A sensorless speed control method for doubly-fed induction machine (DFIM) operating with constant frequency but in variable speed mode is presented in this project work. The control method is based on rotor speed estimation technique by a reactive power model reference adaptive system (MRAS) observer. The presented technique does not depend on any kind of flux evaluation and also independent to the resistance variation of either stator or rotor. The MRAS observer has a capacity for speed catching operation. PI controller is designed and also optimized using algorithm for better dynamic behaviour of the machine. MATLAB Simulink model and the simulation results are shown to check the effectiveness of the observer and also of the controller.

No abstract

The objective of this paper is to implement different tools available in machine learning/artificial intelligence to classify faces and identify different features, highlights, and correlations or similarities between different celebrity faces which can apply in everyday security purposes to identity virtually if the authorized personnel is using certain access or not. The material present in this paper is a literature review of a machine learning model developed by the students. This is a classical problem of machine learning executed using a support vector machine. Images are separated based on sub-images. Each sub-image has been classified into a responsive class by an artificial neural network. The website then fetches the data from the back end and classifies the image into the corresponding personal.

No abstract