随机模糊神经网络在目标状态信息融合中的应用

States information fusion based on stochastic fuzzy neural network

提出一种新的基于随机模糊神经网络的多传感器状态信息融合方法 ,研究和比较了基于单值模糊神经网络和基于随机模糊神经网络的雷达与红外传感器状态信息融合。仿真结果表明 ,当输入被噪声污染时 ,基于随机模糊神经网络的方法离线学习次数更少 ,能更有效地防止噪声的干扰 ,并且融合误差更小。

A stochastic fuzzy neural network is developed with parameter and structure learning. The stochastic fuzzy neural network is investigated for radar and infrared information fusion. The proposed method is compared with the fuzzy neural network method. Simulation results show that the proposed method is of much smaller training time and higher track accuracy.