A novel hybrid Coyote-Particle Swarm Optimization Algorithm for three-dimensional constrained trajectory planning of Unmanned Aerial Vehicle

Unmanned Aerial Vehicles (UAVs) have been considered the future of transportation systems. However, mostly the reported optimizers struggle to estimate the flyable trajectories within acceptable accuracy and time bound under various constraints, particularly in a complex 3D environment. Therefore, this work proposes a novel hybrid optimizer (HCPSOA) by combining the Particle Swarm Optimization (PSO) and Coyote Optimization Algorithm (COA). Further, the chaotic logistic map and dynamic weight adjustments have been incorporated to enhance the exploration-exploitation capabilities. After validating the promising performance of HCPSOA against popular metaheuristics (COA, PSO, Improved COA, Glowworm Swarm Optimization, and Hybrid Fireworks PSO) for several benchmark functions, the proposed HCPSOA has been employed to estimate the flyable path by formulating a novel cost function and smoothened by cubic B-spline curve. The simulated results reveal that the HCPSOA provide a non -colliding path with up to 10.00% lesser average cost for the considered real world scenario (map 3), which confirms its supremacy for the estimation of the safe and flyable relative to other compared algorithms.& COPY; 2023 Elsevier B.V. All rights reserved.

Automated summary

This study proposes a novel hybrid optimizer (HCPSOA) for unmanned aerial vehicles (UAVs) by combining Particle Swarm Optimization (PSO) and Coyote Optimization Algorithm (COA). The chaotic logistic map and dynamic weight adjustments are incorporated to enhance exploration-exploitation capabilities. The results show that HCPSOA outperforms other algorithms in accurately estimating flyable paths in complex environments.