Abstract: In multiple sensors track fusion environment，aiming at the difference about the number of the sensors detected the target in every moment，this paper designs a model assamble composed by four models，at the same time，applies the tracking gate and curve-fitting theory in this model assamble and filters use counterpart filter which matches the model，and integrates the result into the fusion center eventually.The simulation result shows that the processing about the measurement outlier and one sensor localization problem are solved effectively in this model assamble，and the method about multiple sensors track fusion based on multi-model is availably at the same time.

Key words: multi-model, multiple sensors, tracking gate, curve-fitting, fusion

摘要： 在多传感器航迹融合环境中，针对各时刻检测到目标的传感器个数不同的情况，设计了一个由4个模型构成的模型集，将跟踪门和曲线拟合理论应用于该模型集中，并针对不同模型用相匹配的滤波器进行滤波，其结果送入融合中心进行统一融合。仿真结果表明：该模型集有效解决了量测野值点的处理和单传感器定位等问题，同时验证了基于多模型多传感器航迹融合方法的有效性。

关键词: 多模型, 多传感器, 跟踪门, 曲线拟合, 融合