Multiobjective Optimization Based on Response Surface Methodology with Consideration of Input Dependent Noise
Multiobjective Optimization Based on Response Surface Methodology with Consideration of Input Dependent Noise
复制标题
DOI：
10.9746/sicetr.50.792
复制
发表时间：
2014
期刊：
影响因子：
--
通讯作者：
Ryo Ariizumi;M. Tesch;H. Choset;F. Matsuno
中科院分区：
文献类型：
--
作者：
Ryo Ariizumi;M. Tesch;H. Choset;F. Matsuno
在包括控制问题在内的许多工程问题中，需要对多个冲突准则进行策略优化。然而，如果存在可能依赖于输入的噪声和/或在实验在时间和/或金钱上昂贵的情况下引起的评估数量的限制，则这是非常具有挑战性的。本文提出了一种多目标优化（MOO）算法，用于噪声函数的代价估计。该算法结合异方差高斯过程回归方法和标准高斯过程回归方法，从噪声样本中创建合适的代理函数，并找到下一步要观察的点。该算法与现有的MOO算法进行了比较，然后应用于优化蛇形机器人的侧绕步态。
In many engineering problems including control problems, optimization of the policy for multiple conflicting criteria is required. However this is very challenging if there exist noise, which may be input dependent, and/or the restriction in the number of evaluations, which is induced in the case where the experiments are expensive in time and/or money. This paper presents a multiobjective optimization (MOO) algorithm for expensive-toevaluate noisy functions. By incorporating a heteroscedastic Gaussian process regression method as well as standard Gaussian process regression, the algorithm creates suitable surrogate functions from noisy samples and finds the point to be observed at the next step. This algorithm is compared against an existing MOO algorithm, and then applied to optimize the sidewinding gait of a snake robot.