Improved genetic algorithm applied in calibrating a three-axis measuring device

It is very difficult to avoid the ortho-error and gain error in three-axis measuring device by means of improving fabrication technics, and the measure precision is greatly limited by these errors. A model for measure system error was built by the mapping relationship between the measure system and a virtual ortho-system. Gray coded improved genetic algorithm was brought forward for calibrating ortho-error and gain error among three measuring axes offline. Simulation results indicate that a triaxial measure system can be calibrated efficiently by this algorithm, and a new approach was provided for solving this kind of problem.