Research On Problem Solving Process And Methodology For Conceptual Design Of Complex Product In Hall For Workshop Of Metasynthetic Engineering

Hall for Workshop of Metasynthetic Engineering (HWME) is a creative achievement made by Chinese scientists to deal with complex system. It guides people to integrate the wisdom of the experts, high-performance of computers and a variety of data, information into a unified and powerful problem solving system. Today, HWME has been widely applied in the fields of geography and environment, engineering, medicine, social economy, military and so on. But it is a challenge to construct a HWME facing to conceptual design of complex products due to its complexity and the limitation of current supporting technologies.The function and structure of a system are closely related to each other, so the system platform of dealing with complex problems, how to establish a reasonable structure of HWME is crucial to the capability of this system platform. In this paper, on the basis of the conceptual design of some complex product, we make research on the problem solving process and solving method of HWME from the angle of problem solving. Aiming at providing supporting technologies for HWME system applied for the conceptual design of complex product, this research includes the modeling of problem solving process of complex product, problem decomposition and allocation, problem solving and the theory and methods of multi-objective optimization technique.At first, based on meta-synthesis theory, we make a deep research on problem solving process and solving method of conceptual design of complex product. Scholars at home and abroad have adopted evolutionary computation to solve complex decision problems of economy and society. To use their ideas for reference, we propose a mathematical description of problem solving process in HWME and establish a HWME problem solving process modeling based on evolutionary computation.Problem decomposition is the basis of any complex problem solving. So in the process of conceptual design of complex product using HWME, the decomposition of this problem should be made firstly. Nowadays, many researchers have made research on techniques of problem decomposition involved in conceptual design of complex product, but they still didn't find a universal method. In this paper, based on the analysis of the particularity of problem decomposition of complex systematic problem in HWME, we propose a method of problem decomposition on the basis of the research of expert group. And based on interactive genetic algorithm and the characteristic of design structure matrix, we propose an intelligent method of complex systematic problem decomposition to solve the decomposition in general. Practice shows that using this method can avoid modification in large scale and improve the searching efficiency in problem-structuring of complex system analysis.In the process of conceptual design of complex product, with the increase of scale and complexity of the problem to be solved, a key subject to be solved in the problem solving process of HWME is how to harmonize subjects, how to dispatch the resources to make the complex problem solving procedure more reasonable and ordered. The existing complex problem solving procedure in HWME is a kind of empirical workflow standard. As a human-computer cooperative system, HWME need intelligent support from computer. Based on the analysis on the similarity between complex problem solving process in HWME and ant colony algorithm, we make corresponding mathematical model and propose the dynamic allocation of task in complex problem solving and design the ant colony algorithm suitable for solving the problem which achieves good result.The key task of complex product conceptual design is to form a design of the product through generation and choice of concept. In HWME, there is a challenge of how to get the most satisfying and optimum solution to improve solving efficiency during the process of finding conceptual design of complex product after the expert group discussion in HWME. From the angle of complex problem solving and based on current artificial intelligent technology, we propose a group discussion method based on improved interactive genetic algorithm (IGA) model. Improved interactive genetic algorithm gives full play to the advantage of human and computer and provides a effective way to make complex product conceptual design and to solve such difficulties in HWME as the manipulability of man-machine cooperation and the effective conversion from qualitative to quantitative. Compared to the traditional method of opinion consensus and thought convergence in HWME, it is more effective and easier to achieve agreement in expert group. And it confirms the reality of discussion online in HWME and can do repeated circular argument conveniently.During the process of forming product design of complex product conceptual design, the designers are supposed to think fully about various constraint conditions and design objective and meet the problem of making multi objectives in given feasible region as best as possible. In this paper, we propose a new multi-objective optimization method for complex product conceptual design based on the combination of co-evolutionary algorithm and interactive genetic algorithm. We can form a multi-objective plan for complex product conceptual design through the cooperation of population of co-evolutionary algorithm. And during the evaluation process of the design, we can integrate designers'subjective preference, intuition and experience and creative knowledge into the formation of design through IGA to combine man and computer tightly. We take the design of functions of mobile phone products as an example to prove the high effectiveness of this method for complex product conceptual design related to multi-objective and human-computer interaction.