Innovative Control of an Overhead Crane System

Satisfactory real time control of the overhead crane facilitating fast transit (< 3s) and minimal swing (< 4°) was achieved in the x, y and z directions. Control was achieved in each individual direction using experimental modelling techniques, careful sampling frequency selection, NMSS representation and LQR controller design. To overcome non-linear model variation, a gain scheduling algorithm consisting of seven x and y controllers for varying suspended load lengths was implemented. The experimental system model and controller design methods could easily be repeated, thus enabling adaptability of the algorithm across different overhead crane systems. Safe operation and efficiency of overhead crane systems requires maximal load transit speed and minimal swing. The contradictory nature of these two goals presents a unique and interesting control problem. Pre-existing overhead control systems do not address issues of portability across different crane systems and are generally overly complicated. This thesis presents a simple, easily repeatable experimental system modelling, representation and controller design method that enables rapid control system prototyping and adaptability. Current implementations present excellent real time results in x, y and z directions.