Linear control laws for singular linear systems

✍ Scribed by Sobral, Manoel (author)

• Publisher: Elsevier Science
• Year: 1964
• Tongue: English
• Weight: 551 KB
• Volume: 278
• Category: Article
• ISSN: 0016-0032
• DOI: 10.1016/s0016-0032(64)90585-x

✦ Synopsis

For a class of quadratic performance indices the optimal control law is a combination of maximum effort (bang-bang) and singular. The singular control law is linear and it is optimal in a hyperplane in the n-dimensional state-space. For practical purposes it is desirable to restrict the class of admissible control laws to be linear. This investigation presents a method of finding a linear control law which is optimal in the sense that it is the singular control law in the singular surface and the best possible linear law elsewhere. Classical calculus of variations and the more sophisticated maximum principle of Pontryagin are the mathematical tools used; the former provides a simple and straight forward method for obtaining the singular solutions while the latter is used to extend the linear singular law to the entire state space.