Online Optimization of Pickup and Delivery Problem Considering Demand Forecasting

In this paper, we consider an online pickup and delivery problem. Online optimization is to solve the pickup and delivery problem at a certain sampling time and update the route. The delivery area is modeled as an undirected graph. We consider the problem in which demand forecasting is taken into account under the constraints on the delivery of goods. This problem is reduced to a mixed integer linear programming (MILP) problem. By solving the MILP problem considering demand forecasting, we can realize optimal delivery when the number of orders changes rapidly with time.