基于灰色马尔可夫模型的无线传感器网络目标跟踪算法

基于预测的目标跟踪是无线传感器网络目标定位与跟踪中很重要的一种方法，但在实际环境中由于目标运动模式的动态变化等原因，传统预测算法对目标位置的预测往往不准确。为了克服以上不足，提出了一种基于灰色马尔可夫模型的目标跟踪（GMMTT）算法。将具有震荡特点的Markov模型引入到分段灰色预测中，使目标定位既能获得较好的精度，又能适应目标运动模式的变化。预测得到目标位置后，当前跟踪节点将跟踪信息传送到目标下一时刻将要到达的预测区域，提前唤醒该区域内的节点，用尽可能少的节点实现目标有效的跟踪。实验结果表明：GMMTY算法在跟踪概率和跟踪精度方面具有较好的性能。

Grey Markov model-based target tracking algorithm for wireless sensor networks

Prediction-based target tracking is a very important method for target location and tracking in wireless sensor networks, however, a common difficulty in traditional prediction algorithm is that the target location prediction is not accurate in real environment, due to dynamic change of target motion mode, etc. To overcome these shortage, propose a grey Markov model-based target tracking （GMMTT）algorithm. Introduce Markov model with turbulence characteristics into segmentation grey prediction, which can not only achieve good precision of target localization, but also adapt to change of target motion mode. After getting predicted position of target, current tracking nodes propagate tracking information to the predicted area where the target will arrive in the next moment, and wake up nodes in this area in advance, so that effective tracking of target can be achieved with as few nodes as possible. Experimental result shows that GMMTI＂ algorithm has superior performance in tracking probability and tracking precision.