| Engineering design > optimization theory and computer technology have combined closely, forming engineering optimal design which is a modern design method. Through over 30 years of arduous work, researches on optimization theory have made great achievements. However, applications of optimization technique in real world engineering design are not advanced synchronously. The main reason roots in the great complexity of engineering design.To improve the effect of optimal design and to spread the application of optimal design in the engineering fields, the idea of comprehensive optimal design should be applicated. Chapter 1 analyzes the developments of optimization theory and the applications of optimal design carefully. In the recent years, the developments of relevant science and technology fields provide opportunities for the researches of optimization theory and the extensive applications of optimal design. Artificial Intelligence is a new rising discipline which studies the simulation of intelligent behavior on computer. Since mid 1980's, many researchers of domicile and overseas gradually have maken use of Artificial Intelligence in the optimal design. The paper makes research the intelligent methods in the optimal design. In chapter 2, a sort of genetic algorithm based on estimation range is presented. By the means of probability analysis, the convergence area of the optimal result can be estimated. In this way, the search range can be shrinked > the search precision can be improved and the quality of result can be ameliorated. The basal genetic algorithm bears only cross-over operator and mutation operator which makes it has the weak point in the search ability, thus the prematurity often occur and the result often converge to the vicinity of the real optimal point. With the help of the search technique based on traditional optimal methods, gradient operator cross search operator and evolution operator are exploited to strengthen the local search ability of genetic algorithm. The following examples show that the operators mentioned above can enhance the search ability of genetic algorithm. Chapter 4 studies Simulated Annealing algorithm. To spread it in engineering, the classical Simulated Annealing algorithm is ameliorated. The " numerous point Annealing " strategy is presented to strengthen the global search ability of the algorithm which makes it has the character of parallelism. Through break Markovian Chain, search ability is strengthened and the probability of escaping from local optimal point is augmented. In addition, the strategy of remaining optimal result in the search process improves the search efficiency. Chapter 5 studies theinanalysis of sensitivity in the process of optimal design. With the help of the method, in the process of optimal design, designers can know how various variables affect the object function. However, because of using Tailor formula, the information of derivative has to be offered in the traditional analysis of sensitivity. The technique of sensitivity analysis presented is based on the foundation of analysis of variance which dose not need the information of derivative and has the advantage of calculating feasibility. In Chapter 6, the idea of generating optimal preliminary scheme through Case-Based Reasoning is presented. The viewpoint that the engineering design case can be decomposed is discussed. Through the research of engineering design case, general engineering design case can be summed up several finite types according semantic object model which lays a foundation to the design of case library. Further, the search model of engineering case based on the restricted object is abstracted.Visualization in Scientific Computing is the displine which studies how to transfer scientific data into visual and easy unstanding information. The other part of paper emphasizes mainly the application of visualization in optimal design. In Chapter 7, the idea of combining visual modeling and intelligent modeling is presented. In addition, the relevant supporting...
|
PDF Full Text Request