Comfortable Voter Design in Highly Effective Bandwidth Usage and Dynamic Threshold Selection for Safety Critical System

Abstract

Safety and availability of safety critical systems are increased by using masking method. The main aim of this research paper is designing dynamic bandwidth selection for fuzzy voter for Railroad Crossing Control System (RCCS) and masking the faults. The failure of RCCS leads to damage of property and at-times loss of human life. Safety and availability of the safety critical system is calculated. The major limitation of existing fuzzy voters is uncertainties associated with the inputs and three values (a, b and c) are required as threshold. Only fixed factors are used for a particular set of information and acquainted information varies. In this paper, a dynamic fuzzy voter with three automatic values (a, b and c) are proposed in each voting cycle and existing standard weighted average voting algorithm, flexible weighted average voting algorithm , weighted average voting with soft threshold algorithm and Fuzzy weighted average voting are surveyed and their merits and demerits are discussed. Safety overall performance is empirically analyzed by using the fixed fluffy heavy average voter, powerful fluffy voters, fixed fluffy voter with ratings and fluffy powerful fluffy voter with ratings on a TMR system for 10000 voting periods and RMSE is measured for fixed and powerful fluffy voter. The approach uses some form of voting to avoid the classical fuzzy rules used for inference and compute scores that can be used as weighted factors directly instead of computing several rule outputs. We also briefly compare Static bandwidth fuzzy Weighted Average Voter, Dynamic bandwidth fuzzy Weighted Average Voter, Static fuzzy voter without weighted Average Voter and Dynamic fuzzy voter without Weighted Average Voter.

Keywords: Fuzzy Voter, Safety Critical Systems

View PDF