Title: Defect detection-based sewage overflow forecasting using CDS²KNN and ESI-aroma

Authors: Vikram Sadashiv Gawali

Addresses: School of Electronics and Communication Engineering, Dr. Vishwanath Karad MIT World Peace University, India

Abstract: Global changes in climatic conditions and urbanised environments require reliable drainage systems (DSs). By analysing the characteristics of DS, the sewage flood in urban areas is managed with prior forecasting. However, prevailing works failed to forecast the future overflow of sewage due to their defects. This paper forecasts future overflow based on defects and water level in the drainage. The process starts with data collection. Then, the data undergoes preprocessing. Further, the dimensionality of the data is reduced with MDS. Water level, gas leakage, and defects in drainage are categorised utilising CDS²KNN. The images for defect detection are then collected and pre-processed. The resultant image is segmented utilising PPSOA. Thereafter, defect types are categorised utilising CDS²KNN. Seasonal data is collected and clustered via the KGHWMA model. The future overflow is forecasted utilising ESI-AROMA. When compared with the prevailing methods, the proposed system attains an improved prediction accuracy (98.987%) and root mean square error (RMSE) (0.12).

Keywords: CDS²KNN; PPSOA; KGHWMA; ESI-AROMA; drainage management system; defect classification; overflow prediction.

DOI: 10.1504/IJMME.2025.150836

International Journal of Mining and Mineral Engineering, 2025 Vol.16 No.4, pp.374 - 403

Received: 10 Mar 2025
Accepted: 10 Sep 2025
Published online: 23 Dec 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article