基于RBF神经网络的组合导航融合算法

Integrated Navigation Fusion Algorithm Based on RBF Neural Network

提出一种基于径向基函数(RBF)神经网络的组合导航容错算法。该算法将局部滤波器状态估计分组引入作为融合中心的RBF神经网络,通过RBF神经网络的局部特性,实现全局估计的自适应性和容错性。该算法等价于对局部估计的模糊推理。仿真结果表明,该融合算法有较高的估计精度,能够及时检测出传感器故障并在融合网络中予以隔离,不致影响全局估计。

Data fusion algorithm of fault-tolerant multi-sensor integrated navigation system based on the radial basis function （RBF） neural network is proposed. By the local character of the RBF neural network, the estimated states of local filters are divided into groups and inputted into the RBF neural network for the data fusion, so the global estimation is adaptive and fault-tolerant. Simulation results indicate that the algorithm has a good accuracy. It can detect and isolate the sensor fault, so it not likely to affect the global estimation.