The Research And Application Of Trough Line Based On Ant System Theory

The importance of weather forecast enhances increasingly with the rapid development of economy, involving practical interests in many fields including social production and people's lives. Meanwhile, as fast developing of the computer technology, auto-processing has become urgent to deal mass real-time monitoring data.There are considerable difficulties to ensure accurate forecasting because there are mass factors that influence the synoptic and even more meteorological conditions causing synoptic variation. Synoptic chart has been widely applied in meteorological field to exactly display meteorological exploration Data. According to the construction experiences, forecasters summarized key elements which influence synoptic system. Besides the essential theoretical basis, they analyze and judge the synoptic conditions according to the key elements, giving trend forecast. The key elements now are playing an important role, and analytical demand is increasingly improved in practical applications. To realize the automation analysis becomes to the research hotspot.Presently, computer auto-recognition trough line system isn't exists in meteorological operation analysis system, therefore, the objective of this research subject is to fill the gap in this field to import the auto-recognition of trough line in the future system. This research will save the forecasters from artificial search and hand draw trough line in synoptic chart. It will also provide rapid,accurate,effort-saving method for key elements analysis, to promote auto-recognition popularization and development in meteorological field.The thesis discusses the trough line, which takes an important role in synoptic system on the base of describing the application requirement and present situation of forecast analysis. After the analyses of synoptic system changes and basic characteristics, it summarizes a simple algorithm strategy of auto-recognition and establishes appropriate trough line recognition method according to a series of algorithm framework of ant system and the swarm intelligence theory. The algorithm includes the following parts:enclosed and normal isobars recognition in pressure field; trough points in pressure trough and center recognition; clustering of the trough points based on ant algorithm. It obtained for the first time the exact cyclone centre recognition in pressure trough in meteorological field, end the current situation that the resent systems couldn't give a comprehensive analysis data center and prevent a new complete recognition. Is also obtained a complete analysis algorithm in trough line recognition and gave a satisfied experimental proof on the 4th numerical forecast data in MICAPS from National Meteorological Center.