Closed Loop Subspace Identification of a Real Time Non Square MIMO System

When creating model-based control systems, the process model is extremely important. Here, uses a system identification method to create the process model. There are two different ways that this identification process develops: the open loop and closed loop approaches. We’ve set up a conical tank system on a laboratory scale using a non-square MIMO architecture as part of the project. Both with and without feedback techniques yield the system function for this conical tank system. We discuss the reasoning behind using these two methods. We use our laboratory-scale conical tank system to test the open loop system identification method, and we build a PI controller based on the knowledge we obtained from this process. We next use our conical tank system to implement the PI controller in real-time, and we record the data for closed-loop identification. We use a modified version of the N4SID algorithm to conduct the locked in kind of technique computation, which provides exceptional model achievement than the CVA method. The proposed / suggested improved N4SID identification technique is suitable for type of non square system matrix bsased MIMO systems.