Damned PWM rectifier

I have focused on the shit PWM rectifier for several days.And now,I got problems down.I simulated my three-phase pwm rectifier model which is based on my Power electronics technology textbook on the Matlab/simulink platform.The principle of this model is very simple.The control aims include outputting a constant DC voltage,and input current must be sinusoidal without phase shift with respect to the input voltage.So PWM rectifier, compared to conventional diode rectifier and phase-controlled recitifer,which is famous for owning unity power factor.It does not pollute the power-supply network.But things seems not that easy when I started to simulate this model on simulink.My matlab version is v2006,this version of matlab is a resource-consuming monster for my celeron(R)1.7Ghz,256M memory machine.It wasted lot of my time starting the software,waiting it free memory etc...What's more,the model I thought is simple enough,will not have any problem.But the simulation results disappointed me deeply.The output voltage can not be controlled.Voltage compensator saturated all the time.The most strange thing is,what I built is a three-phase boost rectifier,but the output voltage just only can be adjusted downward,not upward.It made man mad.Lasting for several gloomy days,I nearly gave it up.I tried to find my answers in those IEEE papers.But not out of my surprise,no one will post articles based on such a simple model using a most basic hysteresis modulation method.Nearly all of the paper have one or two creative point,such as from topology to algorithm.I have no idea then but to turn my help to fundamentals of power electronics ebook.This book the master piece of Colorado power electronics center,which most concerned DC/DC converter.This book is really good,but it does not related much to three phase system.

After last virus attack,I reinstalled my system,and of course,I had to reinstall my matlab.This time I did not install version2006 but version6.5,a lower version but a resource saver.This afternoon,I rebuilt the pwm rectifier in my matlab6.5,after debuged some algebra loop errors,I got my model worked.Out of my surprise,it works so great,and the simulation result is very right!It is hard for me to interpret why it works on Matlab6.5 but not on 2006.Here I got the simulation results screenshots:
The simulation result above is based on the parameters listed below:
Three phase balanced AC voltage source peak amplitude:100V,frequency=50Hz
Input inductance:1e-3H.
Input impedance:1ohm.
Output filter capacitance:1e-3F.
Load resistance:100ohm.
Reference output voltage:400V.
PI controller of voltage loop:kp=0.01,ki=1,output ranges from-100 to 100.
Current hysteresis regulator band:0.01.
PS:The relay submodel should be concerned carefully. It should output zero when switch on and output one when switch off.