针对GPS高程转换问题,给出了基于径向基神经网络转换GPS高程的模型。用实际观测数据对该模型进行了试验,结果表明,用径向基神经网络转换GPS高程精度高于二次拟合法和BP神经网络法。径向基神经网络能够有效克服BP神经网络局部极小值的缺点,并且具有较高的收敛速度,在GPS高程转换方面具有广阔应用前景。 更多还原
【Abstract】 This paper introduced a Radial Basis Function(RBF) Neural Network mode,which was applied to convert GPS height to normal height.The model was tested with observed data.The results showed that RBF Neural Network conversion accuracy than Quadratic fitting and BP Neural Network.RBF Neural Network can effectively overcome the local minimum shortcomings of BP Neural Network and has a high convergence rate;it has broad application prospects in the GPS height conversion. 更多还原