ROLL FORCE ESTIMATING METHOD USING FUZZY SET THEORY

摘要

PURPOSE: A roll force estimating method using fuzzy set theory is provided to improve thickness accuracy of front and normal parts of the strip by repeatedly compensating error estimated values in which errors are estimated using the fuzzy set theory in the estimation of roll force during hot finish rolling for roll force estimated values. CONSTITUTION: The roll force estimating method using fuzzy set theory comprises a first step(10) of inputting input variables required in the estimation of roll force, that is, roll forces, roll gaps, roll speeds, strip thicknesses and strip widths, and determining cluster numbers(M), cluster center(X0), cluster radius(r) and learning frequency(NUM) for each of the input variables; a second step(20,30,40,50) of proceeding to the next step if the cluster exists after judging whether cluster comprising actual values of the input variables does exist or not, and carrying out the current step again after forming a new cluster in case that the cluster does nor exist; a third step(60) of respectively separating a case of the cluster comprising actual values of the input variables and a case of the cluster not comprising the same for each of the plurality of clusters after the second step, and respectively calculating cluster representative value and cluster development frequency for each of the cases; a fourth step(90) obtaining roll force error estimated values using each of the cluster representative value and the cluster development frequency obtained in the third step after carrying out the third step for each of the input variables; and a fifth step(100) of supplying the calculated final roll force estimated values into a rolling controller by adding the roll force error estimated values obtained in the fourth step to the roll force estimated values, thereby calculating a final roll force estimated value after obtaining roll force estimated values using the input variables.