Final Year Project - Modelling of Black-box system using System Identification Appraoch

Title:  Modelling of Black-box system using System Identification Appraoch

Writter: Muhammad Azfar Bin Zukarnain

A low cost hardware such as Arduino have many applications and widely used in industry and also research study. This project will use Arduino as low cost hardware to build a cheap Data Acquisition card(DAQ). This thesis consisting of apply the Black-box modelling in system identification approach. It will consist non linear time-invariant RLC circuits and DC motor. The System Identification will be used to find suitable model structure within which a good model is to be found. In this thesis, it consists three processes or techniques that arranged all through the undertaking’s term with a specific end goal to accomplish the pre-decided destinations easily and viably. The three processes are Hardware development, Software development and Graphical user interface(GUI) development. Hardware development show the experimental procedures and black-box hardware design meanwhile software development explained the process development of modelling the electrical circuit and DC motor system by using Simulink and System Identification Toolbox(SIT) in MATLAB and last but not list, GUI development is a process building the user interface using GUIDE toolbox app inside MATLAB application. Results from this project were obtain through some experiments that have been conducted such as, experiments that model an RLC circuit to know its transient response behaviour when injecting step input from Arduino and Data Driven Control for DC motor modelling. Thus, from this experiment, the data were analysed and validate using SIT. Project findings for this thesis can be conclude that, to do modelling for RLC circuits and DC motor using low-cost hardware can be done, but need to consider the capability of the hardware in term of its operating frequency and ability to sample data.

 Keywords: Simulink; Arduino; DAQ; Black Box; System Identification

The main objective of this project is to use low cost hardware, to connect Simulink with real system for lab practices in Control Engineering bachelors such as SKEM students at Universiti Teknologi Malaysia, UTM Skudai. Currently, expensive DAQs are being used to connect through a computer. When computers or DAQs are outdated, it implies a very high cost. In the case of internal DAQs, connected on an internal bus of the computer, the difficulties are bigger. New Arduino based DAQ must be adapted to the lab prototypes in order to avoid hardware changes or modifications on it.

 Project Background

A Black box model of a system is a system that has no specific information of the character or physics of the model involved. In other words, black box models no physical understanding is accessible or utilized, but the picked demonstrate structure has a place with families that are known to have great adaptability and have been "fruitful previously"[1]. Black box modelling can be useful to do fitting data that related to a particular mathematical structure of the model. Black box modelling also can be described as a trial-and-error process, the parameters of various structures are being estimated and the results also being compared [2].

The Black box models will be using System Identification. System Identification is methods to build mathematical models of dynamic systems by using measurements of the system’s input and output signals. Dynamic systems is the estimations of the yield signals rely upon both the momentary estimations of its info signals and furthermore on the past conduct of the framework. For this project, electrical circuit of Resistor, Inductor and Capacitor(RLC) will be used as the dynamic systems. A model is relationship between system’s input and output variables in term of mathematics. Thus, Dynamic system model are basically described in term on differential or difference equations, transfer functions, state-space equations, and pole-zero gain models [3].

 Practically speaking, engineering problems are hard to fathom. Regularly, numerical strategies or mathematical issues are utilized as systematic answer for such Issus might be non-existent. Numerical methods in themselves are normally iterative in nature requiring a few middles of the road ventures keeping in mind the end goal to touch base at an answer. An RLC circuit (also known as a resonant circuit or a tuned circuit) is an electrical circuit consisting of a resistor (R), an inductor (L), and a capacitor (C), connected in series or in parallel. An RLC circuit is called a second-order circuit as any voltage or current in the circuit can be described by a second-order differential equation for circuit analysis. A linear RLC circuit has a very useful characterization which is given by its Transfer Function, which is (more or less) the frequency domain equal to the time domain input-output relation. These methods has no related with the interior structure of the plant, and it only allowed limited control of the closed-loop behaviour when feedback control is used [11]. In this paper, it is specifically discussed about to make a black box model of an RLC circuit by using System identification methods and use Arduino to measured output and for triggering the input to the circuit.