Robot trajectory optimization based on bio-inspired algorithms

Tema ovog diplomskog rada je izrada modula za planiranje kretanja koji se temelji na biološki inspiriranim algoritmima. Planirane trajektorije smještene su u 2D i 3D prostoru. Prostor je okoliš koji je korisnik definirao ovisno o željenom broju i položaju prepreka. Provode se različite metode kako bi se ispitao njihov utjecaj na performanse i na razinu konvergencije algoritma. Pokazano je da svaka verzija algoritma ima svoje prednosti i nedostatke. Za svaku verziju algoritma provodi se temeljita statistička analiza, a rezultati su dokumentirani i prezentirani u radu.The subject matter of this master's thesis is a design of a motion planning module based on a biologically inspired algorithm. Planned trajectories are placed in both 2D and 3D space. Space is a user-defined environment which contains an arbitrary number of placed obstacles. Different methods are implemented to examine their impact on the performance and the level of convergence of the algorithm. It is shown that each instance of the algorithm has its benefits and shortcomings. A meticulous statistical analysis is performed for each version of the algorithm, and the results are documented and presented in the thesis