Skip to main content
Probabilistic Mechanism Analysis with Bounded Random Dimension Variables
Mechanism and Machine Theory
  • Kang Luo
  • Xiaoping Du, Missouri University of Science and Technology
In the traditional probabilistic mechanism analysis, random dimension variables are typically assumed to be normally distributed. This treatment may not be practical because a normal random variable is unbounded, changing from negative infinity to positive infinity, but an actual dimension variable is bounded with its tolerance range. This work intends to remedy this problem. The approach is to treat dimension variables as truncated random variables within their tolerance bounds. Since the traditional methods are not accurate for truncated random variables, a new probabilistic mechanism analysis method is developed. Its major steps include the linearization of the motion error with respect to truncated random variables, followed by the application of the Saddlepoint Approximation. The proposed method can accurately estimate the probability distribution of the motion error. The proposed method is applied to probabilistic mechanism analyses of a slider-crank mechanism and a four-bar mechanism.
Mechanical and Aerospace Engineering
Document Type
Article - Journal
Document Version
File Type
© 2013 Elsevier, All rights reserved.
Publication Date
Citation Information
Kang Luo and Xiaoping Du. "Probabilistic Mechanism Analysis with Bounded Random Dimension Variables" Mechanism and Machine Theory (2013)
Available at: