Multidisciplinary Collaborative Optimization Method Research Based On Reliability And Physical Programming

With the continuous development of society and technology, the products which are very common in our daily life are becoming more and more complex. A product can be seen as a complex system. As for design of complex systems, the traditional optimization and design theory has many defects, such as low efficiency, long product development cycle and so on. Therefore, a new theory named multidisciplinary design and optimization was created by scientists. Utilizing the coupling relationship among disciplines is strongly emphasized by MDO. Through coordinating values of each variable continuously, the global optimal solution can be obtained. The coupling relationship among disciplines must be fully considered originally. In order to realize concurrent analysis and design, the design progress should be organized and controlled by distributed computing network. The MDO theory and its technology system is deeply studied in this paper. MDO methods are deeply discussed and are applied to solve mathematical example.Standard collaborative optimization method has great difficulty in calculation. A quadratic function is applied to illustrate its difficulty, and the fundamental cause is analyzed by mathematical theory. Three kinds of improvement measures are deeply analyzed in this thesis,and are applied to solve a mathematical example. Several aspects are compared, such as error, iterations and so on. Through the comparison, collaborative optimization with slack variable has better performance.In modern society, reliability is a significant index to measure the quality of products. In order to improve the reliability of products designed by MDO, first-order reliability method is fully studied in this thesis, and is also proposed to calculate the reliability of systems in the structure of MDO. Hence, collaborative optimization method based on reliability is proposed and its model is established in this paper. Through verification by a mathematical example, the proposed method is feasible and very effective. The FORM-CO method is used to design and optimize the speed reducer. Through comparison, the reliability is significantly improved.Preference information, for example engineering experience, can be introduced in the progress of design and optimization, so physical programming is frequently used by engineers. Physical programming method is deeply studied and fully discussed in this paper. It is also introduced to the structure of MDO, and collaborative optimization method based on physical programming is proposed. The proposed method is applied to design and optimize tactical missile. Not only has the preference information been considered, but also the global optimal solution is very satisfactory.