Adaptive dynamic programming for optimal control of van der Pol oscillator

Adaptive dynamic programming has been considered as one of the effective methods to solve the optimal control of nonlinear systems and it has lots of applications in academic research and engineering. Firstly, we formulate the optimal control of van der Pol oscillator with the least energy consumpti...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Chinese Control and Decision Conference s. 1537 - 1542
Hlavní autoři: Ji, Zheng, Lou, Xuyang
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.06.2018
Témata:
Adaptive dynamic programming
Adaptive systems
Cost function
Dynamic programming
Mathematical model
moving-data window recursive least-square method
neural network
Nonlinear systems
Optimal control
Oscillators
van der Pol oscillator
ISSN:1948-9447
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Adaptive dynamic programming has been considered as one of the effective methods to solve the optimal control of nonlinear systems and it has lots of applications in academic research and engineering. Firstly, we formulate the optimal control of van der Pol oscillator with the least energy consumption. Then, using an adaptive dynamic programming method to dynamically solve the approximate solution of this optimal control problem. A simplified neural network model is used to approximate the cost function. In addition, an admissible initial control policy for van der Pol oscillator is designed and Lyapunov stability criterion is established. On the basis of predecessors, we present a moving-data window recursive least-square method, which can overcome the computational difficulties and the problem that the parameter of oscillator maybe unknown and various slowly. Finally, the implementation steps of the method are proposed, and the feasibility is illustrated.
ISSN:1948-9447
DOI:10.1109/CCDC.2018.8407371