Abstract
The voting system which has been studied normally consists of n units. It is divided into two types in this thesis. One is the weighted voting system, and the other is the un-weighted voting system. Each of these provides a binary decision (0 or 1), or a decision of abstaining from voting. The weighted voting system output is 1 if the cumulative weight of all 1-opting units is at least a pre-specified fraction of the cumulative weight of all non-abstaining units. Otherwise, the system output is 0. The un-weighted voting system output is 1 if the number of unit of all 1-opting decisions is at least a pre-specified fraction of the cumulative units of all non-abstaining ones. Evaluating the reliability of the voting system is an important topic in the field of planning, designing and control. Compared with other studies in the field, in the present thesis, an intuitive Monte Carlo simulation (MCS) was first developed to find the estimated reliability of un-weighted voting system. Then, the response surface methodology (RSM) with the Box-Behnken design (BBD) and the algorithm of Neural Network are used to obtain the reliability function. In the case of the present study, using the Neural Network is more effective than using the BBD. In the last section, the reliability of a real case presidential election is evaluated.