Improved genetic algorithm with Levy operator for friction parameter identification and compensation in shield segment erector system

The shield tunnel segment erector is a six-degree-of-freedom heavy-duty electro-hydraulic construction machinery with complex structure, frequent operation, and large load fluctuations. Nonlinear friction is a key factor affecting its control accuracy. To address this, this paper proposes a nonlinear friction identification and compensation method for it. Firstly, an improved Stribeck friction model conforming to the segment erector system’s friction characteristics is established. To solve local optima and premature convergence in genetic algorithms, an adaptive parameter adjustment mechanism based on stagnation detection and Levy flight mutation operator is introduced to improve friction parameter identification accuracy. Finally, a nonlinear friction compensation method was designed and validated on the development and testing platform of the 6 m level full-size pipe segment intelligent assisted assembly control system independently built by our team. Experimental results show that the proposed method reduces the steady-state error by a maximum of 52.56%, the root mean square error by a maximum of 48.64%, and the maximum absolute error is significantly decreased. The tracking accuracy and motion stability of the segment erector are effectively enhanced.