Objective-oriented Genetic Algorithm Based Dynamical Sliding Mode Control for Slurry Level and Air Pressure in Shield Tunneling

Xiaofei Li,Guofang Gong
DOI: https://doi.org/10.1016/j.autcon.2019.102987
IF: 10.3
2020-01-01
Automation in Construction
Abstract:To keep the slurry pressure balance and excavating face stability in shield tunneling, a modified Smith predictor based and a disturbance observer based dynamical sliding mode control (MSP-DSMC, DO-DSMC) systems are respectively presented for the slurry level regulation and air pressure holding in this paper. The established controller tuning methods can not only balance the set-point response, disturbance rejection and actuator preservation effectively, but also make the control systems robustly stable to the parameter variations, modeling errors, random disturbances and time-varying delays. Through incorporating the domain knowledge into the genetic parameters and ameliorating the genetic operators, a novel objective-oriented genetic algorithm (OGA) based on the periodic supervisor is developed to solve the controller tuning problems. The simulation results demonstrate that the tuned MSP-DSMC and DO-DSMC systems can accurately and robustly track the desired slurry level and air pressure without any overshoot, which are practicable and beneficial for the actuator systems. Both the superior convergence properties of the OGA and the excellent management abilities of the proposed controllers are confirmed by the comparative researches.
What problem does this paper attempt to address?