The biped robot with heterogeneous legs(BRHL) is the test-bed of a novel humanoid robot,which consists of an artificial leg and an intelligent bionic leg, the artificial leg is used to simulate the amputee's healthy leg and the other works as the intelligent artificial limb. The gait identification system in intelligent bionic leg based on the process neural networks is studied. The system depends on the six-dimension force sensor to detect the ground's counterforce on diffierent terrain, identifies the current terrain by the process neural networks, controls the output force of the damper according to the gait information and realizes the bionic leg following with the artificial leg. The model of the process neural networks is presented, and the related arithmetics are calculated. The experiment results show the rationality and feasibility of the proposed method.
