This study proposes an optimal scheduling model for rehabilitation based on the deterioration prediction of existing pipes by using the deterioration survey method for a water distribution system. The deterioration prediction model divides the deterioration degree of each pipe into 5 degrees by using the Probabilistic Neural Networks (PNN). Furthermore, the maximum residual service time is estimated by the calculated deterioration degree for each pipe and pipe diameter. The optimal rehabilitation model by integer programming (IP), based on the shortest path, can calculate the time and cost of maintenance, rehabilitation, and replacement. Consequently, the model proposed by the study can be utilized as a quantitative method for the management of a water distribution system.
- Optimal rehabilitation
- probabilistic neural networks
- maximum residual service time
- © IWA Publishing 2005