The k-nearest neighbor (kNN) method only uses samples’ paired distance to perform fault detection. It can overcome the nonlinearity, multimodality, and non-Gaussianity of process data. However, the nearest neighbors found by kNN on a data set containing a lot of outliers or noises may not be actual or trustworthy neighbors but a kind of pseudo neighbor, which will degrade process monitoring performance. This paper presents a new fault detection scheme using the mutual k-nearest neighbor (MkNN) method to solve this problem. The primary characteristic of our approach is that the calculation of the distance statistics for process monitoring uses MkNN rule instead of kNN. The advantage of the proposed approach is that the influence of outliers in the training data is eliminated, and the fault samples without MkNNs can be directly detected, which improves the performance of fault detection. In addition, the mutual protection phenomenon of outliers is explored. The numerical examples and Tenessee Eastman process illustrate the effectiveness of the proposed method.

It is difficult to correctly classify all faults by using only one classifier, and the performance of most classifiers varies under different conditions. In view of this, a new decision fusion system is proposed to solve the problem of fault classification. The proposed decision fusion system is innovative in two aspects: the use of combined weights and a new improved voting method. The combined weights integrate the subjective and objective weights, where the analytic hierarchy process and entropy weight-technique for order performance by similarity to ideal solution are used to determine the subjective and objective weights of different base classifiers under multiple performance evaluation indicators. Moreover, a new improved voting method based on the concept of classifier validity is proposed to increase the accuracy of the decision system. Finally, the method is validated by the Tennessee Eastman benchmark process, and the classification accuracy of the new method is shown to be improved by more than 5.06% compared to the best base classifier.

Stacked autoencoders have been widely used in soft sensing of industrial process data. However, most of them extract features by minimizing the reconstruction error of the input data at each layer, leading to accumulated loss. Moreover, the extraction of quality-related features is crucial for soft sensing. To address these issues, this paper firstly introduces a novel stacked dual-guided autoencoder (SDGAE) to better characterize intricate data patterns and learn quality-associated features. Compared to SAE, the idea of SDGAE lies in stacking a series of hierarchical dual-guided autoencoder (DGAE), aiming to allow each DGAE to accurately reconstruct the original input data. It also enables the model to simultaneously extract features that are highly correlated with the output variable by introducing a supervisor term. Furthermore, practical scenarios typically involve limited labeled data and a substantial volume of unlabeled data, which renders the network deficient in adequate generalization capabilities. To address this predicament, we further propose a deep semi-supervised stacked dual-guided autoencoder (SSDGAE), it is able to obtain useful feature representations and structural information from unlabelled data in the pre-training phase in addition to the labelled data as SDGAE is able to do. Finally, the proposed methods are verified on two real industrial case for its effectiveness and superiority. The results indicate that our approach can simultaneously utilize both labeled and unlabeled data to enhance the predictive performance of quality variables.

This study explores positioning and obstacle avoidance control for autonomous surface vehicles (ASVs) by considering equivalent elliptical-shaped obstacles. Firstly, compared to most Barrier Lyapunov function (BLF) methods that approximate obstacles as circles, a novel BLF is improved by introducing an elliptical obstacle model. This improvement uses ellipses instead of traditional circles to equivalent obstacles, effectively resolving the issue of excessive conservatism caused by over-expanded areas during the obstacle equivalence process. Secondly, unlike traditional obstacle avoidance approaches based on BLF, to achieve constraint control of angle and angular velocity, a method based on model predictive control (MPC) is introduced to optimize local angle planning. By incorporating angular error constraints, this ensures that the directional error of the ASV remains within a restricted range. Furthermore, an auxiliary function of directional error is introduced into the ASV’s linear velocity, ensuring that the ASV parks and adjusts its direction when the deviation in angle becomes too large. This innovation guarantees the linearization of the ASV system, addressing the complexity of traditional MPC methods when dealing with nonlinear second-order ASV systems. Ultimately, the efficacy of our proposed approach is validated through rigorous experimental simulations conducted on the MATLAB platform.

This article addresses the consensus control problem for a specific class of discrete-time stochastic multiagent systems (MASs). When transmitting the local measurement output to the local controller and neighboring agents, a multiple description coding (MDC) scheme is introduced to reduce the communication burden and enhance data transmission reliability in a resource-constrained environment. The MDC scheme encodes each signal into two descriptions, which are then transmitted through separate channels, and decoding schemes are employed to address the different characteristics of the arrival of the two descriptions, ensuring the boundedness of the decoding error. The proposed consensus control scheme uses the relative decoded measurement errors between local agents and their neighbors. The aim is to design an output-feedback control scheme that ensures the error dynamics of the controlled MAS reach exponentially mean-square boundedness. Sufficient conditions are established for the existence of the controllers through stochastic analysis techniques, and the desired controller gains are parameterized using the feasibility of certain matrix inequalities. The effectiveness of the proposed coding-decoding-based consensus control scheme is verified through a numerical simulation.