Numerical solving for optimal control of stochastic nonlinear system based on PF-PSO
Yong Xiong, Xiaohao Qu · 2016
Getting general solution for optimal control of stochastic nonlinear system is a hard task, the main difficulty is state variable and the control variable influencing each other,so system does not satisfy the separation theorem.To solve this problem, this paper use horizon optimal control strategy, adopting particle filtering method to get system state variable and variance function in every step with regard to measure value, then changing multi-variable function about control variable into optimizing index function, finally, using particle swarm optimization algorithm to solve global extreme value of this multimodal function, then, numerical solution for the problem was found.Simulated experiment result indicates that this method is an appropriate method for optimal control of stochastic nonlinear system and shows fairly good effect.