Air-ground collaborative inspection path planning for heterogeneous tasks

Abstract

Complex infrastructure environments impose distinct mobility constraints on inspection robots, rendering single-type systems inefficient. To address the challenge where aerial vehicles struggle indoors and ground vehicles cannot reach high altitudes, this study investigates an air-ground collaborative path planning problem characterized by strict task heterogeneity. A mathematical model is constructed to minimize the system’s maximum inspection time by strictly matching task types with specific robotic capabilities. To solve this complex combinatorial optimization problem efficiently, a hybrid evolutionary algorithm is proposed that integrates the local search advantage of simulated annealing into the gray wolf optimizer framework. Comparative experiments demonstrate that this hybrid approach outperforms existing simulated annealing genetic algorithms, bi-trajectory search methods, and standard gray wolf optimizers in terms of solution accuracy, convergence speed, and stability. The findings confirm that the proposed strategy effectively balances workload among heterogeneous robots, offering a robust solution for intelligent automated inspection.