基于模糊交互多模型的机动目标被动跟踪算法

针对双观测平台纯方位测量的机动目标,提出了自适应归一化的模糊交互多模型算法。算法利用目标的方位信息,通过模糊推理机制自动调整过程噪声水平。提出了限定记忆归一化方法来自适应归一化模糊系统的输入变量,使得输入变量值始终保持在一个合理的范围内。仿真结果表明,与基于增长记忆归一化方法和经验法的模糊交互多模型算法相比,基于限定记忆归一化的模糊交互多模型算法的跟踪误差分别降低了9.56%和9.29%,能更好地跟踪机动目标的位置、速度和加速度。

Maneuvering target tracking algorithm based on fuzzy interacting multiple model

Fuzzy interacting multiple model algorithm（IMM） with automatic normalization technique is proposed for tracking a maneuvering target with bearing-only measurements of two passive observers.In the algorithm,the fuzzy logic based on the bearing-only information is incorporated in a conventional IMM method to automatically determine the process noise covariance.Normalization algorithm with limited memory is proposed to normalize the input variables of the fuzzy systems,accordingly which can keep the output value in a reasonable range.The simulation results show that the proposed fuzzy IMM algorithm with limited memory normalization technique has 9.56% and 9.29% tracking error reductions comparing with the fuzzy IMM algorithm with growing limited normalization technique and experiential normalization technique.And thus it has better performance on tracking the position,velocity and acceleration of maneuvering target.