基于PSO-RBF神经网络的卫星定位补偿模型设计

Design on Compensation Model of Satellite Positioning Based on PSO-RBF Neural Network

随着北斗系统的完善,卫星定位得到了越来越多的应用。但是卫星定位信号易受到干扰,尤其是当行驶车辆处在比较复杂的道路中,使得定位发生漂移现象。针对此问题,提出了粒子群-径向基神经网络联合的车辆卫星定位模型。由于传统卡尔曼滤波不能较好地处理漂移点,通过神经网络阈值与车速、航向角以及经纬度之间的时序相关性,对定位车辆进行阈值判断,并且利用RBF神经网络进行训练,从而得到补偿模型,实现车辆位置的优化。实验表明,面对定位信号干扰,联合补偿模型可以提高车辆卫星定位的精准度以及可靠性。

With the improvement of the Beidou Satellite Navigation System,satellite positioning has been applied to more and more areas.However,the satellite positioning signal is easy to be interfered,especially when the vehicle is in a complex road,which makes the positioning drift.To solve this problem,a vehicle satellite positioning model based on particle swarm-radial basis neural network is proposed.Since the traditional Kalman filter can't deal with drift points well,the positioned vehicle threshold is judged through the time sequence correlation between the neural network threshold and vehicle speed,heading angle,longitude and latitude,and the RBF neural network is used for training,so as to obtain the compensation model and optimize the vehicle positioning.The experiment shows that the joint compensation model can improve the accuracy and reliability of vehicle satellite positioning in the face of positioning signal interference.