153 Chapter 7 Maintenance Strategies for Sewer Pipes with Multi-State Deterioration and Deep Reinforcement Learning Paper published at L. A. Jimenez-Roa, T. D. Simão, Z. Bukhsh, T. Tinga, H. Molegraaf, N. Jansen, M. Stoelinga, “Maintenance Strategies for Sewer Pipes with Multi-State Degradation and Deep Reinforcement Learning”, in PHM Society European Conference, vol. 8, no. 1, pp. 14, 2024, doi:10.36001/phme.2024.v8i1.4091. Abstract Large-scale infrastructure systems are crucial for societal welfare, and their e!ective management requires strategic forecasting and intervention methods that account for various complexities. Our study addresses two challenges within the Prognostics and Health Management framework applied to sewer assets: modelling pipe deterioration across severity levels and developing e!ective maintenance policies. We employ Multi-State Deterioration Model (MSDM) to represent the stochastic deterioration process in sewer mains and use Deep Reinforcement Learning (DRL) to devise maintenance strategies. A case study of a Dutch sewer network exemplifies our methodology. Our findings demonstrate the model’s e!ectiveness in generating intelligent, cost-saving maintenance strategies that surpass heuristics. It adapts its management strategy based on the pipe’s age, opting for a passive approach for newer pipes and transitioning to active strategies for older ones to prevent failures and reduce costs. This research highlights DRL’s potential in optimising maintenance policies. Future research will aim improve the model by incorporating partial observability, exploring di!erent Reinforcement Learning algorithms, and extending this methodology to comprehensive infrastructure management.
RkJQdWJsaXNoZXIy MjY0ODMw