Using Neuro-Evolutionary-Fuzzy Method to Control a Swarm of Unmanned Underwater Vehicles

Piotr Szymak

Abstract


The paper presents the research whose the main goal was to build a control system for a swarm of Unmanned Underwater Vehicles UUVs for predator-prey problem. To control a swarm of the vehicles new Fuzzy System with Neural Aggregation of the fuzzy rules FSNA was proposed. To learn the FSNA innovative Cooperative Co-evolutionary Genetic Algorithm with Indirect Neural Encoding CCGA-INE was used. At the beginning of the paper, the introduction to the subject of the paper is included. Next, the principles of operation of new FSNA and its tuning method CCGA-INE are presented. In the end, the results of numerical research of FSNA controlling a swarm of the underwater vehicles in a predator-prey problem are presented.

Keywords


neuro-fuzzy system, co-evolution, indirect encoding, control of unmanned vehicles swarm

Full Text: PDF