Toyohiro Umeda

Scheduling system supporting estimated overload for processing plant

This paper discusses a novel method of estimating plant overload and shows an application which supports the making of a feasible production plan for a make-to-order manufacturing type machining plant. The feature of the plant is that jobs beyond the ability of machines can be outsourced for meeting deadlines, therefore an accurate estimation of overload is effective for both cost and delivery times. The verification and validation of the method are shown with application to a typical job shop model, and the effectiveness is shown as a result of application to an actual plant.

Berth Assignment Scheduling for Shipment by Branch and Bound Method with Estimating-Cost Search

In this paper, we describe a scheduling algorithm for berth assignment of shipment in iron and steel works. In the berth scheduling we discuss here, it is necessary to consider both minimization of the waiting-time of shipment for each ship and maximization of the efficiency of work in each berth. To determine the optimal combination between berths and ships, and the optimal order of shipment, we develop an algorithm based on branch and bound method. To improve the efficiency of optimal solution search in this algorithm, we introduce an estimation term of cost-increase for unallocated ships in the evaluation function structure, and use a heuristic search in determination of an initial solution. Moreover, to apply to a larger problem, we also use a practical solving method for divided search space. By using this algorithm, the scheduling system can provide a useful planning result of daily berth assignment within practicable time.