Research on perception and control technology of construction robots in the context of human-robot collaboration

Construction robots face challenges of high loads and complex tasks during operations on complex steel structure facades, and the existing control systems urgently need improvement in terms of stability and precision. This study investigates the optimization of construction robots through a hydraulic drive-based control system integrated with the PID-MRAC (Proportional-Integral-Derivative Model Reference Adaptive Control) control strategy. The research findings indicate that the PID-MRAC control algorithm exhibits the highest stability and precision in pressure regulation. Under 50% load, the average pressure with PID-MRAC is 15.18 MPa, while it reaches 22.28 MPa with a standard deviation of 0.52 MPa under 100% load. In the motion accuracy test, PID-MRAC achieves 30.0 rad/s within 2 seconds and maintains a range of 30.6 to 31.2 rad/s throughout the test, outperforming the RF-LSTM and BP-PID control algorithms. Results from path planning and obstacle avoidance performance tests show that PID-MRAC has a relatively small fluctuation range of path length, from 49.7 meters to 55.2 meters within 0 to 10 seconds, with an obstacle avoidance success rate ranging from 89.5% to 95.8%, significantly better than other algorithms. Overall, the improved control system significantly enhances the operational efficiency and safety of construction robots, effectively upgrading their adaptability and intelligence level in diverse environments, thereby promoting their widespread application in practical engineering projects.