The objective of this work is to design a controller that may maintain the level of a conical tank which has nonlinear dynamics. Because of the nonlinearity present in the dynamics of the plant, the controller parameters have to be dynamic as well, so that the performance of the system may be enhanced. Hence sliding mode controller which is robust is designed with an adaptive mechanism, so that it could cope up with the varying dynamics of the system. In this paper, three different algorithms were used to study the system behaviour by using simulation. These algorithms were also implemented in real time. When its performance was observed in real time, the adaptive sliding mode controller proved to outperform when compared to reaching law and super twisting algorithm based sliding mode controllers.

No abstract

No abstract

No abstract

Speech is the most natural way of people to communicate with one another. It is a vital medium for communicating a person's thoughts, feelings, and mental condition to others. The process of identifying the intellectual state is the recognition of basic emotion through speech. In human life, emotions are incredibly significant. In this project, the emotion is recognized from speech using Support Vector Machine (SVM) and Random Forest classifiers. These are supervised machine learning algorithms used for both classification and regression problems. SVM classifies data by creating N-dimensional hyper planes that divide the input into different categories. The classification is accomplished using a linear and non-linear separation surface in the dataset's input feature. Random Forest is a classifier that combines a number of decision trees on different subsets of a dataset and averages the results to increase the dataset's predicted accuracy. These classifiers are used to categorize emotions like happiness, rage, sadness and neutral for a certain incoming voice signal. Here, the system is trained and developed to recognize emotion in real-time speech. The result demonstrates that the Random Forest classifier is significantly better, when compared to the SVM classifier.