A Lean Path for Learning Control in Undergraduate Electronics Programmes

Abstract

Listen

Translate article

In recent years, the development and spread of machine learning, robotics and other novel application areas, made indispensable to include such topics in undergraduate programmes for students in the area of electrical and computer engineering. However, this entails that traditional subjects must be covered in less time while maintaining their relevant topics for getting an engineering diploma. In this paper, one presents an approach to streamline the learning and teaching of automatic control. As surveyed in the paper, many universities divide the study of control theory in an initial course with system theoretic concepts followed by two other dedicated to linear systems. Various approaches appear on how to divide the topics over these two units. The study of single-input single-output systems defined by differential equations and the study of transfer functions can be argued to be mainly rooted on the need to introduce the Proportional Integral Derivative (PID) controller. In this paper, motivated by the concept of differential observation, we propose a new organization of the control curricula where a first course on systems theory gathers all relevant general tools and a second course, specific to automatic control, presents both continuous and discrete-time linear control systems in the state-space formulation together with ways to get working control systems in face of actuator saturation, measurement noise and parametric uncertainty. In this approach, PID controllers are presented as a special case of state feedback. The structure can also be useful for Universities developing their curricula in countries where control theory teaching is still being developed.