| 摘要 |
PURPOSE:To improve a response accuracy without making the constitution complicated by storing the past increment manipulated variable, the deviating quantity of a trial error and calculating the increment manipulated variable to satisfy the prescribed conditions. CONSTITUTION:Based upon a target value ri of a time (i) from a target value memory 1, a constant alpha<i> to determine a constant Wi from a constant memory 3 with the weight coefficient to the response and the time of the control system, a past increment value manipulated variable mi-j from a past increment value manipulated variable memory 4 and a deviation (i) of the previous trail from a deviation memory 2 of the previous trial, an arithmetic device 5 calculates an increment manipulated variable mi to satisfy a formula I. Since the manipulated variable mi is the optimum value in which the weight sum of the square value of the control deviation to the future target value is minimum, the response accuracy is increased without making the constitution complicated. In the formula, N is the number of times of the sampling to that the response of the control system can sufficiently regulate and (a) is the constant to consider the error of the previous trial.
|
