An improved real-time path planning of mobile robot in a complex and dynamic environment

Many traditional path planning of mobile robot are unable to adapt to the complex and dynamic environment. This paper proposes an improved real-time path planning algorithm of mobile robot based on cellular neural network by idea of artificial potential field. The target neuron has the maximal positive active value which is damply spread to the whole state space by local lateral connections of neurons in the path planning. The mobile robot was attracted to the target through the neural activity propagation, while the obstacles put away the mobile robot to avoid collision by making themselves keep the lowest active value. The experimental results indicate that the path planning algorithm was continuous, optimal, and the mobile robot could respond quickly to the complex and fast changing environment.