本课程系统讲解最优控制的两大核心方法:变分法与动态规划。内容首先从变分法出发,介绍如何求解不同边界条件、多函数耦合及各种约束下的泛函极值,进而给出连续时间最优控制的充要条件。课程将重点分析典型问题,如线性二次调节、跟踪控制,以及应用庞特里亚金最小值原理处理含状态约束的最优控制问题,包括最短时间、最省能量、最省燃料等经典模型。作为对比与补充,课程还介绍利用动态规划求解最优控制的基本思想,并讨论其与变分法之间的联系与区别。在多智能体系统中,课程将拓展到微分博弈框架,讲述如何结合最小值原理与动态规划获得分散式最优控制策略。最后,课程全面覆盖上述方法的离散时间版本,以契合现代控制系统在工程和应用中的普遍需求。
Gain a Splash of New Skills - Coursera+ Annual Just ₹7,999
AI Product Expert Certification - Master Generative AI Skills
Overview
Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Syllabus
- Part I Extrema of a Functional via the Variational Method
- 1.1 Fundamental Notations
- 1.2 Extrema with Fixed Final Time and Fixed Final State
- 1.3 Specific Forms of Euler Equation in Different Cases
- 1.4 Sufficient Condition for Extrema
- 1.5 Extrema with Fixed Final Time and Free Final State
- 1.6 Extrema with Free Final Time and Fixed Final State
- 1.7 Extrema with Free Final Time and Free Final State
- 1.8 Extrema of Functional with Multiple Independent Functions
- 1.9 Extrema of Functional with Constraints
- 1.10 Exercises
- Part II Optimal Control via Variational Method
- 2.1 Necessary Condition for Optimal Control
- 2.2 Optimal Control with Fixed Final Time
- 2.3 Optimal Control with Free Final Time
- 2.4 Linear-Quadratic Regulation Problems
- 2.5 Linear-Quadratic Tracking Problems
- 2.6 Exercises
- Part III Pontriagon's Minimal Principle
- 3.1 Pontryagin’s Minimum Principle with Constrained Control
- 3.2 Pontryagin’s Minimum Principle with Constrained State Variable
- 3.3 Minimum Time Problems
- 3.4 Minimum Fuel Problems
- 3.5 Performance Cost Composed of Elapsed Time and Consumed Fuel
- 3.6 Minimum Energy Problems
- 3.7 Performance Cost Composed of Elapsed Time and Consumed Energy
- 3.8 Exercises
- Final examination
Taught by
MA ZHONGJING