A Pump Operation Scheduling Algorithm for Large-Scale Water Supply Systems

Abstract

The operation of integrated water supply systems has become increasingly complicated owing to reorganization of waterworks. We developed a new method to plan pump operation schedules for large-scale water supply systems. In the proposed method, we define solution space as a network model in which the nodes are cumulative flow rates and the edges are the flow rates at each time step. We treat pump scheduling as a shortest path problem, and the A* search algorithm is used to find the optimum path. One of the characteristics of the proposed method is that the search area is restricted by constraints, such as the water level of tanks. The experimental results show that calculation times increase quadratically, nevertheless, the number of solution states shows an exponential growth with increasing number of pumps. We confirmed that the proposed method can be used effectively for large-scale water supply systems.