Intelligent Pavement Maintenance Optimization Technique for Automatic Driving

<div class="section abstract"><div class="htmlview paragraph">Compared to manual driving, autonomous driving is more prone to the rapid development and deterioration of pavement distress due to the concentration of driving paths. Therefore, a reasonable and efficient maintenance strategy is required. To address the challenges posed by the numerous constraints and objectives in the maintenance strategy generation process, this paper proposes a multi-objective optimization-based method for generating pavement maintenance strategies. The approach leverages advanced pavement distress detection technologies to establish an initial maintenance program, incorporating a range of constraints and maintenance objectives, such as cost-efficiency, performance longevity, and environmental impact. The method applies a genetic algorithm (GA) to iteratively refine and optimize the maintenance strategy, ensuring that the solutions align with both immediate and long-term performance goals for autonomous vehicle operations. A case study utilizing real-world road data demonstrates the effectiveness of the proposed optimization method. The results indicate a significant improvement in the maintenance strategy's overall benefit index, achieving a value of 4.37, with a 1.3-fold increase in benefit performance ratio. Furthermore, when compared to conventional maintenance approaches that apply a single repair method (e.g., micro-surfacing, hot in-place recycling, or milling and overlay) across the entire route, the optimized planning resulted in notable performance gains. Specifically, the benefit performance ratios of the optimized plan increased by 6.92% for micro-surfacing, 2.31% for hot in-place recycling, and 1.54% for milling and overlay, demonstrating the advantages of tailored, multi-objective optimization. This optimization method not only provides essential technical support for the intelligent maintenance of autonomous driving routes but also offers valuable insights for future multi-objective decision-making in transportation infrastructure management. It lays the groundwork for more effective and sustainable road maintenance strategies in the era of autonomous driving.</div></div>