Water plays an indispensable role in the gelation of proteins, but its function still remains unclear. In this work, the variation of water species with the structural changes of globular proteins was investigated using temperature-dependent near infrared (NIR) spectroscopy. Ovalbumin (OVA) was used as a model protein, which forms a gel-like structure as the temperature increases through three phases, i.e., phase I (native), phase II (molten globule state), and phase III (gel state). The structural change and the content variation of different water species in the three phases of gelation were analyzed by two-dimensional correlation NIR spectroscopy and Gaussian fitting. A decrease in the water species with two hydrogen bonds (S₂) was found and the change follows the same phases as OVA. In the first two phases, the change occurs after those of other water species but in the third phase, the change is faster than that of free water species. The result indicates that in the native and molten globule states, S₂ is located in the hydration shell of OVA to maintain the stability of the protein structure, and then in the gel state, high temperature weakens the hydrogen bonding of S₂ and leads to the destruction of the hydration shell, making OVA clusters form a gel structure.

The existing Fault Detection and Diagnosis (FDD) model of chillers requires considerable normal and fault data. The acquisition of these data is time-consuming and expensive, and the model is only suitable for special units, which makes it difficult to popularize FDD technology in the operation and management of chillers. At present, a 120-ton chiller has only a small amount of normal and fault data compared with the abundant data of a 200-ton model of the same series. This study investigates the FDD model of a 120-ton chiller and considers similar characteristics of the refrigeration cycle of the same series of chillers. A training set can be created using the 200-ton prior-knowledge data and 120-ton. However, this training set is imbalanced, and the common imbalanced processing Synthetic Minority Oversampling Technique (SMOTE) synthesis mechanism has an overlap problem. This study adopts two adaptive imbalance processing technologies called the Adaptive Synthetic Sampling Approach (ADASYN) and Borderline SMOTE (BSM) that can solve the imbalance problem and SMOTE oversampling overlap problem during knowledge transfer. A support vector machine FDD model with 100%–400% oversampling ratios is established. The best model is ADASYN under 100% oversampling ratio, with a diagnostic accuracy rate of 94.33%.