PID Control for Models
Update: February 6, 2018
PID control is taking its first step forward this week. Derek and Daniel will be taking the helm on this aspect. Daniel spoke with Dr. Tyagi and he granted us access to LabVIEW. We will be using LabVIEW for the GUI (Graphic User Interface) and the PID control aspect of the models. A bump test was done with the steady state water flow. We will be using this data to develop the tuning parameters for the model.
Update: February 10, 2018
Daniel and Derek began the week by solidifying their understanding of PID controls. They also are firm believers in open source software and codes. So, in the spirit of open source, we have made available a macro-enabled Microsoft Excel spreadsheet with a sample PID code to play with. It is simply a cosine wave being controlled to a set point. It doesn't require the derivative expression to achieve good results but can be used to illustrate the point. Please download and change up the code how you see fit.
CLICK HERE TO DOWNLOAD PID EXAMPLE EXCEL FILE!
As has been covered before there are three components to a PID control the proportional, integral, and derivative. Daniel and Derek have made some progress on the valve and below are initial tests ran using only a proportional component. More work needs to be done but, we are hopeful that by eliminating some of the delays and slack in the valve we can get a more responsive control.
Update: February 19, 2018
We had been working with LabVIEW but decided that we would instead switch to Matlab and use App Designer. We created our first GUI layout and are working on learning how to code it for Matlab. We are hoping to finish the code for the GUI this week. We also had the first working PI controller prototype that attempted to maintain RBP of 1 psi. This was coded using the Arduino IDE. Below is a video of the prototype PI working.
Update: February 25, 2018
The GUI using Matlab has been completed. We currently have it reading and writing data as well as performing proportional control. We still need to work out a couple of bugs with the code but you should expect a fully functioning bench model GUI this week. We had a bit of hiccup on this front as well. Daniel's computer died this week and had to be taken apart and have the hard rive removed to not lose the GUI. Hopefully this will be the one and only casualty of the project. Derek is currently working on the GUI layout for the Pilot scale model and we should begin programming the GUI for that model this coming week as well. We have a video of the GUI dry test showcasing a safety feature.
Update: March 4, 2018
All the bugs have been worked out of the GUI for the Bench scale model. We had to replace the variable power supply on the bench model because it was creating interference on the potentiometer readings. We have a video of us running the GUI on just water while using a PI controller. Enjoy!
Update: March 19, 2018
We have finished our first successful PID controller for the bench scale model that attempted to mitigate the negative effects of slug flow. The results showed that our manually tuned PID controller was able to lower the energy of the slug flow regime. The bench model PID controller revealed that the concept of this type of control method for slug flow control could work but it still needs more research and improvement as the controller was unstable. We will be working on results for the Pilot model PID next.
Update: April 6, 2018
Both the PID controllers for the bench and pilot models are working. Both PID control strategies were able to mitigate high energy frequencies.We are happy with our results and have many ideas of how to improve this technique in the future. Please check our achievements page to see how we faired on some our deliverables.