中文摘要：

变结构多模型方法（VSMM）能够根据目标机动的实际情况实时确定结构不断变化的模型集合，取代传统IMM方法中的固定模型集．使参与状态估计的模型分布更集中，模型数量相对较少，从而提高算法的跟踪精度．多速率模型（Mu｜tirate）能够对原始测量数据进行多分辨分解，有效的抑制量测噪声，从而提高原始数据的信噪比．文章将多速率模型引入变结构多模型方法，提出一种新的机动目标跟踪算法一多速率变结构多模型方法（MR—VSMM），该方法分别从改善原始数据与改善模型集合两方面对传统IMM方法进行了突破．同时，采用多个多速率模型，并将多个模型的滤波结果相互交织实现对机动目标的全速率跟踪．仿真实验证明，该算法较传统的IMM方法跟踪精度得到了一定程度的提高．

英文摘要：

Variable structure multi-model method can produce a model set which is more adjacent to the real mode of system. VSMM replace the fixed model set of IMM using this variable structure model set, which make the distribution of models is more centralized. VSMM improves the performance of algorithm availably compared to traditional IMM. Multi-rate model is used to realize lessening the measure noise through the decomposition of the target＇s original measured results. In the result, multi-rate model can improve the signal-to-noise of original data. The paper imports multi-rate model to VSMM, and advances a new algorithm multi-rate variable structure multi-model method. This new method makes a breakthrough in two aspects-improve original data and model set. At the same time, MR-VSMM uses several multi-rate models and realizes the full-rate tracking of maneuvering target through interlacing the result of these models. The result of simulation proves that this method is more accurate than IMM.