基于ANFIS的自适应机动目标状态估计算法

Adaptive maneuvering target state estimation algorithm based on ANFIS

针对基于当前统计(current statistics,CS)模型的机动目标状态估计算法对机动目标加速度的极限值依赖性大的缺陷,提出了一种利用自适应神经网络-模糊推理系统(adaptive neuro-fuzzy inference system,ANFIS)自适应调整目标状态噪声方差的方法。首先利用ANFIS算法对目标机动强度进行估计,进而对目标状态噪声协方差矩阵进行自适应调整;然后利用粒子滤波(particle filter,PF)算法对目标状态进行估计。仿真结果表明,与该方法能够有效提高目标状态估计的精度。

In view of the fault that the traditional maneuvering target state estimation algorithm based on current statistics （CS） model is greatly dependent on the statistical characteristic of the system state vector, an adaptive maneuvering target state estimation algorithm is proposed. Firstly, the adaptive neuro-fuzzy inference system （ANFIS） is used to adjust the system noise covariance matrix in target tracking system, after that, the particle filter （PF） algorithm is used to estimate the target state. The simulation results show that the proposed algorithm can obtain good tracking precision.