Research On Visual Semantics Quantification And Optimization Of Design Parameters For CNC Machine Tools

CNC machine tool is an important foundation of advanced manufacturing technology.It is of great theoretical and practical significance to study the visual design of NC machine tool for the development of product market.Based on the research method of perceptual engineering,the thesis takes CNC machine tool as the research object,builds the semantic quantification model for the three elements of product visual modeling,color and texture,and outputs the mathematical relationship model between perceptual image vocabulary and product design elements.It provides guidance for design,improves the subjective collocation of designer in the stage of product design,and provides an objective and reasonable method for selecting design elements.The research includes representative perceptual image vocabulary acquisition,visual semantic quantization model construction and genetic algorithm optimization.The perceptual image vocabulary of CNC machine tools is collected by literature induction,and the distance relationship between the two words is obtained by the first screening of word frequency analysis and the semantic grouping experiment.By using multidimensional expansion method and clustering method,the representative perceptual image words of CNC machine tools are obtained,which are used as the lexical basis for the construction of semantic quantification model of modeling,color and texture.Based on the analysis of the representative brands of CNC machine tools at home and abroad,the initial product samples are obtained,and the modeling elements and typical classification of CNC machine tools are obtained by means of morphological analysis,and the key modeling elements are selected through the analytic hierarchy process.The semantic difference method is used to obtain the perceptual evaluation matrix.Based on the quantitative I-type theory,the mathematical relationship between the perceptual image vocabulary and the design elements is obtained by regression analysis between the perceptual evaluation value and the design elements,and the quantification model of machine tool modeling semantics is constructed.Based on PCCS color system,the color matching mode of CNC machine tools is qualitatively analyzed,and the representative color samples of CNC machine tools are obtained by orthogonal experimental design and semantic analysis,and the texture of CNC machine tools is qualitatively analyzed.Based on the qualitative analysis of the material quality of CNC machine tools,the characteristics of material design of CNC machine tools are understood,and the texture parameters of semantic quantization model are determined with Key-Shot design and rendering tools,and the representative samples of texture are obtained based on the orthogonal thought.The perceptual evaluation matrix is constructed by semantic analysis.With BP neural network as the modeling tool,the semantic quantification model between the perceptual image words and design parameters of color and texture is constructed.Genetic algorithm is used to optimize the design parameters to achieve the goal of outputting the optimal design parameters according to the user’s subjective image score.Combined with the model output before and after optimization,the product design scheme of CNC machine tool is generated,and the reliability of the model is verified by calculating the relative error.Based on the semantic quantification model and the control variable method,this paper explores the influence of the combination of different modeling elements,color and texture parameters on the scores of different perceptual images.And the genetic algorithm is used to optimize the model to achieve the goal of choosing the design parameters according to the user’s subjective image score.By using the optimization model and the score prediction model,the complete product design scheme of CNC machine tool is generated,and the reliability of the model is verified by calculating the relative error.