Engineous: A unified method for design automation, optimization, and integration

Engineous, the generic software shell described in this chapter, provides a new method for unifying the automation, optimization, and multidisciplinary integration of product design. Engineous combines advanced computational techniques – such as machine learning, expert systems, and object-oriented programming – with such conventional techniques as numerical optimization. Using captured human knowledge, Engineous automatically invokes the appropriate analysis codes and design procedures, deduces the key design parameters, and manages the use of optimization packages. When needed, genetic algorithms are used to uncover unconventional designs. Engineous was tested on a set of six engineering optimization problems that have been proven difficult to numerical optimization techniques. Results demonstrated that it was more robust and efficient than either genetic algorithms or numerical optimization used in isolation. Engineous has produced designs with higher predicted performance gains than current manual design processes, has reduced turnaround time by 10-to-1 (on average), and has yielded new insights into product design. It has been used in a wide variety of applications, including the design of aircraft engine turbines, molecular electronic structures, cooling fans, DC motors, and electrical power supplies, and the concurrent preliminary design, aerodynamic detailed design, and mechanical detailed design of 3D turbine blades.