Emergency Rescue Path Planning for Urban Emergencies Based on Improved GA and PSO

Abstract

With the acceleration of urbanization and the frequent urban emergencies, traditional rescue path planning methods have response delay and low path efficiency. To improve the efficiency of urban emergency rescue and resource scheduling capabilities, a path planning technique for urban emergency rescue based on improved Genetic Algorithms (GA) and Particle Swarm Optimization (PSO) is built. The research introduces Tent chaotic mapping initialization particles, adaptive segmented inertia weights and exponential learning factors for the PSO algorithm to enhance global search capabilities, and adds traction acceleration terms to avoid local optimality. For the GA, a combination strategy of elite retention and roulette is used to improve selection efficiency, and the Metropolis criterion is combined to optimize the cross-mutation operation, and an adaptive variable neighborhood search mechanism is introduced to strengthen local search. In the experimental setting, the Python simulation platform is used to compare with baseline methods such as RRT*, D* Lite and MOPSO. The test indicators include response time, planning success rate, path length, number of convergence iterations, etc. Experimental results showed that under six concurrent events, the response time of the research method was 6 seconds, which was significantly better than that of the comparison method. When the dynamic obstacle density was 40/km², the planning success rate reached 90.2%. When the scene complexity was 200 nodes, the single planning calculation time was 150ms. The research method converged at the 100th iteration, and the fitness change rate was reduced to 1.3%, showing faster convergence speed and better stability. The above results show that the proposed method is superior to traditional methods in terms of timeliness, robustness and optimization capabilities, which is suitable for emergency rescue path planning in complex urban scenarios.