Bipedal trajectory control based on neurofuzzy networks

This paper presents a bipedal trajectory control technique based on a neurofuzzy controller and a neural network emulator. The neurofuzzy controller is a five-layered neurofuzzy network, it provides the control signals in each stage of a walking gait. The neural network emulator is a conventional three-layered feedforward neural network. It emulates the robotic dynamics and provides the error signals which can be used to back propagate through the controller in each stage. This technique can generate dynamic walking gaits along a pre-specified reference trajectory on sloping terrain.