Development of an Object-oriented Optimization Software for Industrial Utilization

Conventionally, optimization problems are usually solved by first deriving governing equations for objective functions before they are actually processed by optimizing algorithms, and the procedure for arbitrary problems must be determined case by case. However, this is usually unpractical to meet industrial needs, because it takes too long to fully understand the problem academically. Therefore, an alternative way to solve such an optimization problem emerges by exploring the trends heuristically with numerical simulations. This paper presents the structure of an interfacing software developed from a generic perspective of object-oriented programming. The concept of the structure is to construct an environment for simulation software, optimizing algorithms and objective functions to share information and solve the problem together by coordinated cooperation. It is shown that an optimization problem can be solved efficiently by an easily constructed and flexibly expandable software with a relatively low cost. Some practical examples are also listed to demonstrate that the concept possesses great applicability to fulfill both industrial and academic needs