Control Strategy Of Tire-coupled Road Simulator For Waveform Replication

Hydraulic road simulator is one of the key equipme nts in the indoor roadsimulation test, it is used to provide the test vehic le with t he equiva lent vibrationenvironment whe n the test vehic le is driven on the actua l road, in order to testthe durability of the vehic le and its parts. Upon the background of the project todeve lop6-channe l tire-coupled road simulator, which is developed by IEST(Institute of Electro-hydraulic Servo Simulatio n&Test System) of Harb inInstitute of Techno logy. The tire-coupled road simulation platform is driven bymultiple s ingle-channel hydraulic servo mechanis m, every whee l is driven by thesingle servo actuator through the tray and tire. When the entire vehic le is undertest on the road simulator, the interactio ns between channe ls are very obvious,thus the accuracy of affecting wave form reproduction on road simulator iseffected. Base on the coupling characteristics of the system, the control strategyof the tire-coupled road simulator for time-domain waveform replication hasbeen theoretically stud ied.Mathematical model of the single channe l hydraulic system of the roadsimulator is built separately in Simulink. Ana lyze the influence of the carvertica l vibration dyna mic characteristics on system open-loop frequencycharacteristics theoretica lly. On the basis of mathe matica l model, the single-channe l positio n closed-loop and proportiona l controller are designed inSimulink. The characteristic curve of the acceleration response of axle shaftwhich is re lative to the reference inp ut comes out. And it is a lso used as atheoretical model to study the time-doma in waveform reproduction decouplingcontrol a lgorithm.The servo control method alone can’t meet the accuracy required by roadsimulation test, the identification and iterative control method should be applied.FRF(Frequency Response Function) H1estimator is used in the road simulationwaveform replication algorithm to estimate the FRF of the system. It’s wellknown that the H1estimator is influe nced by the no ise in the input s igna ls andgenerates an under-estimation of the true FRF. In consider of the noise both inthe input and output signals, the revised H1estimator is proposed, if inp ut signa lis straightness spectrum and the estimator can be simplified. The FRFidentificatio n simulatio n verified the revised H1estimator is closer to the true FRF than the traditiona l H1estimator.Consider ing the coupling effect of the road simulation system, the FRFestimatio n Matrix at certain frequenc ies or frequenc y bands will be singular, andwhich makes it difficult to operate the inverse solution. In order to solve theinverse proble m, the decoup ling control a lgorithm based on the generalizedinverse theory is proposed, and singular va lue decomposition is used to calc ulatethe pseudo inverse, the decoupling process is comp leted, and the wave formreplication decoupling control algorithm flowchart is give n. Since the numericalalgorithm of singular va lue decomposition is of the high precis ion, when theFRF matrix is nearly singular the inverse solutio n proble m occurs morbid. So thesingular va lue truncatio n method is proposed to ensure the stability o f numericalalgorithm. When there’re fluctuations in the amplitude-frequency characteristicsof the syste m, in particular, the existence of anti-resonance peaks, the estimat ionresult is greater than the true va lue, and the singular va lue correctio n method isproposed to ensure stability and accuracy of iteration.The s imulation s ystem model is built based on the single-channe l syste mtransfer function o f road simulator, and the improved H1estimator and thedecoupling iterative algorithm based on the theory of pseudo inverse are used fortime-doma in wave form replication s imulation ana lys is. The simulatio n resultsshow that when considering the coup ling characteristics of the syste m singularvalue correctio n method and singular value truncatio n method are help ful toguarantee the stability of iterative algorithm and improve the road simulationwaveform replication accuracy.