There are a large number of bearings in aircraft engines that are subjected to extreme operating conditions, such as high temperature, high speed, and heavy load, and their fatigue, wear, and other failure problems seriously affect the reliability of the engine. The complex and variable bearing operating conditions can lead to differences in the distribution of data between the source and target operating conditions, as well as insufficient labels. To solve the above challenges, a multi-scale attention mechanism-based domain adversarial neural network strategy for bearing fault diagnosis (MADANN) is proposed and verified using Case Western Reserve University bearing data and PT500mini mechanical bearing data in this paper. First, a multi-scale feature extractor with an attention mechanism is proposed to extract more discriminative multi-scale features of the input signal. Subsequently, the maximum mean discrepancy (MMD) is introduced to measure the difference between the distribution of the target domain and the source domain. Finally, the fault diagnosis process of the rolling is realized by minimizing the loss of the feature classifier, the loss of the MMD distance, and maximizing the loss of the domain discriminator. The verification results indicate that the proposed strategy has stronger learning ability and better diagnosis performance than shallow network, deep network, and commonly used domain adaptive models.

In this paper, a reinforcement learning (RL) based fault tolerant control (FTC) strategy is investigated for the bleed air temperature control system (BATCS) of an aero-engine with time delays, parameter uncertainties and multiple types of faults. By the design of offline training networks, the deep deterministic policy gradient (DDPG) algorithm is employed to update the parameters of training networks. Then a data driven FTC input combined with exponential weighted moving average (EWMA) filter can be obtained to achieve the fault tolerant tracking control of BATCS. Finally, a comparison simulation between PID and RL based FTC scheme is demonstrated to verify the feasibility of the research.

In plastic injection molding (PIM), the process parameters determine the quality and productivity of molded parts. The traditional injection molding process analysis method mainly relies on production experience. It is lack of advanced and rationality and seriously increases production costs. In this paper, a hybrid multiobjective optimization method is proposed to minimize the warpage, volumetric shrinkage and cycle time. The method integrates orthogonal experimental design, numerical simulation, and the metamodel method with multiobjective optimization. The orthogonal experiment chooses seven parameters as the design variables to generate sampling data and determines key factors that affect product quality by the numerical simulation. A gradient-enhanced Kriging (GEK) surrogate model strategy is introduced to construct the response predictors to calculate objective responses in the global design space. Multipopulation differential evolution (MPDE) is conducted to locate the Pareto-optimal solutions, where the response predictors are taken as the fitness functions. This study shows that the proposed GEK-MPDE method can reduce warpage, volumetric shrinkage and cycle time by 5.7 %, 4.7 %, and 18.1 %, respectively. It helps plastic industry to realize collaborative scheduling of multiple tasks between different production lines by providing a low-cost and effective dynamic control method.