Collision Avoidance Control for Excavator Bucket Based on Model Predictive Control

To address the real-time collision avoidance control problem of excavator buckets under complex working conditions, this paper proposes a collision avoidance strategy based on mechanism modeling and event-trigger model predictive control (MPC).Firstly, a geometric model is developed to accurately describe the dynamic distance and velocity vectors between the bucket, stick connection cylinder, and excavator arm.Secondly, a kinematic model based on the instantaneous center velocity principle is established to monitor critical pose parameters in real-time. Additionally, an event-trigger MPC strategy is designed, dynamically triggering optimization control based on a two-level warning threshold to ensure real-time performance. The anti-collision strategy proposed in this paper shows better effects than the traditional PID control in most working scenarios. It can effectively avoid collisions in complex and variable working environments, and has good adaptability and stability.