Search results

To meet the requirement of establishing an effective schedule for the assembly process with overall detection and rework, this paper aims to address a new problem named resource-constrained multi-project scheduling problem based on detection and rework (RCMPSP-DR).

First, to satisfy both online and offline scheduling, a mixed integer programming model is established with a weighted bi-objective minimizing the expected makespan and the solution robustness. Second, an algorithm that combines a tabu search framework with a critical chain-based baseline generation scheme is designed. The tabu search framework focuses on searching for a reasonable resource flow representing the execution sequence of activities, while the critical chain-based baseline generation scheme establishes a buffered baseline schedule by estimating the tradeoff between two aspects of bi-objective.

The proposed algorithm can get solutions with gaps from −4.45% to 2.33% when compared with those obtained by the commercial MIP solver CPLEX. Moreover, the algorithm outperforms four other algorithms in terms of both objective performance and stability over instances with different weighting parameters, which reveals its effectiveness.

The represented RCMPSP-DR considering the overall detection and rework is an extension of the scheduling problem for large-scale equipment. An effective algorithm is proposed to establish the baseline schedule and determine the execution sequence of activities for the assembly process, which is significant for practical engineering applications.

In gearbox fault diagnosis, identifying the fault type and severity simultaneously, as well as the compound fault containing multiple faults, is necessary.

To diagnose multiple faults simultaneously, this paper proposes a multichannel and multi-task convolutional neural network (MCMT-CNN) model.

Experiments were conducted on a bearing dataset containing different fault types and severities and a gearbox compound fault dataset. The experimental results show that MCMT-CNN can effectively extract features of different tasks from vibration signals, with a diagnosis accuracy of more than 97%.

Vibration signals at different positions and in different directions are taken as the MC inputs to ensure the integrity of the fault features. Fault labels are established to retain and distinguish the unique features of different tasks. In MCMT-CNN, multiple task branches can connect and share all neurons in the hidden layer, thus enabling multiple tasks to share information.