Robot Path Planning Based on Simulated Annealing and Artificial Neural Networks

Abstract

As for the limitations of algorithms in global path planning of mobile robot at present, this study applies the improved simulated annealing algorithm artificial neural networks to path planning of mobile robot in order to better the weaknesses of great scale of iteration computation and slow convergence, since the best-reserved simulated annealing algorithm was introduced and it was effectively combined with other algorithms, this improved algorithm has accelerated the convergence and shortened the computing time in the path planning and the global optimal solution can be quickly obtained. Because the simulated annealing algorithm was updated and the obstacle collision penalty function represented by neural networks and the path length are treated as the energy function, not only does the planning of path meet the standards of shortest path, but also avoids collisions with obstacles. Experimental results of simulation show this improved algorithm can effectively improve the calculation speed of path planning and ensure the quality of path planning.

Application of Type-1 and Type-2 Fuzzy Logic Controller for The Real Swarm Robot

“…Basically, besides fuzzy logic, there are a lot of approaches that are used as intelligent computation methods, for instances: ant colony optimization algorithm [16], artificial neural network, neuro-fuzzy [17], particle swarm optimization [14], genetic algorithm [18] [19], simulated annealing algorithm [20] [21].…”

Robotic Path Planning using Genetic Algorithm in Dynamic Environment

“…Considering the work proposed by previous researchers. In [5] [7] improved simulated annealing combined with artificial neural network for finding a solution for path planning problem. In [6] ACO is applied in dynamic environment using grid environment.…”