Mining Algorithm of Regular Route of Target of Sea Battlefield based on Neighborhood ε-dist

Mining on regular route of target is an important research subject of analysis on sea battlefield situations. Aiming at the problem that there is enormous target data in sea battlefield target, which is difficult to obtain target whole-process track data, and so rules of route may only exist in part of area, this paper proposes one mining algorithm on regular route of target in sea battlefield based on neighborhood ε-dist. This algorithm firstly performs pre-processing on massive sea and air target track data, reduces the complete target track into sequence of several straight-line track segments by using the simplified method based on straight-line, then brings out a kind of track segment similarity distance SSD according to mutual relationship between pairs of straight-line track segments, and based on this distance, neighborhood ε-dist is defined. By using the density clustering algorithm based on neighborhood ε-dist, track segments is divided into several non-intersecting clusters. For each cluster, core track segment will be extracted, which will be combined to form mining results on regular route of target. Experiments show that method proposed in this paper is better than current algorithm in the aspect of mining accuracy of rules in part of route, which can be applied to mining and analysis applications of rules in target route under environment of target data in sea battlefield.