Recent progress in intelligent perception and control of excavators

Hydraulic excavators, as crucial equipment in modern construction engineering, have their intelligentization level directly impacting construction efficiency and quality. This paper systematically reviews the research progress in intelligent perception and advanced control methods for excavators. In terms of intelligent perception, it focuses on analyzing environment reconstruction, object recognition, and pose detection technologies based on multi-sensor fusion, while discussing innovative applications of RTK-IMU, LiDAR, and deep learning algorithms in improving perception accuracy and real-time performance. Regarding control methods, it provides a detailed evaluation of four major strategies—PID cooperative optimization, sliding mode control, robust adaptive control, and time-delay control—comparing their technical principles and performance characteristics, as well as analyzing their advantages and disadvantages in trajectory tracking accuracy, robustness, and real-time performance. The study reveals that while existing technologies have achieved significant progress in individual performance metrics, challenges remain in multi-actuator cooperative control, nonlinear dynamic compensation, and full-process optimization. Future research should further explore an integrated “perception-decision-control” architecture, incorporating emerging technologies such as deep reinforcement learning to achieve higher levels of autonomous operation capability. This review provides systematic technical references and research directions for the intelligent development of excavators.