基于卡尔曼滤波和TITAN路径的雷电预警

Lightning proximity early warning technology based on kalman filter and titan path

摘要:

雷电灾害是一种常见的自然灾害,受其发生的随机性、瞬时性、地域性影响,预警预报具有一定的难度。为提高雷电短临预报的准确性和及时性,该文提出了一种新型的基于clique聚类识别和卡尔曼滤波算法进行雷电识别及追踪外推的方法,并在自主开发的广州市雷电监测预警系统中采用该算法和传统的TITAN风暴路径算法分别实现了未来1小时逐6分钟的雷电路径预报。通过2020年5月至10月广州地区闪电定位数据对两种算法的检验分析表明：两种算法基本性能接近,均能有效识别、追踪和预测出大部分雷暴移动路径,且经过优化的新算法在各时次的预测命中率上均已优于传统基于TITAN风暴路径的预测算法,对于提高雷电临近预报的准确性有一定参考价值。

Abstract:

Lightning disaster is a common natural disaster, which is affected by the randomness, instantaneity and regionality of its occurrence, so it is difficult to early-warning and forecast. In order to improve the accuracy and timeliness of lightning short-term and impending forecast, a new method based on clique clustering recognition and Kalman filter algorithm for lightning identification and tracking extrapolation is proposed. In the self-developed lightning monitoring and early warning system of Guangzhou City, the algorithm and the traditional Titan storm path algorithm are used to realize the lightning path prediction in the next hour by six minutes. By using the lightning location data in Guangzhou from May to October 2020, the two algorithms are tested and analyzed. The results show that the basic performance of the two algorithms is similar, and they can effectively identify, track and predict the thunderstorm moving path. Moreover, the optimized new algorithm is better than the traditional prediction algorithm based on Titan storm path in each time, which improves the accuracy of lightning proximity prediction has certain reference value.