| Internal combustion bullet train does not need to run on electrified railway,so it has good economic benefits and a wide range of use.At present,dual-layer vibration isolation system is used to passively isolate the vibration response caused by the toppling moment and unbalance force of the internal combustion powertrain mounted under the vehicle body.With the technological innovation of power unit and light weight of car body,the passive vibration isolation mode,which cannot change the stiffness,damping and installation position of vibration isolator,is difficult to meet the increasingly stringent national standards and the pursuit of passengers for ride comfort.Compared with the passive control,the active control can adjust the system parameters in real time through the controller according to the current working condition,so as to achieve better vibration reduction control.However,the structure of active control system is complex and the economic benefit is poor.The semi-active control has the same vibration isolation effect as the active control.At the same time,the system structure is simple,the energy consumption is low,and the system stability is better.The quasi-zero stiffness vibration isolation and the corresponding control theory are expected to solve the contradiction between the stability and the vibration isolation performance of the two-layer vibration isolation system of power unit,and the research on it has certain theoretical reference significance and engineering application value.In this paper,a multi-objective optimization algorithm is first adopted to target the basic dynamic reaction force,the vibration intensity of the unit and the system decoupling degree.The vertical stiffness of the system is used as the design variable to optimize the genetic algorithm.The number and installation position were optimized,and the vibration isolation performance of the system before and after optimization was compared.Then,the static and dynamic experiments of the improved magnetic suspension vibration isolator are carried out.Based on the theoretical model of electromagnetic force,the static model of electromagnetic force is established by function fitting.Based on the experimental data,a more practical model of electromagnetic force is established by the non-parametric modeling method of genetic neural network.The influences of misalignment of the armature,tilt of the armature,magnetic coupling,magnetic leakage and input current direction on the electromagnetic force are also discussed.Then,considering the strong nonlinearity of the magnetic levitation isolator and the wide frequency domain,strong coupling,and multiple operating conditions of the power unit,a robust sliding mode control algorithm was used to design the magnetic levitation isolator to be applied to the first and second stages.Level controller,and perform simulation verification on the two-freedom model,flexible dynamic system and single-degree-of-freedom system of the power unit,At the same time,considering the hysteresis phenomenon of the actual dynamic electromagnetic force,the dynamic electromagnetic force model is brought into the simulation system.Considering that the armature misalignment has a great influence on the electromagnetic force model.In this paper,an adaptive control algorithm is used to detect the armature offset of the magnetic suspension vibration isolator.The influence of armature offset on the model of the magnetic suspension vibration isolator is compensated by the strategy of different currents being passed into the upper and lower electromagnets.Finally,the low frequency resonant frequency and high frequency non-resonant frequency control experiments are carried out on the single-layer vibration isolation test platform.The control effect of the two-layer vibration isolation system magnetic suspension vibration isolator is verified in the stable condition on the two-layer vibration isolation test platform of the power unit.The results show that the reasonable setting of the stiffness,number and installation position of magnetic suspension vibration isolators can well reduce the dynamic reaction transmitted to the foundation and ensure the stability of the unit.The electromagnetic force error caused by structural deformation can be reduced by improving the structural stiffness of magnetic suspension vibration isolator.The genetic neural network is used to model the dynamic electromagnetic force of the magnetic suspension vibration isolator.The accuracy of the model is higher than that of the method of coefficient correction based on the theoretical formula.It can accurately reflect the nonlinear,hysteresis and dynamic characteristics of the magnetic suspension isolator.The magnetic suspension vibration isolator should adopt the input mode of the same direction current,design the size of the air gap reasonably,avoid the tilt and deviation of the intermediate armature,and reduce the system magnetic leakage as far as possible.Through double degree of freedom vibration isolation system and single control simulation calculation shows that the current of maglev vibration isolator quasi zero stiffness change at any time according to the working condition of sliding mode control and current fixed semi-active control compared to the passive vibration isolation can be effective in reducing force transmissibility,dynamic reaction force and system stability,the impact is not big to the vibration intensity of the unit.Although the semi-active control method has poor stability under variable working conditions,its performance under stable working conditions is similar to that of sliding mode control,and the control method is simpler and the system reliability is better.By adopting the adaptive sliding mode control method,the offset of the intermediate armature can be well estimated,and the strategy of adjusting the incoming current of the upper and lower electromagnets can be adopted to reduce the adverse effect of the armature offset on the vibration isolation system.By applying semi-active control of magnetic suspension vibration isolator to single-layer vibration isolation system,the root mean square value of dynamic reaction force can be reduced to 25.4% and 41.07% of passive control in low frequency resonance and high frequency conditions,and the peak value of dynamic reaction force can be reduced to29.57% and 30.55% of passive control.The semi-active control of magnetic suspension vibration isolator is applied to the two-layer vibration isolation system,and the dynamic reaction force transmitted to the foundation can be reduced by 57.7% and 54.2% at low frequency,and the dynamic reaction force transmitted to the foundation can be reduced by9.6% and the vibration intensity of the unit can be reduced by 46.7% at high frequency.It shows that the magnetic suspension vibration isolation control method applied to the two-layer vibration isolation system is effective.The research results of this paper are of reference significance for the modeling of magnetic suspension vibration isolator.In order to further improve the vibration isolation performance and the stability of the unit under the whole working condition,a new technical route has been developed,which has a good prospect of engineering application. |
PDF Full Text Request