Nonlinear Optimization

3 credits · Graduate

Key issues in engineering design are the optimization of the design parameters and optimization of overall system performance. The objective of this course is to expose the student to modern techniques in finite dimensional optimization. Topics in unconstrained optimization will include steepest descent, conjugate gradient and quasi-Newton methods. In the field of constrained optimization, topics will include Kuhn-Tucker theory and algorithmic methods such as reduced gradients, gradient projection, penalty and barrier methods. The use of constructive dual methods may also be included. Throughout the course, students will be encouraged to apply the theory to engineering decision problems.