Transcript John report
Current Work • • Grasp Database – Created database to store all data from tracking system – Created Matlab scripts to import/export all data into a struct – Working on real time gathering of data though there seems to be discrepancies in Tracking Tools dll file and exporting to csv. We have emailed them requesting info. WAM is back – Feels a lot smoother, especially joint 4 – Re-ran friction graphs. Came back same shape What's Next Finish John Wen's controller Gravity compensation is working by optimizing the DH, mass, etc parameters Estimate friction torque vs joint velocity and build that in Add PID on top for very accurate control. He said that we can neglect the position dependent friction as well as coriolis/centrifugal since they will be small (integral will take care of them) He said this is the same as State Feedback controller except that we do not explicitly choose the poles