Parallel Hybrid Metaheuristic on Shared Memory System for Real-Time Uav Path Planning

Abstract

In this paper, we present a parallel hybrid metaheuristic that combines the strengths of the particle swarm optimization (PSO) and the genetic algorithm (GA) to produce an improved path-planner algorithm for fixed wing unmanned aerial vehicles (UAVs). The proposed solution uses a multi-objective cost function we developed and generates in real-time feasible and quasi-optimal trajectories in complex 3D environments. Our parallel hybrid algorithm simulates multiple GA populations and PSO swarms in parallel while allowing migration of solutions. This collaboration between the GA and the PSO leads to an algorithm that exhibits the strengths of both optimization methods and produces superior solutions. Moreover, by using the "single-program, multiple-data" parallel programming paradigm, we maximize the use of today's multicore CPU and significantly reduce the execution time of the parallel program compared to a sequential implementation. We observed a quasi-linear speedup of 10.7 times faster on a 12-core shared memory system resulting in an execution time of 5 s which allows in-flight planning. Finally, we show with statistical significance that our parallel hybrid algorithm produces superior trajectories to the parallel GA or the parallel PSO we previously developed.

Recent advances in unmanned aerial vehicles real-time trajectory planning

“…28). Only 28% of the surveyed papers (Zheng et al 2003;Cocaud 2006;Allaire et al 2009;Guo et al 2009;Swartzentruber et al 2010;Chen et al 2011;Wan et al 2011;Holub et al 2012;Yan et al 2012;Ozalp and Sahingoz 2013;Roberge et al 2013Roberge et al , 2014Q. Wang et al 2014;Zhan et al 2014;Gardi et al 2015a;Ling and Hao 2015;Wen et al 2015) considered both 3D terrain and obstacle avoidance; these are the ones for which we will further analyze the flyability.…”

“…1. Roberge et al (2013Roberge et al ( , 2014 presented a UAV trajectory optimization method that used a cost function, which employed the 3D arcs of circles and helical curves trajectory smoothing technique for which Labonté (2012Labonté ( , 2015Labonté ( , 2016Labonté ( , 2017aLabonté ( , 2017b provided the equations for computing both the power and the fuel requirements of the UAV trajectory. These works will be taken as a reference point for the rest of our analysis.…”

“…Roberge et al (2013) implemented both a GA and a PSO and demonstrated that its GA implementation provides better trajectories than the PSO implementation. Roberge et al (2014) combined a GA with a PSO to improve the quality of the generated trajectory. Other than Roberge et al (2013Roberge et al ( , 2014, no one used a cost function that incorporates both the power and the fuel criteria.…”

“…Roberge et al (2014) combined a GA with a PSO to improve the quality of the generated trajectory. Other than Roberge et al (2013Roberge et al ( , 2014, no one used a cost function that incorporates both the power and the fuel criteria.…”

“…Roberge et al (2013) set the deadline at 10 s, explaining that trajectory generated in their testing scenarios had a length between 5 and 60 km, and that within 10 s the fastest UAV (within their comparison set) was traveling 1.7 km, which was deemed acceptable. Roberge et al (2014) reduced the deadline to 5 s for a distance of 882 m. In Liu et al (2013), the minimum flight time for a trajectory segment was 5 s. In their simulation the UAV was flying at 50 m/s, therefore, their shortest trajectory segment was 250 m long. Ronfle-Nadaud (2009) set her deadline to ensure 150 m was given to two UAV facing each other to change their respective headings, this being the same distance as in the near-midair collision regulation (Federal Aviation Administration 2014).…”