首页|期刊导航|指挥控制与仿真|基于改进闭包多限制条件的船舶周期模式挖掘

基于改进闭包多限制条件的船舶周期模式挖掘OA

Ships'periodic pattern mining based on improved multi-constraint closure condition

中文摘要英文摘要

针对海洋船舶周期模式挖掘中预估周期困难、稀有模式易遗漏、判别约束不具有向下闭包性等问题,提出了一种面向多支持度多周期距离的基于改进闭包多限制条件(IMCCCDC)的船舶周期模式挖掘算法.该算法一方面改进了模式周期的计算方法,可在无须人为预估数值的情况下计算出较为可靠的周期值;另一方面提出了基于改进闭包多限制条件的周期模式判别方法,使得不同模式的最小支持度阈值和最大周期阈值能够根据模式内船舶的出现频数动态调整,从而解决稀有模式的遗漏问题.此外,还给出了 IMCCCDC 算法的向下闭包性和不遗漏性的数学证明,确保了挖掘结果的完备性.利用 AIS 数据检验算法,结果显示:相较于已有的比较算法,IMCCCDC 算法能够以相对较低的运行时间和内存消耗挖掘出最多的频繁周期模式,求得的周期能够更加准确地表征模式出现的规律.

To solve the problem that the period is difficult to predict,the rare patterns are easy to miss and the discrimina-tion constraints are not anti-monotonicity during ships'periodic pattern mining,the multi-support and multi-periodic distance-oriented algorithm based on improved multi-constraint closure conditional discrimination constraint(IMCCCDC)is proposed.On the one hand,the calculation method of the pattern's period is improved to enable the value of the pattern's period more reliable without the need for artificial pre-estimated values;On the other hand,a periodic pattern discrimination method based on improved multi-constraint closure condition is proposed,so that the minimum support threshold and the maximum period threshold of different patterns can be adjusted according to the appearance frequency of the items in the pattern to solve the omission of rare patterns.In addition,the mathematical proof of the anti-monotonicity and the no-omission of the IMCCCDC algorithm are also given.The real AIS data is used to test the algorithm,the results show that comparing with the existing comparison algorithms,IMCCCDC algorithm can mine the maximum number of frequent periodic patterns with rela-tively low running time and memory consumption,and the obtained periods can more accurately characterize the patterns'oc-currence regularity.

张亚伦;杨露菁

海军指挥学院作战指挥系,江苏 南京 210000海军工程大学指挥信息系统系,湖北 武汉 430000

信息技术与安全科学

数据挖掘周期模式AIS数据

data miningperiodic patternsAIS data

《指挥控制与仿真》 2026 (5)

51-60,10

10.3969/j.issn.1673-3819.2026.05.008

评论