| Sheet metal forming is widely used in automobile, ship-building, aerospace etc. It plays an important part in national economy. The key of sheet metal forming is the forming process planning. Not only is the sheet metal forming process planning scientific, but it is also artistic. The process planning of sheet metal forming contains human creative power and intuitive experience. Traditionally, the forming process planning was designed by engineer according to book knowledge and intuitive experience. And because of that, it led to the restrictions for the application of sheet metal forming process planning systems. The most important causes of uncertainty in the sheet metal forming process planning are lack of information, abundance of information, and ambiguity. Therefore, the effective handling of the uncertainty and nonlinearity constitutes the most urgent problem for the industry and academic world. Bionic intelligence is partly corresponding to the biological information processing system. It has the power of logical reasoning and associative learning for uncertainty and imprecise factors. In the complex system, it can deal with the modeling or optimization problem, which was constituted by the uncertainty factors. Then, in this paper, the bionic intelligence is introduced to solve the uncertainty, imperfection, nonlinearity problems of the sheet metal forming process planning. Not only will the application of improve the intelligent automation of sheet metal forming process planning, but it is also of great significance in theory and in practice.In this dissertation, the bionic intelligent design theory for sheet metal forming process planning and the key technologies based on the fuzzy theory and immune algorithm of bionic intelligence was studied. The fuzzy synthetic evaluation method was directly introduced to construct the stamped parts forming performance evaluation system, which was used to evaluate the forming performance of the stamped parts. The sheet metal forming process planning system of drawing parts was constituted based on fuzzy theory, and it can simulate the design flow of human being. The calculated algorithm of sheet metal forming technological parameter was proposed. On the basis of fuzzy clustering and statistical theory, the fuzzy clustering knowledge analysis system for springback was built up. The techniques of fuzzy system theory were directly introduced to the blank design of sheet metal forming. And then, the flanging blank fuzzy design system was present. The immune clone algorithm was researched based on immune theory, and the metamodeling was established in the flanging of sheet metal forming, and the process is optimized by the immune clone algorithm.Due to the vagueness and uncertainty in the sheet metal forming process planning, the fuzzy theory was introduced to this field in this paper, which was a special tool that can simulate the design process of human being.Evaluating the stamping performance is a key step in sheet metal forming process planning, which is affected by several factors. In order to get the stamping performance, the knowledge of experts should be considered. Fuzzy synthetic evaluation is an effective decision-making method in the comprehensive evaluation of things affected by multiple factors, and can deal with uncertainty problem with domain expert knowledge. Firstly, the influencing factors were selected by the importance of factors. Shape complexity factors, material performance factors, precision factors were selected to establish the first level influencing factor set. Shape complexity factors could be separated into a triple including geometry factor, maximum projection distance factor between two points in XY plane, the ratio factor between height in the Z direction and maximum projection distance between two points in XY plane. The membership functions of the influencing factors were established according to the different effects of each factor. And then, the evaluation set was built based on the evaluation results of the forming performance. The multi-grade Fuzzy synthetic evaluation algorithm for the stamped part was developed. Finally, the forming performance of a typical stamped part was evaluated based on the developed fuzzy synthetic evaluation model, and the perfect result was achieved. According to the result of the evaluation, the membership function and weight set weighed with the evaluation of the stamped parts.As well known, stamping process design involves lots of empirical knowledge, which prolongs the cycle of design and manufacturing, limits the competitiveness of enterprises. With the development of computer technologies, CAD/CAM/CAE technologies have been widely applied to the design of stamping processes and dies, but it still needs a large number of engineers' experiences. Fuzzy theory was introduced into the process of traditional stamping process design, and a drawing process design system was established, which mainly includes input and output module, knowledge data-base module, process design module. First, according to the object-oriented method, the representation model of stamping knowledge was established. Then, referred to the tool provided by the fuzzy theory converting knowledge data-base to non-linear mapping, the intelligent adjustment model of design process was established. Finally, a typical part was taken as an example to demonstrate the running process of the system, and the result showed that the system is effective.The traditional sheet metal forming process planning needs a gear mount of parameters. The process parameters can be found in the compendium, which was only built up on simple experiment condition. So, the generalization capacity of process parameters has some limitations, which led to process parameters limitations. For that reason, huge amounts of experiential knowledge should be considered in the sheet metal forming process planning. The algorithm for process parameter modification was developed based on fuzzy theory. One is fuzzy modified coefficient, which was used to multiply with classic process parameters in order to obtain modified process parameters. Another is about modified process parameters, which was calculated from the fuzzy system embedded with classic process parameters and influencing factors. At last, the effectiveness of process parameter modification for drawing rate was validated, and the calculating process was demonstrated.In the industrial domain of plastic, there is mass of data of experiments and FEA simulation existed in books, test report, engineer experience, and the data is in the state of discrete. In order to discover the implicit and potential pattern or knowledge, the fuzzy clustering method based on statistic theory was introduced to the springback of sheet metal forming. The springback knowledge analysis model for sheet metal forming was studied firstly. And then, the algorithm for fuzzy clustering was researched in order to establish the fuzzy clustering system for spingback of sheet metal forming. After that, the fuzzy clustering model of spingback was obtained after the fuzzy clustering knowledge analysis, and the effectiveness and accuracy of the proposed model and the developed program for springback fuzzy clustering were validated. On the basis of the research of the springback experiments of Mg alloy plate, the fuzzy clustering method was directly introduced to the springback of Mg alloy plate, and the fuzzy clustering knowledge analysis for Mg alloy plate springback was proposed. The membership function and fuzzy clustering model of Mg alloy plate springback were established to get the fuzzy clustering model of Mg alloy plate. The forecasting result of fuzzy clustering model of Mg alloy plate springback is effective and accuracy.How to design blank shape is the key problem of sheet metal forming. The fundamentals of flanging process were studied deeply, and the new flanging groups were presented. According to the flanging groups, the framework and function of combined blank fuzzy design system of flanging was developed, including straight flanging blank fuzzy design system, stretch flanging blank fuzzy design system and shrink flanging blank fuzzy design system. In order to construct the flanging blank fuzzy design system, the method of Table Look-up Scheme for sheet metal forming process planning was presented. The borders of the flanging part and the blank were separated to construct input-output data set for Table Look-up Scheme method. On the base of the research of Table Look-up Scheme method for knowledge base, the input-output space was separated in order to construct membership function of process variable. The input-output data was used to establish the IF-THEN rules. The problem of rules conflict was solved reliable by rule intensity. After that, the flanging blank fuzzy design system based on Table Look-up Scheme was constructed by product inference engine, singletom fuzzifier and center average defuzzifier. Due to the subjectivity of membership function, rule conflict and rule procurement, the adaptive fuzzy system for flanging blank fuzzy design was described. The influence of separating number of fuzzy space and the shape of membership function was studied. The Gaussian membership function is more suitable for the adaptive fuzzy system for flanging blank fuzzy design than the other types membership function. Fuzzy space would be divided into 3 to 5 parts in the term of computational efficiency and accuracy. The stretch flanging part was taken as an example to demonstrate the running result of the system. With the applications to the reverse flanging part, the effectiveness and accuracy of the blank design capacity of the system was validated by the experiments.To combine the immune clone algorithm and metamodeling method for sheet metal forming optimization, a method for sheet metal forming optimization based on the immune clone algorithm was developed. The mechanism and framework of the immune clone algorithm were studied, such as clone operator, mutate operator and select operator. The immune clone algorithm was programmed by VC++. Because of the complexity and nonlinear of sheet metal forming, it is difficult to establish the model of sheet metal forming. So, in the dissertation, the metamodeling method was used to construct the multi-objective optimization model of sheet metal forming. Firstly, the design results based on numerical simulations were used to construct response surface functions by least square method in order to establish the multi-objective optimization model of sheet metal forming. Secondly, optimal solutions were gained from the optimization of response surface functions by the immune clone algorithm. Lastly, the optimal solutions from immune clone algorithm were verified by finite element analysis. By taking of the stretch flanging part as an example, the multio-bjective optimization design model consistent with the wrinkle and thickness was constructed. The immune algorithm was used to solve the multi-objective optimization design model and the validity and reliability of optimal solution was verified by numerical analysis of the forming process.
|
PDF Full Text Request