Operation Scheduling for an Unmanned Coal Port System with Constraint Programming

This study investigates a coordinated inbound and outbound operation scheduling problem for Huanghua Port, one of the largest coal ports in China. The annual throughput of Huanghua port keeps around 200 million tonnes for the last three years, which accounts for 7% of the overall coal consumption in China. As a leading coal terminal in automation, Huanghua port implemented an inbound-to-outbound unmanned operation system in 2020. Intelligent logistics management is the next target to achieve a system-wide operation efficiency. This could be a challenging task due to several safety issues. Firstly, two stockpiles sometimes must be reclaimed simultaneously to accomplish a blending process when loading onto vessels. This may increase their collision risk if the two reclaimers are on the same track. Secondly, operations on one stockpile consist of stacking operations while coals are dumped onto the stockpile, and reclaiming operations while coals are reclaimed from the stockpile. Therefore scheduling of these operations must be aligned such that coals on stockpiles do not exceeds their safety inventories. Moreover, ship loading operations are delicate due to the stability and strength requirements of vessels. In this case, reclaiming and ship loading of the cargos of one vessel must follow a pre-defined sequence. These issues make the operation planning and scheduling of the port logistics system very complicated, and most of these tasks are stilled handled by humans. However, even the most experienced planners can make only short-term local decisions. Once the problem sizes are getting larger, operation conflicts may occur which leads to a system interruption. In this study, the scheduling problem is modeled and solved with constraint programming, a fundamental technique in artificial intelligence [1]. Specifically, the afore mentioned safety issues are captured and addressed with some global constraints provided by IBM ILOG CP solver [2]. Based on a comprehensive experimental analysis, parameters of solving engines are tuned for a better solution. Computational results verified the correctness and effectiveness of this approach. Real world data shows that the presented method is able to effectively improve the operation efficiency and service level of the coal port.