Search

The major contribution of this paper is shows how to control the speed of Permanent Magnet Synchronous Motor (PMSM) by using Proportional Integral and Derivative (PID) controller. In this work, proposed PID controller is used in both speed and current feedback loops to control the speed and torque respectively. The proposed PID controller having the advantages which are less oscillatory response, less overshoots and takes less settling time for the speed response compared with existing PID controller used in speed control loop. The PMSM drive is used in electric vehicle, electric traction, robotics and industrial dive applications. The PMSM having the salient features which are high efficiency, high torque density, low torque ripple, high power density and robust construction compare to dc, induction, Switched Reluctance Motor (SRM) and Brush Less Direct Current (BLDC) motors. In this paper, the proposed PID controller is developed based on Field Oriented Control (FOC) strategy is well used to achieve the good speed response of PMSM drive using MATLAB R2019b/Simulink software. The proposed PID controller provides an excellent steady state and dynamic speed response in terms of rise time, peak time, settling time and overshoots. In addition, proposed PID controller has overcome the drawbacks of Proportional (P) and Proportional Integral (PI) controllers. The P and PI controllers which process only first order unstable systems can be stabilized and higher order processes cannot be stabilized respectively. These controllers compromised if the system dynamics change over time. [ABSTRACT FROM AUTHOR]

The growth of abnormal cells in the brain causes brain tumours. Statistical calculations on the population of the United States reveal that amongst those diagnosed with a brain tumour, the percentage of patients who have for five years or more after the tumour is discovered is only approximately 36 diagnosis of a brain tumour leads to a decreased chance of a patient's demise. The classic procedure used in general hospitals to diagnose brain tumours is magnetic resonance imaging. Once the MRI reveals the presence of abnormal cell growth, a biopsy is performed on the tissue sample of the suspected tumour mass. Examination of biopsy results helps classify the type of tumour. The aim of this project is split into two sections. The first one is to use MRI scans of the brain to identify the exact type of tumour present in the brain using a Convolutional Neural Network with a static Graphic User Interface to make the results easily accessible. The second section of this project is to compare different types of convolutional neural networks (Modified CNN, VGG16 and ALEXNet) on various parameters to determine which of those would be the most helpful with our project necessities. [ABSTRACT FROM AUTHOR]