论文部分内容阅读
针对飞行器追逃对抗的二人零和微分对策问题,提出基于数据的积分策略迭代自适应动态规划算法,以求解数学模型未知系统的控制律。该算法利用固定时段内有效的状态和输入信息,建立数据模型,并对其进行基于值函数和控制策略的算法迭代,在平面拦截系统完全未知的情况下得到追逃双方的近似最优策略。仿真结果表明,所得到的双方控制策略能在有限界内无限接近最优解,验证了所提出算法的有效性。
In order to solve the two-person zero-sum differential problem of flight pursuit and confrontation, an iterative adaptive dynamic programming algorithm based on integral strategy of data is proposed to solve the control law of the unknown system of mathematical model. The algorithm uses the valid state and input information in a fixed period of time to establish a data model and iterative algorithm based on the value function and control strategy to get the approximate optimal strategy of both evading and avoiding the plane interception system. Simulation results show that the proposed control strategy can approach the optimal solution infinitely within the finite bound, which verifies the effectiveness of the proposed algorithm.