Operational stability control of a buried pipeline maintenance robot using an improved PSO-PID controller

Ensuring the structural safety of buried pipelines is critical for urban areas due to their capacity for transporting important liquid supplies and wastes. Due to the dangerous (i.e., harmful gas) and narrow environment of pipelines, using robots instead of manual work has become an excellent solution to maintain the pipelines. However, because of the poor perception (i.e., vision and global positioning system signal), the robots controlled manually are easily yawed or even overturned, resulting in an unstable and dangerous maintaining process. Therefore, our research aims to address the following research question: How can we evaluate and improve the operational stability of pipeline maintenance robots in the poor-perception pipeline environment? Aiming to address this question, this paper (i) proposes a framework of operational stability and its control strategies to evaluate and reduce the robot's overturning probability, (ii) analyzes qualitatively the operational stability and its influential factors using zero-point moments and fuzzy theories, and (iii) implements a hybrid genetic algorithm-particle swarm optimization (GA-PSO) and proportional integral derivative (PID) controller to improve the operational stability of the buried pipeline robot. The proposed framework is simulated by the case of the Dadong Lake deep sewage tunnel system in Wuhan and its maintenance robot. Compared with the ZN-PID and BPSO-PID methods, this method can improve tracking accuracy by 88.9% and 78.6% during verification, respectively. Furthermore, laboratory-level experiments using a scaled-down robot are carried out to validate the effectiveness of the controller by using it in practice. The results of simulations and experiments show that the buried pipeline robots can be effectively controlled by the proposed framework.