Prediction of Fault Occurrences in Smart City Water Distribution System Using Time-Series Forecasting Algorithm

The proposed research work is focused on forecasting the future requirements of water supply based current requirement and also identifying possibility occurrences cracks leaks using ARIMA (autoregressive integrated moving average) model. experiments were conducted real-time experimental hardware. pressure data obtained their p -value less than 0.05, which represents stability in forecasted range between 0.451379 N/m2 2.022273 N/m2. frequency ranges 1.706869 3.065836 (maximum peak) −0.81046 1.042164 (minimum peak). Forecasted at damaged condition lie 2.880788 3.29797 4.866227 5.664348 Similarly, for next 1 year 614.6292 (liters/week) 620.0099 (liters/week), forecast value with maximum ranging from 617.0086 to 628.5465 minimum peaks 611.0967 612.2914 (liters/week). above are a single distribution pipeline.