Hello everyone,
My issue is about a water distribution system that I am working on a zone of the system where does not exist any plan for the place of pipes. However, we have the place of actuators like different types of valves, pressure relief valves, pressure meters, flow meters and tanks. Also, we have the place of demands where suffer from pressure loss. Now the question is how we can do pressure management using actuators to maximize water pressure for all demands of the zone based on the previous recorded data while we are going to minimize water loss as well as pipes damaging. Please let me know if you have any idea or you know any suitable paper for this issue.
Thanks.
It is an interesting project. The first step should be collecting data. Open more or less systematically taps and valves and record angles, numbers of rotations, pressures, flow. With that data you can train a neural network. In the end the neural network is a model of your pipe system. You can then do simulations by varying inputs and try to achieve the desired outputs.
In hydrology, fluorescein sodium is used to find out the actual waterways. I would be surprised if there weren't any microbiology algorithms to infer pathways. Network optimization is often done with Ford Fulkerson.
Regards,
Joachim
It is an interesting project. The first step should be collecting data. Open more or less systematically taps and valves and record angles, numbers of rotations, pressures, flow. With that data you can train a neural network. In the end the neural network is a model of your pipe system. You can then do simulations by varying inputs and try to achieve the desired outputs.
In hydrology, fluorescein sodium is used to find out the actual waterways. I would be surprised if there weren't any microbiology algorithms to infer pathways. Network optimization is often done with Ford Fulkerson.
Regards,
Joachim
Dear Joachim,
Thank you very much for your answer. I have the recorded data of the system for over two years in a database. However, the problem is that the main part of the zone is old and there is no actual network for this zone. I have just the place of valves, pressure meters, flow meters and tanks as well as their GIS and recorded data. But there is no network graph for using Ford Fulkerson . Regarding these data and the pressure loss, I have to find out a meaningful pattern among actuators and demands.
Based on these conditions, do you have any idea?
Thanks again for your consideration.
Vahid
@Aramide,
Creating viable water channels is very costly. However, data mining is not very costly. So, I am trying to find a data-driven model to make a kind of pseudo graph for predicting the whole network.
Do you have any idea about this issue?!
Maybe of interest, but the knowledge about the existence of additional pipe connections does not increase the possibilities to optimize.
Article Locating the Source of Diffusion in Large-Scale Networks
Article Hierarchical structure and the prediction of missing links in networks
Regards,
Joachim
- Badges
- Science topic
- Similar topics
- Computer Science
- Data Mining