...Markov决策过程(MDP)的最优控制问题,也就是最优控制问题的随机离散版本,文献提出了值迭代算法(Value Iteration Algorithm)和策略迭代算法(Value Iteration Algorithm)等一系列算法和相关理论。
基于4个网页-相关网页
Along with the increasing of iteration times, iterative range will gradually dwindle until approach to true value, and demarcation iterative algorithm can be constructed.
随着迭代次数的增加,迭代区间将逐步减小,直至逼近真值,由此即可构造定界迭代算法。
Simulation results show that by using this iteration algorithm, the flattened gain can reach the possible highest value of the original amplifier, and noise penalty is the minimum.
模拟计算结果表明,用该方法得到的滤波谱能使放大器的增益平坦在原放大器可能得到的最高增益水平,同时因滤波器的引入带来的噪声特性恶化也最小。
A revised fuzzy control algorithm was developed to accelerate iteration convergence in numerical fluid dynamic simulation by adjusting the value of the under-relaxation factor.
在原来研究工作的基础上,提出了一种改进的模糊控制方法,用以调整粘性流场迭代计算中亚松驰因子的值。
应用推荐