Model Predictive Control for Unmanned Excavator Based on Skilled Operator’s Operation Trajectory

The operational trajectory of a skilled operator holds significant guiding implications for the motion control of unmanned excavators. This study collects and analyzes the operational trajectories of experienced operators under specific working conditions, acquires standardized data, and proposes fluctuation judgment data in terms of joint motion speed and angle values at specific stages. Utilizing standardized data, a motion control trajectory is generated and combined with the Lagrange function and MPC algorithm to achieve decoupling of the four composite motions of excavators. Simulation verification and experimental comparison are conducted using the MPC algorithm. By adopting MPC motion control without altering the structure, hydraulic system, or electrical system of the experimental object, operating efficiency reaches 108.1% compared to that of skilled operators while significantly reducing fluctuations in operation trajectory. The average reduction in relative standard deviation and maximum fluctuation value, by 29.3% and 3.14deg/s respectively, signifies enhancements in velocity stability and position stability of over 36.2% and 12.85deg respectively, thereby presenting a novel approach for controlling work devices on unmanned excavators.