Abstract:In order to predict the fire heat release rate (HRR) and tunnel temperature field during subway tunnel fires more accurately, a novel approach based on an improved genetic algorithm-optimized back propagation (GA-BP) neural network is proposed. This work incorporates a regularization term into the fitness function of the genetic algorithm, which enhances the GA-BP algorithm. The algorithm integrates the strengths of genetic algorithms and neural networks, utilizing the efficient optimization capability of genetic algorithms and the powerful nonlinear fitting ability of BP neural networks to establish mapping relationships among multiple parameters in fire scenarios. The improved algorithm shows a 66.2% reduction in average error compared to traditional neural networks, indicating faster convergence and higher accuracy. This method is unaffected by tunnel environment variations and complex fire propagation mechanisms, enabling the precise prediction of fire HRRs and the entire tunnel temperature distribution with only a subset of temperature sensors. Experimental results indicate that the algorithm can accurately estimate the fire HRR using partial sensors, with an average error of 0.11 MW. Furthermore, even when certain temperature sensor values are unstable, the algorithm predicts an average error of 2.77 °C for the entire tunnel temperature field, demonstrating the network's strong generalization capability and robustness. Comparative experiments show that the improved GA-BP algorithm outperforms random forest, Transformer, long short-term memory network, and back propagation neural network in predicting fire HRRs. This research contributes a reliable decision-making foundation for intelligent firefighting systems.

Study on temperature prediction of subway tunnel fire based on improved GA-BP algorithm

Self-adaptive GA for Optimization of Neural Network Model of Pipe Network for State Estimation

A multi-GA-BPNN fusion algorithm and full-scale experimental verification for fire warning in the underground pipe gallery

An Improved Updatable Backpropagation Neural Network for Temperature Prognosis in Tunnel Fires

Predicting Coal Ash Fusion Temperature Using Hybrid of Ant Colony Algorithm and BP Neural Network

BP Neural Network-Based Adaptive Spatial-Temporal Data Generation Technology for Predicting Ceiling Temperature in Tunnel Fire and Full-Scale Experimental Verification

A Multiple Back Propagation Neural Network Fusion Algorithm for Ceiling Temperature Prediction in Tunnel Fires

Regression Prediction of Critical Exhaust Volumetric Flow Rate in Tunnel Fire with Two-Point Extraction Ventilation based on Neural Network Method

Study on Real-Time Heat Release Rate Inversion for Dynamic Reconstruction and Visualization of Tunnel Fire Scenarios

Prediction of Long-Term Settlement in Shield Tunnel Using GA-BP Neural Network

Temperature Prediction Method Based on Improved GA_BP Neural Network

A Deep Learning–based Surrogate Model for Spatial-Temporal Temperature Field Prediction in Subway Tunnel Fires Via CFD Simulation

Short-term forecasting of traffic flow based on genetic algorithm and bp neural network

Automatic Data Generation Method for Precise Ceiling Temperature Prediction of Cables Fire in the Utility Tunnel and Full-Scale Experimental Verification

A Multi-Neural Network Fusion Algorithm for Fire Warning in Tunnels.

STUDY ON BACK-PROPAGATION NEURAL NETWORK MODELING OF PREDICTING THE GASIFICATION RATE UNDER THE ELEVATED TEMPERATURES

Optimized BP Network Model for Predicting Igniting Temperature of Pulverized-Coal Based on Genetic Algorithm

Neural Network Approach for RHR Calculation and Prediction in Fire Science

A Physical Model-Free Ant Colony Optimization Network Algorithm and Full Scale Experimental Investigation on Ceiling Temperature Distribution in the Utility Tunnel Fire

Temperature Data-Driven Fire Source Estimation Algorithm of the Underground Pipe Gallery

Optimization of Backpropagation Neural Network under the Adaptive Genetic Algorithm