Research On Intelligent Processing Algorithm Of Metro Vehicle Monitoring Data Based On ART-2

ABSTRACT:With the increasing of subway transit network, ensure the metro vehicle operation safe and reliable has become a critical issue which needs to be addressed urgently in the field of transportation, and settle mass of monitoring data from metro vehicle is the key problem. So in this paper data processing based on artificial intelligent technology, research ART-2 model network, conducted to judge the train status fast and reliable from a large number of real-time status data belong to monitor key equipment.ART-2 network model could realize self-organized learning process before determining category sample from the angle of pattern recognition on data. Updating algorithm changing network weights, activating priority mechanism and setting supervision function by established recognition layer are conducted to solve the issue traditional ART-2 network existed in pattern drifting, low computing efficiency and large storage space based the algorithm of ART-2 network model processed monitoring dat. The improved network model has more stable and reliable clustering effect, faster pattern recognition capabilities and less storage space.This thesis analyzes connotation, physical structure, thought of essence and key problem for metro vehicle condition monitoring. And it proposes improved algorithm and then one of the classical databases of application pattern recognition named IRIS is utilized to test its superiority against shortcoming of ART-2 network models. At last, the improved ART-2 algorithm is utilizing on data processing for monitoring. It shows that the improved algorithm has more stability, reliability and efficiency.