In recent years, deep learning methods have gained much success in the area of intelligent fault diagnosis. However, due to the fact that the working condition is various and the noise is inevitable, degradation of previous model is very serious. To address the challenge of bearing fault detection under strong noise environment, this paper proposed a novel anti-noise deep residual multiscale convolutional neural network with attention mechanism named Attention-MSCNN. First, dynamic dropout is used to improve the anti-noise ability by introducing artificial noise into training process. Additionally, we design a residual connection between input and the convolved features to fully capture the characteristics of the initial input. Finally, a novel denoised multi-head attention mechanism is applied to remove excess noise in raw input and obtain relationships between long time series. The experimental results show that Attention-MSCNN can achieve robust anti strong noise performance with over 85% accuracy on CWRU dataset. On the self collected two-stage gear drive test bench, our model achieve an accuracy of over 99% under strong noise environment. Thus, Attention-MSCNN successfully solve the problem of low detection accuracy of previous models under strong noise environment.

We study the robust finite‐time leader–follower output consensus problems for second‐order multiagent systems (MASs) with mismatched disturbances. For mismatched disturbance, the mismatched disturbance observer is used for each follower to estimate the mismatched disturbance in finite time. Then a backstepping control protocol is constructed via Lyapunov finite‐time stability theorem, and we can prove that the output consensus tracking with strong robustness can be reached for the considered MAS in finite time if the communication topology includes a directed spanning tree, where the leader is the root. Finally, we use two simulation examples to illustrate the correctness of the theoretical results.

This paper addresses H _∞ synchronization for uncertain chaotic systems with one-sided Lipschitz nonlinearity under the output and intrinsic state delays. By utilizing the one-sided Lipschitz condition and quadratic inner boundedness, constructing an appropriate Lyapunov-Krasovskii (LKF), robust controller design conditions based on Lyapunov stability theory are derived for synchronization of chaotic systems under disturbances or perturbations bounded by L ₂ norm. By introducing the delay-derivative limits and delay-interval bounds into LKF, the intrinsic state time-varying delay can be tackled by the delayrange-dependent strategy. Less conservative stability condition can be obtained by the further improved inequality of Jensen inequality and reciprocally convex approach, which can lead to the tighter upper bound for integral inequality. Numerical simulations are provided to verify the validity of the proposed methodology for synchronization of chaotic systems.

This paper researches observer-based $H_{\infty }$ synchronization of coronary artery system with input, output time-delays, and external disturbance. By designing chaotic observers for both the master system and slave system, the estimation of the states for master-slave systems has been accomplished. Based on the free-matrix-based integral inequality and Wirtinger-based inequality, the Lyapunov Krasovskii functional (LKF) is constructed to obtain a less conservative synchronization strategy for the chaotic coronary artery system with the existence of state estimation errors and synchronization error. The effectiveness of this control strategy has been demonstrated by some simulations.