Home  Contact Us  Search  Site Map  Downloads  About Unipass

Eleven MPP Identification Methods 

The most probable point (MPP) is defined as the point with the highest probabilistic density in the failure domain. This point is also known as the most likely failure point and the design point. The MPP locus, denoted as MPPL, is the curve containing all the MPPs that correspond to different levels of limit-state function. A typical MPPL curve in standard normal space (U-space) is illustrated in the figure below.

Figure 1 Illustration of MPPL

UNIPASS™ offers 11 different iteration methods for MPP identification. These methods are listed below.

• U-based/U-linearized MPPL method
• Modified U-based/U-linearized MPPL method
• U-based/X-linearized MPPL method
• Modified U-based/X-linearized MPPL method
•  X-based MPPL (not available in this version of UNIPASS™)
• HL-RF method
•  Modified HL-RF method
• Improved HL-RF method
•  Sequential Quadratic method
• Gradient Projection method
•  Modified Gradient Projection method
•  Simulation Search Method (Non-Gradient Based Method)

Figure 2. MPP Search Method Definition Window

Last Updated 02/08/10

Send mail to info@predictionprobe.com with questions or comments about this web site.
Copyright © 2001-2004 PredictionProbe, Inc.