Predicting Parameters by Neural Network Learning and Nonlinear Controls of the Hydraulic Crawler Excavator Using Them

For vibration attenuation of excavators, this study proposed two controllers: Feedback linearization controller [FLC], and Sliding mode controller [SMC]. The three actuators, consisting of boom, arm and bucket force, simultaneously actuate fine output consisting of boom angle, arm angle and one payload bucket angle. With dynamic modeling using the Lagrange equation, the mathematical analysis of boom, cancer, bucket and terminal was performed, and experimented with a scale of 1/14 excavator to investigate the quality of the two controllers. The performance comparison of the two proposed controllers through system modeling demonstrated the FLC improved the nonlinear system to the linear system, and the performance of the SMC controller improved the vulnerability to non-linearization of the FLC. In addition, in order to overcome the uncertainty structural instability factor using the reduced model, the parameters of the control algorithm were studied in the neural network learning to overcome structural suggestion elements.