Research On The Algorithm Of Surface Roughness Filter Based On The Immune Memory

Surface roughness is an important indicator that reflects the micro-geometry shape error of the workpiece surface, also it's an important indicator that assesses the machining quality of various mechanical workpiece surfaces. The assessment of surface roughness experiences from qualitative stage to quantitative stage. In order to adjust to the meet of the manufacturing development under the new situation, since 2000, our country has revised related standards of the original surface roughness, promulgated and implemented a series of new national standards successively, these standards change and reorganize the original standards dramatically. For the changes of the national standards of surface roughness and the requirement of manufacturing development, we submit this issue and conduct a relevant research.Firstly, relevant theories of the surface roughness measurement and assessment are introduced in this paper, two assessment benchmark of surface roughness are contrasted and discussed, the mean line benchmark adopted in many countries is studied in emphasis, several common assessment mean lines are analyzed, the solution of the least square mean line is deduced, the arithmetic mean line and the least square mean line's respective influence on surface roughness assessment is discussed. On that basis, two filter mean lines, namely, standard filter mean line and Gaussian filter mean line, that used in the new and old national standard respectively are elicited, the respective filter performances is compared and studied, and it is found that Gaussian filter is better than standard filter. Then, the composition and the basic functions of the immune system are introduced, implementation of several typical immune algorithms are introduced, and experiment research is conducted, the result shows that immune algorithm has a reliable data analysis function, selection function, and self-adaptive detection function, etc.On the basis of above research content, this paper focuses on the implementation style of Gaussian filter, constructs the rational approximation fractional of Gaussian function, adopts a proper approximation fractional to transform the Gaussian filter implement style into IIR, achieves a highly efficient, precise effect of Gaussian filter, designs a Gaussian filter used in surface roughness measuring, proposes a new method to choose the Gaussian filter's cutoff wavelength by surface profile space feature, for the traditional roughness measuring instrument has the shortcoming that the cutoff wavelength has to be selected by manual, immune memory cogitation is elicited, the artificial immune system for Gaussian filter is constructed, Then the paper carriers out the simulation experiment verification on the designed filter and the constructed artificial immune system by Simulink simulation function in MATLAB, the result shows that system can extract the surface roughness Gaussian filter mean line reliably and effectively, the self-adaptive function of Gaussian filter in surface roughness measuring is realized, and adjusts the developing requirement in automatic measurement technology.