Abstract
The article's goals are to illustrate the feasibility of implementing a Takagi Sugeno state observer on an embedded microcontroller based platform and secondly to present a methodology for validating a physical embedded system using a Hardware In The Loop architecture, where a simulation software replaces the process. As an application, a three water tank system was chosen. For the validation part, LMS AMESim software is employed to reproduce the process behaviour. The interface to the embedded platform is assured by Simulink on a Windows operating system, chosen as it is the most commonly used operating system. The lack of real time behaviour of the operating system is compensated by a real time kernel that manages to offer deterministic response times. The Takagi-Sugeno observer in the case of this process has the complex form that considers the premise variables to be unmeasurable. The embedded system consists of two Arduino boards connected in parallel, thus offering distributed resources.
Highlights
A consistent increase has been seen in the last decade in the usage of small scaled embedded systems in control, diagnostics and signal manipulation, as these solution become cheaper and more powerful [1]
Because it is preferred to test an embedded system prior to its implementation on a real world process in order to avoid unforeseen errors in modeling or communication, it is a useful practice to test the system in a hardware in the loop configuration, where the process itself is replaced by a simulation software, as the a computer that hosts the simulation software (AMESim) software [3], [4]
The z variables, that are dependent on states, have the name of premise variables, and they contain the nonlinear information for the system
Summary
A consistent increase has been seen in the last decade in the usage of small scaled embedded systems in control, diagnostics and signal manipulation, as these solution become cheaper and more powerful [1]. Published under licence by IOP Publishing Ltd sector nonlinearity transformation [5] to the state space form, giving an exact approximation. This model consists of a blend of linear systems that are coupled together by means of weighting functions. After the construction of the state space model, the T-S transformation is applied, and for the newly built representation, the observer is developed.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
