Showing papers by "Toyono Inakazu published in 1999"
TL;DR: In this article, a water distribution network is modeled as a black box system, and a water quality simulation model by neural networks is presented, which includes the time lag between the input point and the output point.
Abstract: Recently, managing water quality in distribution pipes has become an important subject for water supply systems. However, it is very difficult to grasp the spatial and hourly fluctuation of residual chlorine concentration which is important for water quality. This paper presents the modeling process for determining residual chlorine concentration in a water distribution network. First, the modeling methodology which we propose in this paper is presented. Regarding a water distribution network as a black box system, we make a water quality simulation model by neural networks. Our model includes the time lag between the input point and the output point. Second, we show the modeling process. This process consists of choosing input points and output points, pre-analysis for neural network structure, modeling, and estimation for results of the model. Third, we apply the modeling process to the actual distribution network as a case study, which proved that our model worked well. Our proposed modeling process can be applied in all seasons and at various places. Therefore, it will be a very useful tool for water quality management.