ABSTRACT
The intelligent control system refers to the automatic control devices that may have the adaptive capability and some level of intelligence based on human behaviour, ways of thinking and problem-solving skills in an adverse or unknown and unpredictable environment. For many decades, human has made relentless efforts in developing such intelligent human-like creatures so that they can work efficiently and accurately in collaboration with humans and can understand the environments. We are very much aware of the involved complexities such as uncertainties, variability, high non-linearity and disturbances of a robotic manipulator and the computational complications involved in designing the intelligent and knowledge-based controller for such systems. The main purpose of this work is to enlighten the readers and to provide an in-depth insight into the continuous researches and advancements that are taking place in developing the human-like smart and expert systems with suitable intelligence to collaborate each other in the workplaces. This chapter discusses the work done in designing and controlling a two-link robotic manipulator along with model-based controller design using conventional proportional-integral-derivative (PID)-type control and computed torque control. These conventional control methods are associated with many flaws, such as variational parameters, variational payloads, backlashes and uncertainties, but still these are much widely used in industrial applications due to their ease of control and design and lower costs. Hence, for better performance and desired results we cannot rely on these methods but, the continuous advancements in technologies with time have adopted the new alternatives for many new applications along with their associated complications. The present work introduces mainly two types of intelligent controlling approach, namely fuzzy logic control (FLC) and artificial neural network (ANN) control, which are very popular model-free approach and have easy adaptive properties with known complexities. Subsequently, the advantages and disadvantages for each controller are discussed. Finally, the results and conclusion are provided, and also, on the basis of this study, the future scopes are discussed.
