Research On Inverse Calculation Method Of Asphalt Pavement Structural Layer Modulus Based On Deflection Database

The structural strength of the asphalt pavement determines the bearing capacity of the road,etc.,which is expressed by the modulus of resilience.In engineering,a drop-weight deflectometer is often used to detect the deflection of asphalt pavement,and the modulus of the road structure layer is calculated from the deflection to judge whether the bearing capacity of the road meets the standard.There is a highly nonlinear relationship between the modulus of the asphalt pavement structural layer and the deflection,which makes the inverse calculation of the modulus of each structural layer of the asphalt pavement very difficult.Previous scholars have carried out a lot of research on the inverse calculation of modulus,and proposed many methods such as neural network method,database search method,and regression formula method.When studying the neural network method,it was found that the error of the positive calculation of the bending value using BP neural network was much smaller than the error of the backcalculation modulus;when analyzing the database search method,it was found that the search objective of the algorithm was the minimum mean squared error,and the source of the backcalculation error was mainly in the process of interpolating the positive calculation,and there was no very suitable numerical interpolation method at this stage,so even if the interpolation method was improved,the backcalculation error was not significantly improved,indicating that the optimization of the algorithm potential of the algorithm is limited.In order to improve the effect of modulus backcalculation,this paper proposes a new modulus backcalculation method after analyzing the geometric characteristics of a large number of bending basins,which takes the angle index as the search target;the concept of bending basin line shape is proposed based on the angle index,and all possible bending basin curves are summarized by nine types of bending basin line shapes;according to these Then,based on the monotonic relationship between the angle index and the structural layer modulus,the dichotomous search method is selected as the iterative method,and the highprecision characteristics of the BP neural network for calculating the bending and sinking are combined to construct a new inverse calculation method for the structural layer modulus of pavement based on the bending and sinking basin curve.Through the selection of initial values in the two cases,the rationality and accuracy of the initial value calculation method proposed in this paper are confirmed.The iterative search process in the cases confirms the efficiency of the dichotomous search method and the rationality of this paper in classifying the line shape of the bending and sinking basin.