Optimum Design Of Cotton Knot Control System Based On Genetic Algorithm

Textile industry is an industry which has thousands years of history.The carding machine is one of the main production equipment in the cotton spinning process,and the demand for carding machines is increasing year by year.However,due to the historical reasons of the textile industry development,there are many problems need to be solved,such as textile equipment lags,low degree of automation,inefficiency and high cost.Therefore,developing the "smart card" is one of the most theoretical and practical research directions in the textile machinery industry.In this paper,I applied genetic algorithm,image segmentation and other new technologies to the carding machine,so that the carding machine can automatic statistics and automatic monitoring.The various parameters can automatically be setting,self-diagnosing,self-correcting for the semi-product indicator online.Pipeline equipment need to achieve intelligent management.Make sure the transportation,quality dynamic and equipment status can be digitally managed.This project has done the following aspects of research and develop:First,according to carding process,carding structure and carding theory,this paper analysis the mechanism of the "neps" and the damage to yarn,at the same time I lists factors which engender "neps".I determined the variable parameters by a lot of experiments.Second,according to the characteristics of the genetic algorithm and the carding process,I determined the basic thought,basic theory and main operational steps of the genetic algorithm.Third,based on the design principle for the control unit— "standardization,economy and reliability",I introduced the “PFS precision adjustment cover gauge system” and “FCT cover gauge measurement system”.Researched identification,operation,processing,analyzing and monitoring of the “CCD photoelectric sensor fine tuning system” provides a guarantee for the research.Fourth,through the analysis of the computer vision detection system principle,I optimized the collected images,including identification,imagery,image digitization,image processing and so on.At the same time,we need considering the sampling and quantization of the target control parameters,it is necessary to correct the light intensity unevenness,this brings convenience for the following research.Finally,segment the image by using the Otsu segmentation method which based on the genetic algorithm,and calculated the number of the highlighted neps.Fifth,according to the carding principle,combining with the characteristics of control unit and variable parameters,I designed the project and determined the fitness function.I select,cross,and change the initial population,and simulating the final results by using MATLAB software.In this paper,the experiment and analysis show that the detection system established in this paper can accurately identify the neps in the cotton mesh.According to the quality of the cotton net,the system gives the corresponding mechanical state reminder,and considering the real-time requirements in the algorithm design,this allows the system to lay a foundation for further development of real-time inspection cotton quality systems and real-time control systems.