Automatic Speech Disentanglement for Voice Conversion using Rank Module and Speech Augmentation