Probability Density Evolution Method for Probability Density Function Estimation of Random Variables

A new method for probability density estimation based on the probability density evolution method is proposed in this paper.A virtual stochastic process is constructed so that the basic random variable becomes the value of the stochastic process at certain instants of time.The instantaneous probability density function of the virtual stochastic process is evaluated,and then the probability density function of the basic random variable is obtained by employing the independent random samples.As an example,the ratios of the experimental value to the analytical value of ultimate strain of confined concrete are taken as random variables whose probability density function is estimated with the proposed method.