A Research On Visual Operation Optimization Method Based On MDS-FCM

Selection of optimal variables in process has much relationship with operating control variables. Optimizing is usually setting up numerics for the variables. Optimizing variables for a production process, generally increase productivity, saving the cost or reduce emissions of waste. Many methods have been provided in Multivariate Statistic Process Optimizing, including on-line optimization methods and off-line ones. Generally, off-line optimization methods means analyzing historical operation data and fitting them into linear or nonlinear functions, with the generated functions, analyze optimal operation points.With study on multivariate statistical operation optimizing, that the optimal operation region is also valuable for theoretical study is pointed out. With solving the function of FCM optimal with line-up competition algorithm, a fast and reliable method of FCM objective function was composed. Comparison with genetic algorithm-based objective optimization and standard FCM was made, and shows out that LCA-based method is faster and the objective is better.With integration with MDS and FCM, high-dimensional data was mapped into lower-dimensional space, and then data pattern was clustered in 2-Dimensional graph. Optimal operation region was found and the structure of data was explained by point distances. With 2 true case studied with the method, validity was demonstrated and the factor 'c' in FCM was studied and analysis of determining the value of c was given.At the end, some new technique of ASP.NET and MATLAB was carried on, a browser-based calculation software was developed to do the work of optimizing.