Fault diagnosis is crucial in ensuring the safety and efficient operation of industry processes. However, achieving excellent diagnosis performance in traditional fault diagnosis tasks becomes challenging when the training process lacks test fault information. To address this challenge, a novel model for zero-shot industrial process fault diagnosis based on deep semantic canonical correlation embedding (DSCCE) is proposed in this study. The DSCCE model utilizes an embedding reconstruction model to establish the relationship between fault and their semantic attributes, which is accomplished the task through the transfer of fault semantic attribute descriptions. Initially, DSCCE encodes fault samples and semantic attribute descriptions into an embedding space through two embedding networks. Then, the model employs canonical correlation analysis with two embedding features to assess the consistency between semantic attribute vectors and fault samples. Simultaneously, correlation constraints are applied between fault embedding feature subsets to enhance differential information. Subsequently, the classifier is trained using embedding features to complete the task of zero-shot fault diagnosis. Finally, the effectiveness and superiority of the proposed DSCCE model are verified using two case studies.

Automatic welding system is ideally suited for high-volume production of modern manufacturing. The optimization of workpiece mass manufacturing processes on assembly welding lines is a significant practice in the automotive industry to improve production efficiency. Considering the highly coupled characteristics of multiple sub-problems (including task allocation, robot assignment, welding sequence planning, and dual-function robot scheduling) as well as numerous constraints for workpieces and production line configuration, a new unified optimization model for multi-station multi-robot welding system planning and scheduling (MSMRWS-PS) is established. The objectives are to simultaneously minimize the completion time of single workpiece, the completion time difference between adjacent workpieces and the path length of robot movement. After modeling, a multi-objective evolutionary algorithm based on similarity taboo non-dominated sorting rule and individual population density (STNSGA-D) is proposed. A two-layer encoding method is designed to decompose the coupled problem of welding task allocation and welding sequence planning. In cooperation with the two-layer encoding, an encoding scheme combining dominant coding and hidden coding are devised for robot assignment and dual-function robot scheduling. Finally, STNSGA-D is compared with four state-of-the-art multi-objective evolutionary algorithms (MOEAs) on a set of workpiece test instances. The experimental results show that the comprehensive performance of STNSGA-D is superior to the comparison algorithms. The model and solution proposed in this paper can improve the efficiency of mass production in factories, and has important practical significance for optimizing the high-coupling welding systemsNote to Practitioners—This work is motivated by the actual industrial scene of a welding system for assembling, welding and transferring workpieces. Specifically, it pertains to an assembly welding line comprising assembly welding stations, transfer station, unloading station, stationary welding machines and robots. Notably, this line incorporates two dual-function robots that can independently carry out welding and transfer operations, owing to its special configuration. It possesses the ability to weld one workpiece while clamping another. To coordinate the welding and transfer tasks, reduce the waiting time of robots, and enhance the production efficiency, it is imperative to schedule the operations of the dual-function robots. In addition, the reachable space of both dual-function robots encompasses the same station, which means that one of the robots can weld independently, or both robots can weld cooperatively. As a result, the scheduling of dual-function robots must be conducted concurrently with the robot assignment for station. It is also necessary to assign the welding task for stations and plan the welding sequence for robots. These optimization problems are highly coupled to each other, and considering only one of the sub-problem cannot optimize the entire production line. Hence, we present a comprehensive MSMRWS-S-PS model which ignores collisions and propose a corresponding algorithm tailored to effectively optimize the welding system.

Automatic Relevance Determination (ARD) priors have been widely used to induce sparse reconstructions in Bayesian compressive sensing approaches. In this paper, we propose a new sparsity-promoting prior coined as Double Lomax prior. Its connection with the generalized inverse Gaussian distribution and Rayleigh distribution leads to a tractable full Variational Bayesian (VB) inference procedure here. It is shown that the proposed update procedure includes the canonical ARD update procedure as a special case, but provides a better global convergence performance and results in improved signal reconstructions.

This paper develops a fuzzy chance constrained mixed-integer nonlinear programming (FCC-MINLP) model and the solution methods for refinery short-term crude oil scheduling problem under demands uncertainty. To reduce the calculation complexity of the model, it is transformed into its equivalent fuzzy chance constrained mixed-integer linear programming (FCC-MILP) model by using the method of Quesada & Grossmann (1995). After that the FCC-MILP model is solved through its crisp equivalent algorithm and fuzzy simulation algorithm rely on the theory presented by Liu & Iwamura (B. Liu, K. Iwamura, 1998) for the first time in this area. Finally, a case study which has 265 continuous variables, 68 binary variables and 318 constraints is effectively solved in LINGO 8.0 (J. X. Xie and Y. Xue, 2005) with the proposed approaches