The existing solution for the generalized Weng's model commonly suffers from the problem of strong discrete deviation of the fitting points at the initial stage of production when these parameters are preferred for linear fitting,and the deviation of these deviation points from the overall fitting trend line is larger,and the location of the deviation points is relatively more centralized,which increases the influence of the deviation points on the overall fitting degree,and further affects the accuracy of the model parameter preference.To address the existing problems of the model,this paper proposes a new solution for the generalized Weng's model with logarithmic distribution from the study of the distribution form of the model fitting points,which is different from the existing solution with time as the independent variable of the linear fitting function.The new approach employs the logarithm of the time as the independent variable of the linear function,and the distribution form of the fitting points is also changed from uniform to logarithmic distribution.Practical applications demonstrate that the new solution results in a relatively more dispersed distribution of parameter fitting points at the initial stage of production,with a trend that more closely aligns with the trend of the overall fitting points.Consequently,the initial fitting point deviation error is reduced by 2.1%on average,while the parameter fitting curve deviation error is decreased by 1.7%on average,and the overall output deviation error is diminished by 2.87%on average.These results indicate that the new solution is superior to the existing method in enhancing the overall yield fit.
generalized Weng's modellinear functionfittingdeviationlogarithmic distribution
NETL
NSTL
万方数据
