An Adaptive Controller System using mnSOM

Introduction: Autonomous underwater vehicles (AUVs) have great advantages for activities in deep oceans, and are expected as the attractive tool. However, AUVs have various problems which should be solved for motion control, acquisition of sensors’ information, determination of action, navigation without collision, self-localization and so on. We have been investigating the application of neural network technology into the AUVs focusing on the capability of neural networks (NNs) such as learning, nonlinear mapping. In order to realize the useful and practical robots which can work in the ocean, underwater vehicles should take their action by judging the changing condition from their own sensors and actuators, and are desirable to make their behaviors with limited efforts of the operators, because of features caused by the working environment. Therefore, the AUVs should be autonomous and adaptive to their environment. Considering the above features, we proposed an adaptive control system [1] and a collision avoidance system [2] using Self-Organizing Map (SOM) [3]. AUVs have non-liner dynamics in six degrees of freedom, and the changes of the equipments of robots have influence on the control system. In the adaptive control method [1], the initial state of controller is getting less and less during the process of adaptation. Therefore, a method which keeps the information of previous environment and adapts to new environment is needed. This paper describes an adaptive control method using modular network Self-Organizing Map called mnSOM [5] proposed by K.Tokunaga et. al, and its application to an AUV.