Learning friction estimation for sensorless force/position control in industrial manipulators

We present a novel type of function estimation applied to the field of sensorless force/position control. Friction estimation is used twice in our force/position control: as part of a computed torque controller and in an estimation of external forces exerted by the tool center point (TCP). As a part of a position based Neural Force Control (NFC-P) the estimation friction and external force allows a force/position control without using a force sensor. NFC-P consists of a hybrid force/position controller that accurately generates contact forces to objects with arbitrary flexibility and uncertain distance or shape. NFC-P performs force control by modifying the desired joint angle changes in force direction (position based force control) before they are fed into a computed torque controller. The inverse dynamics of the manipulator is modeled in a computed torque controller. Kinematical mappings guarantee singularity robustness in the entire workspace. Results from real time experiments are presented with a 6 DOF industrial manipulator as test bed. The control functions are implemented on a PC based operating system (RCON). All mappings are solved in a cycle time of 0.7msec (Pentium PC 166)(1).