Design And Implementation Of The Load Fatigue Testing Machine Of The Roadbed

This thesis studied an electronic-hydraulic servo loading system, which is based on the application of Roadbed dynamic response and stability analyses experiment and is used to simulate the stress environment of the roadbed material. Based on the research of that, studying different roadbed material's distortion and the life situation under the different load intensity, will then give the instruction comment to the future construction and the maintenance.At first, consulted numerous related literatures and reference materials at home and abroad, hydraulic system and control system are designed and relative components are chosen or manufactured based on the completion of the loading matching calculations of hydraulic system. Control system's hardware circuit are designed and realized then the control scheme of computer and data acquisition card are finished too. Secondly, combining the basic equations of the institutions, the general mathematical model of force and position control system about valve control unsymmetrical hydraulic ram are proposed and this thesis also gives the simulation study of the mathematical model on MATLAB/Simulink toolbox. Improving the system's performance by the forward feedback control strategy is also studied in this thesis and the validity of this method is confirmed by the simulation and experimentsAnd then an amplitude-phase control method which is based on the Widrow-Hoff learning rules is presented. After the simulation of the method, this method's validity has been proved by the real-time system development environment basing on MATLAB/xPC and the debugging work of the load fatigue testing machine is also completed. The experiment proved that this algorithm does not need to carry on online or the off-line identification to the hydraulic system and it can be easily realized on the projects. This control algorithm is already effective to the movement platform, is also effective to the independent hydraulic system, in the experiment has obtained the good control effect. Finally, the algorithm's stability analyses indicated the learning parameters should be within a certain range so as to keep the system stability.