As a pivotal part of machine driven system, the health states of rolling bearings usually determine the normal operation of the whole equipment. Consequently, it is very necessary to make accurate and timely judgments on the operating conditions of rolling bearings. In this paper, a synthesized diagnosis technology including fault pre-judgment and identification for rolling bearings is raised. In the first part, a threshold value is defined on the basis of the sensitivity of amplitude-aware permutation entropy (AAPE) to bearing faults. Based on this value, whether the bearing has defects is judged. If the defect exists, a feature extraction scheme combining the modified complete ensemble empirical mode decomposition with adaptive noise (MCEEMDAN) and the modified hierarchical amplitude-aware permutation entropy (MHAAPE) is adopted to fully mine the hidden state features. Firstly, the scheme uses MCEEMDAN, which enjoys good time-frequency decomposition capability, to divide the trouble signal into a group of intrinsic mode functions (IMFs). Secondly, the MHAAPE of each IMF component is computed to form the candidate state features. Then, multi cluster feature selection (MCFS) is employed to compress high-dimensional fault features to form low-dimensional sensitive feature vectors required for subsequent classification. Finally, the sensitive feature vectors are input into the random forest (RF) classifier for training and classification, so as to ascertain the different fault type and severity. In addition, different contrastive methods are tested based on experimental data. The experiment results indicate that compared to contrastive methods, the raised scheme enjoys better performance, which can effectively judge whether the bearing is healthy and accurately identify different fault state of bearings.