Quality Risk Analysis for Sustainable Smart Water Supply
Rachamalla Sindhuja Goud1, Kota Venkata Nikhil2, Godavarthi Manikanta Varma3, BP Deepak Kumar4
* 1 B. Tech Student, Department of Computer Science and Engineering, CMR Technical Campus,
Hyderabad, Telangana, sindhu.rachamalla02@gmail.com
2 B. Tech Student, Department of Computer Science and Engineering, CMR Technical Campus,
Hyderabad, Telangana, nikhilkota04@gmail.com
3B. Tech Student, Department of Computer Science and Engineering, CMR Technical Campus,
Hyderabad, Telangana, manikantavarma53@gmail.com
4Assistant Professor, Department of Computer Science and Engineering, CMR Technical Campus
Hyderabad, Telangana, deepakkumar.cse@cmrtc.ac.in
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - With the rapid growth of contemporary cities, building sustainable smart water supply systems is facing significant hurdles worldwide. Water quality has a significant impact on every aspect of our lives and is given priority in urban planning. Previous methods of regulating urban water quality have mostly relied on performing routine tests on several physical, chemical, and biological groupings of quality indicators. However, the inevitable lag time for biological indicators has raised the risk of illness, resulting in mishaps like widespread infections in many major cities. In this essay, we examine the issue, the technological difficulties, and the research questions. Next, we offer a potential remedy by creating a methodology for risk analysis specific to the urban water delivery system. To detect risks and detect changes in water quality, we need the indicator data we gathered from industrial activities.
We provide an Adaptive Frequency Analysis (Adp-FA) method to resolve the data using indicators' frequency domain information for their internal linkages and individual prediction in order to produce results that are understandable. We also look into this method's scalability in the indicator, geographic, and time domains. We choose data sets of industrial quality for the application that was gathered as part of a Norwegian project from four separate urban water supply systems: Oslo, Bergen, Strmmen, and Lesund. Comparing the new method to the traditional Artificial Neural Network and Random Forest approach, we examine the proposed method's spectrogram, prediction accuracy, and time requirements. The outcomes demonstrate that our strategy performs better in most areas. Supporting early alerts for industrial water quality risks and additional decision support is possible.
Key Words: Water Quality, Risk Analysis, Industrial Quality Data, Quality Indicators
