Open Collections

Dynamic compensation and sensor fusion for a GSM-based water quality monitoring network

University of British Columbia

With the increasing demand for water, access to clean water is becoming a more challenging problem for people, in both rural and urban communities. The quantity and quality of fresh water resources, both surface water and ground water, are of major concern worldwide. A continuous water quality monitoring system with access to accurate real-time data can play an important role in water quality tracking and environmental protection. However, evaluation of water quality is complicated; on the one hand, a great number of physical, chemical and biological parameters are usually involved. Hence, multi-sensors network is often deployed for collecting a variety of useful water quality information, such as pH value, ammonia concentration, oxidation-reduction potential, temperature, electrical conductivity, turbidity, and the concentration of dissolved oxygen. On the other hand, objectives of in situ testing are complex and dynamic, and the testing environment in the field is also dynamic and harsh. This thesis develops a wireless data transmission platform to solve the communication problem between the monitoring sensor nodes in the field and the base station. What’s more, an individual sensor is only able to make a judgment using a single parameter as evidence. Simplex information is neither sufficient nor reliable, and some parameters also have mutual interference with each other to some extent. Specifically, there should be a systematic way to integrate information from multiple sensors to obtain more accurate and reliable water quality information. Furthermore, allowance has to be made for the variation in the conditions of a sensor, which will affect the sensor accuracy. Therefore, compensation and fusion of sensory data from disparate sources are very necessary to secure a reliable, accurate, and comprehensive monitoring result. By applying Dempster-Shafer theory and Euclidean Distance, this thesis presents a method of assigning four different parameters in the same scale, and combining them into an integrated and reliable quality evaluation result. The necessary methodologies are systematically presented. They are applied to realistic sensory data to illustrate their application and effectiveness.

Text

2016-07-04

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/58372

Mechanical Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/