OPTIMIZATION OF INITIAL RELIABILITY OF INVERTER DEVICES

At a certain change of interposition of the boundaries of defining influences (external and internal) with respect to the working area it is possible to receive the maximum value of initial product reliability( ), which is determined by using the Laplace function. From the formula given in article it is visible that initial reliability depends on relative dis-tance between the bottom boundary of the working area and the expectation of the lower boundary of the area of capacity ? , on the ratio of the standard deviations of the upper and lower boundaries of area of capacity ? n and on the relative distance between the expectations of the upper and lower boundaries of area of capacity without working area width ? . The problem of finding the maximum value of is reduced to finding the location of boundary of the working ar-ea for given values of n and . Problem is solved by a numerical method. The results showed that the accuracy of the calculation of highly reliable equipment is not provided with opportunities of tabulated Laplace function and a more accurate method of setting a maximum is necessary. Considering reliability according to the scheme of cases and assuming the normal law of distribution, points of rela-tive control are defined. In contrast to the method of numerical approximations obtained formulas are applicable at any values of n, , and allow to find optimization conditions with any accuracy and including the location of the boundaries in which it is impossible to calculate the according to tables of Laplace functions. The resulting equation is approximated and received a condition of optimum reliability in the form of the linear equation which coefficients can be found from tables or by the above formulas. At independence of borders of area of capacity from values of external influences reliability optimization can be performed for each external action separately. In a case when the values of external influences, corresponding boundary values of area of capacity, related to external influences regression equation, the optimization can be considered only in relation of both effects. The error from the use of approximate formulas is less than allowed by the calculation accuracy using tabulated Laplace function.