China
Africa
Explore chapters and articles related to this topic
Unification of Artificial Intelligence with Optimization
Published in Jay Liebowitz, The Handbook of Applied Expert Systems, 2019
The major advantage of the PMA procedure is that it can support the trade-offs between the goals regardless of their association with the optimization model or rule base. Three types of trade-offs can be supported systematically (Lee and Song, 1995): Ask the impact of changed goals in the rule base in terms of the objective function in the LP model.Ask the impact of changed goals in the rule base on the other goals in the rule base while sustaining the current objective function value of LP model.Ask the impact of changed objective function value in an LP model on the designated goals in the rule base.
Timelines and Capital
Published in Gennadi Saiko, Bringing a Medical Device to the Market A Scientist's Perspective, 2022
For PMA devices, the survey found that the average total cost from concept to approval was $94 million, with $75 million spent on stages linked to the FDA. The product development costs were around $10 million, and another $9 million were associated with clinical unit development. The results are depicted in Fig. 5.6.
Reliability-based topology optimization using the response surface method for stress-constrained problems considering load uncertainty
Published in Engineering Optimization, 2022
Changzheng Cheng, Bo Yang, Xuan Wang, Kai Long
Generally, it is difficult to make the RBTO work successfully by applying the probabilistic constraints described in Equation (12) directly. For this reason, two approximate reliability techniques including the performance measure approach (PMA) (Tu, Choi, and Park 1999; Youn, Choi, and Park 2003; Meng, Guo, and Wang 2022) and the reliability index approach (RIA) (Periçaro et al.2015) have been developed for describing the probabilistic constraint. Here, the PMA is used due to its great efficiency and stability. Thus, the probabilistic constraint described in Equation (12) can be replaced and the Equation (12) can be further represented as follows: where is the target probabilistic performance value, and can be calculated through tackling a nonlinear equivalent constrained optimization model in standard normal space: in which denotes the target reliability index. It should be noted that the optimum point that makes the performance function take the minimum value is referred to as the most probable target point (MPTP), marked as . In this situation, . Many programming algorithms can be used to find the MPTP. In this article, because of the simplicity and efficiency of the advanced mean value (AMV) approach (Wu, Millwater, and Cruse 1990), it is adopted to find the MPTP, and its formulation is written as follows: where represents the normalized evolution direction of at ; and stands for the gradient vector at , which is calculated by . It is obvious that the core of the AMV is the derivative solution for the performance function with respect to . However, the current performance function consisting of global stress has an implicit nonlinear relationship with random load variables. Even if the gradient solution is performed using the numerical method, it still needs a large number of finite element analysis processes, resulting in long computing time consumption. To reduce the computational time consumption and give the explicit expression between random input load and stress response, a response surface-based reliability design method is presented.