Autonomous parallel parking of a car-like mobile robot by a neuro-fuzzy sensor-based controller

In this paper, a neuro-fuzzy model has been developed for autonomous parallel parking of a car-like mobile robot. In our approach we have focused on the most difficult case of parallel parking which is the case when the parking space dimensions cannot be identified. The proposed model uses the data from three sonar sensors mounted in the front left corner of the car to decide on the turning angle. Fifth-order polynomial reference paths for three different size parking dimensions have been used to generate the training data. The fuzzy model has been identified by subtractive clustering algorithm and trained by ANFIS. The simulation results show that the model can successfully decide about the motion direction at each sampling time without knowing the parking space width, based on the direct sonar readings which serve as inputs. The results which are based on real dimensions of a typical car demonstrate the feasibility and effectiveness of the proposed controller in parallel parking.