Dual-axis cross-coupling controller design using DNA algorithm

This paper presents a new approach to deal with the dual-axis control design problem for a watch surface grinding machine. The cross-coupling effect leading to contour errors is effectively resolved by incorporating a DNA evolutionary algorithm in the PID gain-scheduling control The universal solution seeking capability of DNA algorithm is proposed for finding the optimal weighting parameters of the nine PID control gains for the X and Y axis control loops of the grinding machine. Numerical studies and a real-world experiment for the watch cambered surface polishing performance have been validated successfully. The mechanism of this machine consists of three translational axes and two rotational axes. The platform of the three translational axes belongs to a general one. For the two rotational axes, one rotates about watchcase supporting shaft, another rotates about the clamp of the watchcase. The kinematics equation of this system is constructed to plan the grinding path of the watchcase Use DNA Algorithm to search the optimal nine parameters of all the PID controllers under some searching conditions, which can minimize the contour error. In this paper, CCD smart sensing technique is adopted to measure the cambered watch surface thickness by using robust MRF dynamic image processing. Finally, we use MFC Visual C++ to construct a monitor system and simulate actual grinding work under OpenGL 3D virtual reality.