A Novel Mechanical Model Simulating a Pavement Deflection Measuring System for Non-Invasive Pavement Monitoring Based on a Reliability-Based Optimization Algorithm

This article presents a novel mechanical model for simulating the behavior of pavement deflection measuring systems (PDMS). The accuracy of the model was validated by comparing the acceleration of the new model with the data achieved through experimental tests fusing a deflection measurement system mounted on a Ford F-150 truck. The experimental test for the PDMS is carried out on a random road profile, generated by an inertial profiler, over a 7.4-mile (12 km) loop around a lake near Austin, Texas. Integrating a reliability-based optimization (RBO) algorithm in a PDMS aims to optimize system parameters and reduce vibrations effectively. The PDMS noises and uncertainties make it crucial to use a robust system to ensure the stability of the system. This article presents a robust algorithm for considering the uncertainties of PDMS parameters, including the damping coefficients and spring stiffness of the supporting brackets. Moreover, it considers the variation of system parameters, such as stiffness of the vehicle’s tire and changing in the weight of the vehicle based on the number of passengers sitting in the cars. This approach ensures the system’s high accuracy and reliability, despite uncertainties and variations. Moreover, the presented algorithm calculates the optimum parameters of the PDMS model by minimizing the acceleration of the front and rear lasers and the rotation of the beam about the y-axis at the center of the beam. The accuracy of the RBO is being evaluated by comparing RBO with deterministic optimization methods. This comparison illustrates the accuracy of the RBO method for obtaining a more precise and reliable PDMS system. It also indicates the effectiveness of the RBO method in obtaining a more accurate and robust system, which can be utilized in real-world applications. RBO significantly minimizes the mean angular displacement of the beam (from 8.68° to 2.18°), as well as the RMS of front and rear accelerations (from 5.0 m/s² to as low as 0.66 m/s²), based on variance reductions.