产品多目标意象造型进化设计研究
发布时间：2018-10-05 07:09
【摘要】：消费者重视产品的情感满足,综合运用感性工学、心理学、设计学和智能信息等技术设计蕴含消费者潜在情感需求和喜好的产品显得尤为重要。 本文以分析产品意象造型进化设计的整体流程出发,提出了确定多目标意象、确定研究样本及参数化样本、建立产品意象造型评价系统和建立产品意象造型进化设计系统四个阶段的整体流程,并以其为研究主线。第一阶段中,研究了确定多目标意象的聚类分析法、主成分分析法和多元尺度法等技术方法,并以汽车为例利用SPSS软件通过K.Means聚类分析确定了汽车的“豪华”、“力量”、“稳重”、“亲和”、“可爱”和“动感”六个目标意象。第二阶段中,研究了参数化样本的曲线控制法、参数模型法和频谱分析法等技术方法。提出了汽车造型的表达、分解及参数化。通过曲线控制法定位主造型线的关键控制点坐标值,以此作为研究样本的参数。第三阶段中,研究了建立产品意象造型评价系统的模糊聚类分析、数量化一类和人工神经网络等技术方法,其中人工神经网络包括BP神经网络、径向基神经网络和模糊神经网络。提出了基于模糊神经网络的评价系统输入层、神经元层和输出层的构建,并建立了汽车意象造型评价系统。第四阶段中,研究了产品单目标意象造型进化设计的常用技术方法,其归纳为遗传算法、群智能算法、交互式进化算法和混合算法四种类型的进化算法。提出了产品多意象造型进化设计的模型和基于NSGA-Ⅱ算法的产品多意象造型进化设计技术。利用Matlab开发了基于NSGA-Ⅱ算法的汽车多意象造型进化设计交互系统。对进化结果进行调查分析,表明本文提出的整体流程是可行的,提出的产品多意象造型进化设计的模型和基于NSGA-Ⅱ算法的产品多意象造型设计的应用是有效的。
[Abstract]:Consumers attach great importance to the emotional satisfaction of products, and it is particularly important to design products that contain consumers' potential emotional needs and preferences by using such technologies as perceptual engineering, psychology, design and intelligent information. Based on the analysis of the whole process of evolutionary design of product image modeling, this paper proposes to determine the multi-objective image, determine the research sample and parameterized sample. The whole process of the four stages of product image modeling evaluation system and product image modeling evolutionary design system is established, and the main line of study is the product image modeling evaluation system. In the first stage, the cluster analysis method, principal component analysis method and multivariate scale method are studied to determine the multi-objective image. Taking the automobile as an example, the "luxury" and "strength" of the automobile are determined by K.Means clustering analysis using SPSS software. "steady", "affinity", "lovable" and "moving" six target images. In the second stage, the curve control method, parametric model method and spectrum analysis method are studied. The expression, decomposition and parameterization of automobile modeling are proposed. The coordinate value of the key control point of the main modeling line is located by the curve control method, which is used as the parameter of the research sample. In the third stage, the fuzzy clustering analysis, quantification and artificial neural network are studied to establish the product image modeling evaluation system. The artificial neural network includes BP neural network. Radial basis function neural network and fuzzy neural network. In this paper, the input layer, neuron layer and output layer of evaluation system based on fuzzy neural network are proposed, and the evaluation system of automobile image modeling is established. In the fourth stage, the common techniques of product single-objective image modeling evolutionary design are studied, which can be divided into four types: genetic algorithm, swarm intelligence algorithm, interactive evolutionary algorithm and hybrid algorithm. The evolutionary design model of product multi-image modeling and the evolutionary design technology of product multi-image modeling based on NSGA- 鈪，
本文编号：2252474
[Abstract]:Consumers attach great importance to the emotional satisfaction of products, and it is particularly important to design products that contain consumers' potential emotional needs and preferences by using such technologies as perceptual engineering, psychology, design and intelligent information. Based on the analysis of the whole process of evolutionary design of product image modeling, this paper proposes to determine the multi-objective image, determine the research sample and parameterized sample. The whole process of the four stages of product image modeling evaluation system and product image modeling evolutionary design system is established, and the main line of study is the product image modeling evaluation system. In the first stage, the cluster analysis method, principal component analysis method and multivariate scale method are studied to determine the multi-objective image. Taking the automobile as an example, the "luxury" and "strength" of the automobile are determined by K.Means clustering analysis using SPSS software. "steady", "affinity", "lovable" and "moving" six target images. In the second stage, the curve control method, parametric model method and spectrum analysis method are studied. The expression, decomposition and parameterization of automobile modeling are proposed. The coordinate value of the key control point of the main modeling line is located by the curve control method, which is used as the parameter of the research sample. In the third stage, the fuzzy clustering analysis, quantification and artificial neural network are studied to establish the product image modeling evaluation system. The artificial neural network includes BP neural network. Radial basis function neural network and fuzzy neural network. In this paper, the input layer, neuron layer and output layer of evaluation system based on fuzzy neural network are proposed, and the evaluation system of automobile image modeling is established. In the fourth stage, the common techniques of product single-objective image modeling evolutionary design are studied, which can be divided into four types: genetic algorithm, swarm intelligence algorithm, interactive evolutionary algorithm and hybrid algorithm. The evolutionary design model of product multi-image modeling and the evolutionary design technology of product multi-image modeling based on NSGA- 鈪，
本文编号：2252474
本文链接：https://www.wllwen.com/guanlilunwen/gongchengguanli/2252474.html
