| 摘要 |
<p>An optimization method for obtaining an optimization function falling within a range of an allowance and having large robustness for an input having the allowance such as an error. An optimization area (an area sandwiched between the upper limit function M<+> (x) and the lower limit function M<-> (x)) is set by an input having an allowance such as an error. An equality is established between this optimization area and an optimization function containing an undetermined parameter. A distance of a solution area of a parameter corresponding to the boundary of the optimization area from the boundary is set for the parameter value of the optimization function, and an optimization function f(x) having a parameter value providing the maximum distance is obtained. The dashed line represents an optimization function f(x) having a small robustness and the solid line represents an optimization function f(x) having the maximum robustness and obtained by applying this invention. <IMAGE></p> |
