top of page
Optimal Control & Estimation
Course Description: optimal parameter estimation; linear least-squares; nonlinear least-squares; constrained least-squares; optimal control problem; linear-quadratic regulator; h∞ optimal control; h2 optimal control; convex optimization for control; receding horizon control; linear-quadratic-gaussian; separation principle; optimal state estimation; kalman filter; extended kalman filter; sigma-point kalman filters; bayes filter; particle filter.
Project 1: Elevator system simulation and Analysis [Link]
Project 2: Optimal Battery Parameter estimation [Link]
Project 3: Optimal Airplane control [Link]
project 4&5: Optimal Vehicle state estimation using EKF, UKF and Particle Filter [Link]
Radar Signal Processing
Syllabus: Introduction to environmental perception of autonomous driving; Radar fundamentals: link budget analysis, resolution, radar waveforms , radar waveforms, ambiguity function ,Principle of linear frequency modulated continuous waveform (FMCW) ,Constant false alarm rate (CFAR) detector, Basics of antenna array , Beamforming, Direction-of-arrival (DOA) estimation: FFT, Subspace methods, Compressive sensing, Introduction to MIMO radar, Waveform orthogonality: TDM, DDM, CDM, Sparse array concept, Sparse array.
Term Project: DOA estimation (FFT, MUSIC, Compressed Sensing) with Virtual Antenna array using FMCW simulated data. (Problem Description) Project solution
Research 1: Range Doppler map generation from RADAR Data Cube for ASL signing. (Code)
Research2: Range Angle (DOA) Estimation using MUSIC for two person walking in opposite direction (Code)
Research 3: Point Could generation from Range Angle map by applying CA-CFAR. (Code)
Machine Learning
Outlines:
1) Supervised and unsupervised learning techniques
2) Dimensionality reduction
3) Clustering
4) Assessment of classification algorithms
5) Neural networks and deep learning
HW2: Occupancy detection using combination of features
HW3: Multivariate parameter classification
HW4: Dimensonality reduction using Opti digit dataset
HW5: Support Vector machine
HW6: Gaussian Mixture Model
HW7: Sequency classificaiton usign Hidden Markove Model(HMM)
bottom of page