Fuzzy rule reduction influence on system's accuracy

Considering that some systems have limitation in memory and processing power, storing a full fuzzy rule base might be a drawback. Large rule base might considerably slow down the whole system and significantly affect performance. Thus, the purpose of rule reduction method implementation is simplifying the decision process and making the rule base traversal faster. In this paper several methods for rule reduction are presented and one of them - FURIA is applied to system for fire possibility determining. Applying FURIA, rule base is significantly reduced and tested by simulation of temperature rises in a several cases for high and low temperatures. A data analysis for this measurement shows that decreased rule base has slightly lower accuracy in contrast to a system with full rule base, which means that, by reducing a number of rules, system's energy and memory consumption can be decreased, transmission costs can be reduced and critical event detection made faster.