Self identification and control of four-leg robot based on biological evolutionary mechanisms

This paper provides a method of self identification for robot model, and applies it to four-leg robot's self identification of legs' length. The majority of robots are based on some certain models, which once damaged, may lead to the task fail of robots. The algorithm provided in this paper enables the robot to identify its model, and then the model is applied to create its behavior. This method can keep the robot from total incapacitation when certain damage occurred in the robot.