Burst Mode to handle emergency scenarios.

[0092] It should be noted that the following is an explanation based on specific examples in practical applications. (1) Installation layout. The installation specifications of the optical guidance signs proposed in this embodiment can be as follows: horizontal spacing, for example, one set every 10 meters along the evacuation route, and densification to 5 meters at corners; installation height, for example, 2.2 meters (error ±0.1m), tilted 15° downwards to avoid obstruction; light field coverage, for example, by adjusting the beam angle to 120° through microlenses to ensure ground illuminance ≥1 lux. For special scenarios, such as high-rise buildings, two-way indicator signs can be added to each platform in the stairwell, with a height of 2.5 meters; for example, in smoky environments, multimodal feedback (voice + vibration) can be activated, and the brightness of the sign can be automatically increased to 200 lux. (2) Light parameter adjustment: the enhancement effect of the 8Hz pulse on the BOLD signal in the V1 region (Δ signal strength ≥1.8%) is verified by fMRI, and the resonance effect of the theta wave (4-8Hz) and light frequency under stress is confirmed by EEG. (3) Multimodal feedback: Integrating voice prompts (e.g., 85dB@1 meter) and tactile vibration (e.g., 50Hz frequency) to compensate for visual limitations in smoky environments. The tactile vibration technology works as follows: the actuator type can be a linear resonant actuator (LRA), with a frequency of 50Hz and an amplitude of 0.8G. The drive method can be PWM control, using an H-bridge circuit to achieve forward and reverse vibration. Deployment method: The location can be embedded in the evacuation passage handrail (spaced 2 meters apart) or below the ground guide strip. Trigger logic: When the smoke concentration is >500ppm, synchronous vibration is initiated (delay <100ms). In addition, flexible vibration technology can be used: using an array of micro motors (spaced 30cm apart) to achieve a positional perception error of <10cm. Frequency adaptation, such as dynamically adjusting 50-100Hz according to environmental noise, ensures tactile recognizability.

[0093] In this embodiment, a mapping relationship between light parameters and neural responses is established: an activation threshold model is established for the amygdala (fear response) and prefrontal cortex (decision-making) based on a combination of 450nm blue light and 8Hz pulses. Non-invasive light field modulation is performed: a wide-area uniform illumination system is designed to overcome energy attenuation caused by smoke scattering, ensuring a light intensity ≥100 lux (the minimum threshold for activating melanopsin). Multimodal perception fusion is achieved: LiDAR crowd counting, thermal imaging fire prediction, and EEG stress monitoring are integrated to dynamically adjust light parameters and evacuation routes. Light safety and compliance are met: glare control conforms to design specifications such as DB11 / T 2239-2024 (average brightness ≤1000 cd / m²). 2This system conforms to electrical safety standards. Experiments show that, compared to traditional systems, the emergency evacuation system employing the optical guidance signage and method described in this embodiment significantly improves evacuation efficiency, reducing evacuation time by 28% (e.g., in a scenario with 200 people) and path deviation rate by 42%. For neural response verification, fMRI data shows that blue light pulses increase left thalamic activity by 30% and inhibit right amygdala activity by 15%, indicating a bidirectional optimization of crisis cognition and emotion regulation by the optical guidance signage in this embodiment. Furthermore, this emergency evacuation system is compatible and scalable, supporting integration with BIM systems for pre-assessment of building structural risks, and reserving a 5G-MEC interface for large-scale city-level evacuation networks.

[0094] This embodiment provides an intelligent decision-making method for emergency evacuation in underground rail spaces, including: acquiring real-time sensor data and historical evacuation data, setting neural activation thresholds and energy consumption constraints; processing the real-time sensor data and historical evacuation data based on edge computing nodes to obtain the optimal evacuation path; obtaining optical parameter adjustment instructions based on the real-time sensor data, neural activation thresholds, and energy consumption constraints; and generating LED driving signals and voice synthesis instructions based on the optimal evacuation path and optical parameter adjustment instructions to drive the light source modules of the optical guidance signs in the evacuation channel and provide evacuation voice prompts. In this embodiment, multimodal perception fusion is used to dynamically adjust optical parameters and evacuation paths, thereby improving evacuation efficiency, reducing evacuation time, and lowering path deviation rates.

[0095] Furthermore, this embodiment of the invention also proposes a storage medium storing an intelligent decision-making program for emergency evacuation in underground rail spaces. When the intelligent decision-making program for emergency evacuation in underground rail spaces is executed by a processor, it implements the steps of the intelligent decision-making method for emergency evacuation in underground rail spaces as described above.

[0096] Reference Figure 7 , Figure 7 This is a structural block diagram of an embodiment of the intelligent decision-making system for emergency evacuation in underground rail space according to the present invention.

[0097] like Figure 7 As shown, the intelligent decision-making system for emergency evacuation in underground rail space includes:

[0098] Data acquisition module 10 is used to acquire real-time sensor data and historical evacuation data, and to set neural activation thresholds and energy consumption constraints;

[0099] The path calculation module 20 is used to process the real-time sensor data and historical evacuation data based on the edge computing node to obtain the optimal evacuation path;

[0100] The instruction calculation module 30 is used to obtain optical parameter adjustment instructions based on the real-time sensor data, neural activation threshold and energy consumption constraints;

[0101] The control execution module 40 is used to generate LED driving signals and voice synthesis commands according to the optimal evacuation path and light parameter adjustment instructions, so as to drive the light source module of the optical guide sign of the evacuation channel and provide evacuation voice prompts.

[0102] This embodiment provides an intelligent decision-making system for emergency evacuation in underground rail spaces. Through multimodal perception fusion, it dynamically adjusts optical parameters and evacuation routes, thereby improving evacuation efficiency, reducing evacuation time, and lowering path deviation rates. Compared with traditional systems, the system in this embodiment can significantly improve evacuation efficiency, reducing evacuation time by 28% (e.g., in a scenario with 200 people) and path deviation rate by 42%. For neural response verification, fMRI data shows that blue light pulses increase left thalamic activity by 30% and inhibit right amygdala activity by 15%, indicating that the optical guidance markers used in this embodiment provide bidirectional optimization in crisis cognition and emotion regulation. Furthermore, the system described in this embodiment has compatibility and scalability, supports integration with BIM systems for pre-assessment of building structural risks, and reserves a 5G-MEC interface for large-scale city-level evacuation networks.

[0103] It should be noted that technical details not described in detail in this embodiment of the intelligent decision-making system for emergency evacuation in underground rail space can be found in any embodiment of the present invention and applied to the intelligent decision-making method for emergency evacuation in underground rail space as described above, and will not be repeated here.

[0104] It should be understood that the above are merely illustrative examples and do not constitute any limitation on the technical solutions of the present invention. In specific applications, those skilled in the art can make settings as needed, and the present invention does not impose any restrictions on this.

[0105] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of this invention. In practical applications, those skilled in the art can select some or all of the workflow to achieve the purpose of this embodiment according to actual needs, and no restrictions are imposed here.

[0106] Furthermore, it should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0107] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0108] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory (ROM) / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0109] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A smart decision-making method for emergency evacuation in underground rail spaces, characterized in that, include: Acquire real-time sensor data and historical evacuation data, and set neural activation thresholds and energy consumption constraints; The optimal evacuation path is obtained by processing the real-time sensor data and historical evacuation data using edge computing nodes. The optical parameter adjustment command is obtained based on the real-time sensor data, neural activation threshold, and energy consumption constraint. Based on the optimal evacuation path and light parameter adjustment instructions, LED driving signals and voice synthesis instructions are generated to drive the light source module of the optical guide sign of the evacuation channel and provide evacuation voice prompts.

2. The method as described in claim 1, characterized in that, The wavelength range of the optical guide mark is 450±5nm, and the pulse frequency range is 8±0.5Hz; The light source module includes a high-density LED array, and the light source module has a built-in optical diffusion film.

3. The method as described in claim 1, characterized in that, The acquisition of real-time sensor data and historical evacuation data includes: Real-time sensor data is acquired based on sensor networks; wherein, the real-time sensor data includes millimeter-wave radar population density data, CO concentration gradient data, and building structure risk coefficient; Historical evacuation data is obtained based on building information modeling; wherein, the historical evacuation data includes turbulence model parameters from evacuation drills; The real-time sensor data is transmitted to the edge computing node via a cognitive packet network.

4. The method as described in claim 1, characterized in that, The process of processing the real-time sensor data and historical evacuation data based on edge computing nodes to obtain the optimal evacuation path includes: Kalman filtering is performed on the real-time sensor data based on edge computing nodes, and a real-time dynamic environment map is generated by fusing the filtered real-time sensor data. The real-time dynamic environment map and historical evacuation data are input into the neural symbolic model to obtain the optimal evacuation route.

5. The method as described in claim 4, characterized in that, The step of inputting the real-time dynamic environment map and historical evacuation data into the neural symbolic model to obtain the optimal evacuation path includes: The real-time dynamic environment map and historical evacuation data are input into the symbolic logic layer to obtain the shortest path candidate set; The path safety coefficient of the shortest path candidate set is evaluated based on the DQN neural network to obtain the optimal evacuation path; wherein, the optimal evacuation path includes dynamic arrow direction and voice command.

6. The method as described in claim 1, characterized in that, The step of obtaining the optical parameter adjustment command based on the real-time sensor data, neural activation threshold, and energy consumption constraints includes: The ambient illuminance and remaining battery capacity are obtained based on the real-time sensor data. The ambient illuminance, remaining battery capacity, neural activation threshold, and energy consumption constraints are input into the DNN prediction model to obtain light parameter adjustment instructions.

7. The method as described in claim 6, characterized in that, The step of inputting the ambient illuminance, remaining battery capacity, neural activation threshold, and energy consumption constraints into the DNN prediction model to obtain light parameter adjustment instructions includes: The ambient illuminance, remaining battery capacity, and neural activation threshold are normalized to obtain normalized data. The normalized data is input into the multilayer fully connected network of the DNN prediction model and adjusted based on the energy consumption constraint to output the frequency, duty cycle and light intensity. The light parameter adjustment command is generated based on the frequency, duty cycle, and light intensity.

8. An intelligent decision-making system for emergency evacuation in underground rail spaces, characterized in that, include: The data acquisition module is used to acquire real-time sensor data and historical evacuation data, and to set neural activation thresholds and energy consumption constraints. The path calculation module is used to process the real-time sensor data and historical evacuation data based on edge computing nodes to obtain the optimal evacuation path. The instruction calculation module is used to obtain optical parameter adjustment instructions based on the real-time sensor data, neural activation threshold, and energy consumption constraints. The control execution module is used to generate LED driving signals and voice synthesis commands based on the optimal evacuation path and light parameter adjustment instructions, so as to drive the light source module of the optical guidance sign of the evacuation channel and provide evacuation voice prompts.

9. An electronic device, characterized in that, The electronic device includes: a memory, a processor, and an intelligent decision-making program for emergency evacuation of underground orbital space stored in the memory and executable on the processor, wherein the intelligent decision-making program for emergency evacuation of underground orbital space is configured to implement the intelligent decision-making method for emergency evacuation of underground orbital space as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores an intelligent decision-making program for emergency evacuation in underground rail spaces, which is used to enable the processor to implement the intelligent decision-making method for emergency evacuation in underground rail spaces as described in any one of claims 1 to 7.

Images