ObjectiveFeeding platform is a mechanism designed by imitating manual feeding when sawing wood board. The posture of platform end is the key to ensure sawing quality. A compensation control strategy was proposed in this paper to enhance the accuracy of end posture as the error of platform cannot be simply and efficiently compensated just through hardware.
MethodFirstly, Lagrange method was used to establish the dynamic transfer function of the platform. Through dynamic analysis, the motion characteristics of the key components of the platform were analyzed. By simulation and other methods, according to mathematical solution, the error and its characteristics of the end posture were studied, thus a simple error model was established. Secondly, based on the traditional PID control and the parameter optimization function of RBF neural network, a single neuron PID controller suitable for the feeding platform was designed to compensate the driving displacement relation of each branch chain and to carry out real-time compensation control on the mechanism. Finally, the compensation control strategy was verified and analyzed by the joint method of Matlab and Adams, and the algorithm was transplanted to the controller of the feeding platform successively.
ResultAfter compensation by the single neuron PID algorithm, the offset error of the end trajectory curve of the feeding platform in X and Y directions was reduced from 3 mm to less than 1.5 mm, and the angle error was reduced from 3.5° to 1.5°, and in most curve segments, the end trajectory curve of the platform completely coincides with the command curve.
ConclusionThe SN-PID proposed by this paper can effectively improve the accuracy of the end posture of the woodworking saw feeding platform, and the feeding platform after compensation control can realize the precise cutting task of the wood plate.
DownLoad: