论文部分内容阅读
为了研究路面摩擦系数的预测问题,利用JN-1型道路摩擦系数测试仪和BM-Ⅱ型摆式摩擦系数测定仪在试验场4种路面进行对比试验,建立了测量结果与摆式仪的转换模型。运用JN-1型道路摩擦系数测试仪在公路和城市道路上8种路面进行测试,研究了路面摩擦系数的影响因素。应用广义回归神经网络分析方法,以路面等级为分类基础,建立了基于广义回归神经网络的路面摩擦系数预测模型,通过131组试验数据对网络模型进行了训练,利用11组试验数据对网络模型进行预测结果对比。结果显示,模型预测值与实测值的平均误差为3.0%,模型预测结果与实测结果吻合,表明预测模型的正确性和精确性。
In order to study the prediction of pavement friction coefficient, the JN-1 road friction coefficient tester and the BM-Ⅱ pendulum friction coefficient tester were used to conduct comparative tests on four kinds of pavement in the test ground to establish the conversion between the measurement results and the pendulum instrument model. The JN-1 road friction coefficient tester was used to test eight kinds of pavement on the road and the urban road, and the influencing factors of pavement friction coefficient were studied. By using generalized regression neural network analysis method, the pavement friction coefficient prediction model based on generalized regression neural network is established on the basis of pavement level. The network model is trained by 131 sets of experimental data. The network model is tested by using 11 sets of experimental data Comparison of forecast results. The results show that the average error between the model predicted value and the measured value is 3.0%, and the model prediction results are in good agreement with the measured ones, indicating the correctness and accuracy of the prediction model.