HPD’s Breakthroughs/Advanced Capabilities & Concepts

A Fundamental Breakthrough:

HPD can compute “Exact” Output Distributions!!** (Far supersedes Monte Carlo!!)

(** thus can compute “exact” Prob(Fail)!!)

This enables:

  • Multi-Level Variability Propagation when needed (with Important Implications on attacking problems!)

This is in concert with our Breakthrough Methodology on Stochastic “Modeling” (see HPD)

  • Breakthrough Thinking (e.g., easy Transcendence from Deterministics to Stochastics)
  • Determining FLOWs of: Variabilities (FoVs), Distributions, Stochastic Models, Cpks (mgt desires this); . . .

HPD’s Other Breakthroughs Far Supersede Existing Techniques for:

  • Sensitivity/Contribution Analysis (existing ones have gross omissions & errors!!)
  • Stochastic Optimization for Design Engineering (re. existing ones: (i) Taguchi Methods has severe limitations and a basic fallacy***; (ii) others are by What-If iterations)
  • For Prob(Fail): Can handle several Failure Criteria.
  • For Stoch. Optimization: Can handle several Performance Targets
  • Structuring a Stochastic Problem with the associated “Modeling” (see HPD)! (currently non-existent)

**   Parks, Jean M., 1997, “The two enablers for Holistic Probabilistic Design: Methodology and a complete suite of tools for stochastic analysis.” Proceedings of the Third international Conference on Stochastic Structural Dynamics, Puerto Rico, pp. 9.1-9.29.  (A keynote presentation)

*** Parks, JM., 2001, “On stochastic optimization: Taguchi Methods (TM) demystified; its limitations and fallacy clarified,” Probabilistic Engineering Mechanics, 16(1), pp. 87-101.