Pipeline networks are commonly used for transporting water, gas, oils and other chemicals in industrial, as well as in non-industrial applications. Any damages to these pipelines that remain undetected will eventually lead to disasters that cost billions of Saudi Riyals every year due to the wastage of fluid and the damage it causes to surrounding infrastructure. It is therefore of a paramount importance to have some reliable pipeline monitoring scheme in place. In pipeline monitoring, various quantities, including as flow, pressure, vibration, temperatures, etc are regularly monitored in order to detect harmful defects such as leak, overload, blockage, fluid contamination and pipe corrosion. Traditional monitoring technologies used wired sensors that involve messy wiring and the cost associated with their installation and maintenance may be prohibitive. On the other hand, a distributed network of wireless sensors can be used to continuously monitor pipes, and detect problems such as leaks in them, in real time. These wireless sensors are battery-powered and have limited energies which may reduce their operational life, unless additional energy is somehow harvested. Various ways of harvesting energy in pipelines rely on physical phenomena such as hydropower, solar, wind, vibration, thermal, etc and produce some vital extra energy to both the sensors and their associated electronics. The challenge remains is the need to use the energy efficiently, particularly if the harvested energy itself is limited. The main aim of this research is to extend the work carried out in our previous projects for developing schemes for detecting and localizing single leaks in simple and complex water pipeline networks. In our previous works, it was assumed that a limited energy was continuously available to the sensors and other electronics through their batteries, and the research therein was mainly about optimal energy utilization in various tasks of sensing, processing and communication, and performing data analysis for timely and accurate leak detection. Simulations were carried out to validate the developed algorithm, and several journal and conference papers were published as outcomes of these projects. The focus of this new proposal is on 3 main points: Firstly, to develop a scheme to effectively harvest energy from vibrations produced by the flow (or other means such as hydropower) in a complex water pipeline structure, and also apply prediction models to estimate the expected harvested energy. Secondly, the optimal utilization of harvested energy for monitoring and leak detection using vibration (and possibly other) sensors and applying a Deep Learning (DL) scheme in this sensor network using a wealth of data that is available during monitoring. Thirdly, to experimentally validate the design and effectiveness of both the harvesting and the associated DL schemes. This will be carried out using our home-grown, lab-scale testbed developed in our own lab. The deliverable of this project will be a new scheme that will allow monitoring pipelines by utilizing the harvested energy, and the solution will be validated first through simulations and then evaluated through experiments. The work will also result in journal and conference publications as well as patents and technical reports. Furthermore, the testbed developed in our lab will serve as a good platform for graduate and undergraduate education and research in this area of paramount importance to the kingdom, as well as a working pipeline monitoring system to demonstrate, to visitors from industry to spawn some possible future collaboration with other interested parties.
|Effective start/end date||15/04/19 → 15/10/20|
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.