Search

An improved multi-swarm migrating birds optimization (IMMBO) algorithm is proposed for hybrid flow shop scheduling with sequence-dependent setup times (HFS-SDST), to minimize the total maximum completion time (i.e., makespan). Permutation-based encoding is adopted to substitute the individual. The modified Nawaz-Enscore-Ham (MNEH) algorithm is employed to generate initial population which are assigned to each sub-swarm according to the makespan. For each sub-swarm, the neighborhood individuals of the leader and followers are generated respectively by performing serial and parallel neighborhood strategies. If the follower is better than the leader according to their makespan, they are exchanged to ensure the information interaction of individuals within the sub-swarm. Moreover, the discrete whale optimization strategy is embedded in IMMBO to optimize the leaders of all sub-swarms to enhance the interaction among them. Furthermore, the local search is designed for the optimal individual to further improve the local search ability of the algorithm. Meanwhile, to avoid algorithm premature convergence, the control strategy for population diversification is designed to the leader of each sub-swarm. Finally, based on adjusting the algorithm parameters experimentally, simulation experiments are conducted on four variants of IMMBO to verify the function of each part by testing an adaptation dataset of Ta. Moreover, the IMMBO is compared with three existing algorithms by testing an adaptation dataset of Ta, and the experimental results demonstrate the effectiveness of the IMMBO algorithm to solve the hybrid flow shop scheduling problem.

Process monitoring plays an essential role in ensuring the safe and efficient operation of multi‐unit plantwide industrial processes. However, complicated nonlinear dynamics of these processes pose enormous challenges to the plantwide process monitoring approaches. In this paper, a novel distributed temporal–spatial neighbourhood enhanced variational autoencoder (DTS‐VAE) monitoring scheme is proposed. The scheme selects the temporal and spatial neighbourhood sets of a current sample from a local unit. Time‐series correlation is utilized to construct temporal representative samples to obtain process dynamic characteristics. Likewise, spatial similarity is employed to construct spatial representative samples to obtain spatial patterns in the data. Then the reconstructed and current samples serve as the inputs for a variational autoencoder to extract features from the current sample and its neighbourhoods. Subsequently, a distributed monitor that considers the temporal–spatial characteristics is established for each unit. Finally, a comprehensive evaluation index is developed using Bayesian fusion strategy to improve monitoring performance. The proposed DTS‐VAE monitoring scheme was effectively verified on the Tennessee Eastman benchmark process and a wastewater treatment plant. [ABSTRACT FROM AUTHOR]

Aimed at the no-idle flow-shop scheduling problem (NIFSP) with minimized makespan, a discrete sine optimization algorithm (DSOA) is proposed. Inspired by sine waveforms, the original sine optimization algorithm (SOA) is a global optimization algorithm, which uses the sine function to update the position of search agents. First, the update position strategy to adapt to the combinatorial optimization problem is redefined. An iterated greedy algorithm with a variable removing size is employed to update the position to enhance the exploration ability. Then, a crossover strategy and a selection strategy are applied to avoid the algorithm falling into local optimum. Next, to improve the exploitation ability of local search and the accuracy of the algorithm, an insertion-based local search scheme is applied in DSOA to search for a better solution around the current optimal solution. Finally, based on the Taillard benchmark, the simulation results of performance comparisons are presented. The experimental results demonstrate the effectiveness of the proposed DSOA algorithm for solving NIFSP.

With the development of modern pressure transmitter detection technology and artificial intelligence technology, stricter requirements are made on the stability, real-time performance and measurement accuracy of pressure transmitter in aerospace, petrochemical, nuclear power and other fields. The temperature of working environment may greatly affect the accuracy of equipment, leading to offset in the measured values of the transmitter. Aiming at these problems, this paper proposes a temperature compensation method for high- precision pressure transmitters based on adaptive DBN (Deep Belief Network). DBN extracts features from data in the unsupervised learning phase, and then fine tunes network parameters using a small amount of data in the supervised phase; Utilizing the BWO (Beluga whale optimization) algorithm to achieve a balance between global search and local optimization, effectively improving the optimization performance of DBN networks; Introducing the Metropolis criterion and fitness balance factor to further improve the algorithm's global search capability and model convergence speed. The experimental results show that the data accuracy after fitting can reach 0.004 8%, much higher than the existing highest standard of 0.05 level. By comparative analysis, the accuracy and practicability of the compensation algorithm are verified. [ABSTRACT FROM AUTHOR]

Bearing faults can seriously reduce industrial production efficiency and even endanger the safety of people's lives and property. Monitoring the operating conditions of bearings conducting fault diagnosis are of great significance for ensuring the safe operation of the production process. This paper proposes a Multi-Scale and Attentive Convolutional Neural Network (MACNN) based on attention mechanism for bearing faults classification. Firstly, one-dimensional bearing signals are input into a convolution layer, followed by a maximum pooling layer to suppress noise and reduce information redundancy. Then, four MACNN modules are taken as input, each of which adopts a parallel network structure of ordinary convolution and void convolution. Under the premise without increasing model parameters, the model's receptive field is expanded to extract more fault features for improving the accuracy. In addition, the attention mechanism module is connected at the end of each MACNN model and the feature information is further extracted by using the ability of automatic extraction of important features. The average pooling layer is used in the network structure to prevent the overfitting and the full connection layer from the output of experimental classification results. Furthermore, the Boundary Equilibrium Generative Adversarial Networks (BEGAN) model is adopted to enhance fault data, change the proportion of unbalanced data sets, increase the number of dataset samples, reduce overfitting of MACNN model, and improve diagnostic accuracy. Finally, the experimental results on the Paderborn University Dataset show that MACNN can achieve better performance in feature extraction and fault classification, outperforming the state-of-the-art methods. [ABSTRACT FROM AUTHOR]