Radar Phrase Recognition Algorithm Based on New Word Discovery
Yang Liu, Weigang Zhu, Shouye Lv · 2020
Aiming at the problem that the previous recognition algorithms based on supervised learning are difficult to apply to non-cooperative multi-function radar, this paper proposes a radar phrase recognition algorithm based on new word discovery. The algorithm can effectively identify radar phrases without prior knowledge. It is verified by simulation experiments that with the increase of the input data size, the algorithm can overcome the problem of decreased phrase recognition rate caused by the loss of data quality. To a certain extent, the algorithm achieves the purpose of compensating for data “quality” with data “quantity”, and has good engineering applicability.