The MagLev train are equipped with gap sensor for levitation control. the gap sensor, which is measuring the air gap between electro-magnet and rail, and its controlled constant range. the rail of maglev train can be modified by external condition, which is rail joint, land erosion cause by girder and so on.
These conditions, called jumps in this paper, have a considerable affect on the ride comfort and stability of maglev train. Therefore irregularity of guideway surface should be monitored and managed. There are few kinds of methods of detecting rail irregularities. In a previous research, a jump measuring equipment was used. Another method is the GMS(Guideway Monitoring System) which is in shanghai MagLev line. However, these methods are limited as static measurement demanding lots of time and effort. Instead, the rail irregularities could be measured by using sensors, like as gap, acceleration, current signal, during vehicle running.
In this study, we established the maglev system in 3-dof dynamic model and mathematical analysis were carried out using computer simulation. also we compared simulation data and experimental data to validate the virtual plant. after then, we verified the proposed algorithm, which is double integration, using virtual plant.
If we can use the virtual plant, the limits which are velocity and heigh of jump in small scale maglev train could be overcome. Therefore we designed 3-dof dynamic model based on small-scale maglev train and mathematical analysis were carried out while 3-dof dynamic model passing through the disturbance input using computer simulation. To validate the virtual model, Experiment is performance the same condition using small-scale maglev train. the validation of virtual plant is compared simulation with experiment result.
Also, we proposed double integration algorithm which is wavelet method. using virtual plant, we create the acceleration signal and carried out double integration. The result show that double integration algorithm is not suitable for detecting the guideway jumps. we need an additional study for other algorithms like as kalman Filter and so on. and acceleration sensor should be replaced for additional experiment.