Yanjun Liu

The core of ocean thermal energy conversion (OTEC) is the transfer and conversion of heat energy, and the heat exchanger is a key component of the heat transfer between the surface warm seawater and the lower cold seawater. The working fluid has a significant impact on the efficiency of the entire cycle in the temperature difference cycle. This study aimed to improve heat exchange efficiency. The article studied heat exchangers, used R134a as the circulating medium, and applied ANSYS-FLUENT 2020R2 simulation software to analyze the variation in heat transfer coefficients. We obtained the trend in the heat transfer coefficient of the heat exchanger with the shape of an elliptical tube under the condition of ocean temperature difference cycles. Then we used a long short-term memory network and Adam optimization algorithm to establish the prediction model. The NSGA-II 11 algorithm was used to realize optimization objectives of the highest heat transfer efficiency and the smallest cross-sectional area of heat transfer tubes along the X and Y directions. Finally, the parameters of the evaporator and condenser ultimately resulted in three optimal solutions. The results of this study can provide a certain theoretical basis and reference value for the efficiency analysis, structure optimization, and experimental research of the subsequent ocean differential circulation heat transfer.

The reliability and continuity of data are key issues in deep-sea sediment cone penetration testing. Cone penetration testing employs static force to uniformly insert rods into sediment, a process crucial for assessing its mechanics and layering. The clamping manipulator can perform this operation while accommodating sediment sensors of varying types and sizes. However, its requirement to reset post-penetration creates zero-velocity points that diminish test continuity and should be minimized. To address these limitations, this paper proposes a load-adaptive sediment rig that minimizes zero-velocity points, ensures data continuity, and contributes to sedimentology research. This paper analyzes the mechanical properties and layering patterns of sediment, along with the interaction mechanisms between sediment and mechanical structures. Subsequently, a mechanical structure–sediment-integrated model with adaptive control logic is established. Finally, real sediment data are introduced into the physical model for simulation experiments. The simulation results demonstrate that the load-adaptive rig reduces data breakpoints by 50% and increases the maximum single penetration stroke to 1.8 m. Additionally, the load-adaptive rig provides redundancy between penetration force and stroke, automatically reducing penetration force for greater stroke when encountering low-strength sediments and, conversely, sacrificing part of the stroke for greater force. These improvements significantly enhance the continuity of in situ detection data of sediment.

Mini Marine Environment Chamber System (Mini-MECS) is a mesoscale marine climate and environment simulation experimental device with the parameter control function. Its main equipment is three large experiment chambers with a diameter of 3.2m and a height of 12.5m. It can be used to carry out large-scale marine environmental ecology scientific experiments, which can not only meet the huge demand for the marine mesoscale vertical process simulation, but also ensure the number of parallel comparisons required for scientific verification. Lots of engineering techniques have been applied to construct this large-scale experiment equipment, for example structural design, ocean temperature field simulation and control and multi-parameter joint measurement. This paper introduces its construction process, technologies and experiment applications.

Deep learning based intelligent fault detection for mechanical equipment has become an important research fields. However, due to various equipment and working conditions, it is difficult to collect sufficient fault samples by monitoring sensors, which restricts the accuracy of existing intelligent fault diagnostic approaches. In this paper, we propose a novel generative adversarial networks (GAN) based on local weights-shared multi-generator for the generation of fault data and building a fault diagnosis model. Unlike existing GAN based data augmentation methods, the proposed model contains multiple generators which are divided into several groups with each groups’ prior layers of generators weights-shared. By adversarial training, the discriminator gradually improves the performance of judging authenticity and fault categories with each group of generators acquiring the ability to generate a type of fault data. In this way, the proposed local shared multiple-generator GAN method effectively reduces mode collapse and generates more diverse fault data. The performance of the proposed method was validated on the fault datasets that are not applied in training stage. Compared with traditional deep learning diagnosis methods, the accuracy of fault diagnosis could be improved up to 30%, which indicates that the data generation model can generate diverse data and expand the fault data space.

Intelligent fault diagnosis is an important subject of mechanical system maintenance. Domain adaptation is a method to solve the problem that the model trained on the training set (source domain) is not suitable for the test set (target domain) due to different working conditions in fault diagnosis. In industrial scenarios, there may be multiple-source domains. For this reason, we proposed an intelligent fault diagnosis system (IFDS) with multisource unsupervised domain adaptive network that adapts to single- or multiple-source domains. The proposed method considers the differences between sources and uses source domain data and a small amount of unlabeled target domain data to mine the feature information contained in the data. IFDS uses a feature extractor to learn the feature representations, constructs a domain discriminator for each source domain, and learns domain-invariant features through adversarial training to diagnose target domain faults. The validity of the method in fault diagnosis is verified by various transfer tasks of two public fault-bearing datasets.