Fuzzy logic system promises an efficient way for obstacle avoidance. However, it is difficult to maintain the correctness, consistency, and completeness of a fuzzy rule base constructed and tuned by a human expert. Reinforcement learning method is capable of learning the fuzzy rules automatically. However, it incurs heavy learning phase and may result in an insufficiently learned rule base due to the curse of dimensionality. In this paper, we propose a neural fuzzy system with mixed coarse learning and fine learning phases. In the first phase, supervised learning method is used to determine the membership functions for the input and output variables simultaneously. After sufficient training, fine learning is applied which employs reinforcement learning algorithm to fine-tune the membership functions for the output variables. For sufficient learning, a new learning method using modified Sutton and Barto's model is proposed to strengthen the exploration. Through this two-step tuning approach, the mobile robot is able to perform collision-free navigation. To deal with the difficulty in acquiring large amount of training data with high consistency for the supervised learning, we develop a virtual environment (VE) simulator, which is able to provide desktop virtual environment (DVE) and immersive virtual environment (IVE) visualization. Through operating a mobile robot in the virtual environment (DVE/IVE) by a skilled human operator, the training data are readily obtained and used to train the neural fuzzy system.
展开▼
