We propose and then evaluate a new framework for finding the physical parameters of an artificial environment which give rise to given target physiological characteristics. We assume that a human is a system that takes as inputs the physical parameters of an artificial environment and outputs physiological parameters in response. We define our task as the inverse problem; we must find the best inputs from given target outputs. Our proposed framework solves the inverse problem using evolutionary computation techniques to optimize an artificial environment. We evaluate this framework using a simulation with a vibration environment and verify that it works.