With fuzzy controller's variable error E and error variety EC for inputs,the paper designed a two-dimension fuzzy controller. By E and EC and the output U,the control rule table was obtained. Based on these,fuzzy relation was deduced,and fuzzy control search table gave the way to confirm two quantitative factors of input variable and two comparative factors of output variable.
Through changing the input variables membership function distribution of E and EC, Fuzzy-PI controller's output is simulated under Matlab simulink, the influence rules of membership function distribution on risetime, overshoot,adjust time is obtained by data analysis.
To optimize the dynamic performance of Fuzzy-Pi controller, the influence of membership function distribution on Fuzzy-Pi controller's dynamic performance is researched. According to self-adjusting rules of fuzzy controller, the proportion coefficient Kp and the integral coefficient Ai are corrected with a two order system as research model.
The errors for the calculated stiffness in the theoretical formulae can be directly obtained from the errors for the independant variables, and theoretical formulae should be used for more accurate calculcation or selection of bearings.
The paper not only analyzed the evolution of Chinese logistics industry by labor-division theory, but also did empirical research on the data of 1980-2003, at the same time, adding investment of fixed assets to eliminate omitted variable bias, using ADF test, PP test, Johansen test and Granger casualty test to measure the effect of logistics development level on economic growth.
Absolute constant error (ACE), variable error (VE) and normalized variable error (NVE-normalized to tested range of motion) of matching were quantified for each subject for each of the six angles matched.
Variable error increased slightly with a greater number of intervening saccades.
The landmark aided targeting precision, but did not eliminate the increase in variable error with additional intervening saccades.
The constant error (i.e., bias) and variable error (i.e., SD of mean constant error) of each pointing movement was quantified.
There was no significant difference in constant error when the hand moved slowly, although there was a slightly higher variable error during slow movement than when the hand was stationary.