General methodology and framework for creating haptically enabled fine motor skill training system
MetadataShow full item record
This thesis described the efforts to develop a general framework of haptic based fine motor skill training system for handwriting. Using Global Local Approximation strategy, the proposed Fitting method will provide closest possible replication of expert's skill as a record. A Polynomial Chaos based expansion approach is then used to approximate uncertain system parameters in the interest of providing basic dynamic information of handwriting motion. An optimal control scheme - Linear Quadratic Regulator - was provided to automatically produce a controller that is stable and robust to provide force compensation depending on the information the previous two steps generated. The three steps develop a new methodology in training handwriting and can be further expanded to train more complex skill based tasks such as surgery or playing musical instruments.