INTELLIGENT RIVER TEMPERATURE WARNING SYSTEM FOR A POWER STATION USING A DYNAMIC NEURO-FUZZY SYSTEM

摘要

This paper introduces an intelligent river temperature warning system in real time using the Dynamic Evolving Neuro-Fuzzy Inference Systems (DENFIS) to predict the downstream temperature two hours in advance due to the cooling water discharge to a distance of 1km and 5 km downstream from the thermal power station. Results shows that real time predictions using the DENFIS developed can predict the river temperature with a high level of accuracy and can effectively be used as an intelligent warning system to help engineers to take appropriate action in advance on when the potential exists for river temperature to exceed consented limits.