本文以桥梁挠度数据作为研究对象,对其测量点进行了关联分析,并依据RBF神经网络强大的函数逼近能力,提出了一种基于神经网络模型来恢复不可靠测量数据的方法。在仿真实验中,通过对比实验(该方法的均方误差为2E-9,线性回归法均方误差为0.6974),证实了该方法在理论和实践上的精确性和可行性。
A novel method based on correlation analysis of bridge checking points and the RBF neural networks is proposed for restoring abnormal nonlinear deflection data. Compared with conventional methods (its MSE is 0. 697 4), the proposed approach (its MSE is2e-9) assures high accuracy and the test results accord with practice. Simulation results verify the effectiveness of the proposed method and the discussed theory.