The battery industry has been growing fast because of strong demand from electric vehicle and power storage applications.Laser welding is a key process in battery manufacturing. To control the production quality, the industry has a great desire for defect inspection of automated laser welding. Recently, Convolutional Neural Networks (CNNs) have been applied with great success for detection, recognition, and classification. In this paper, using transfer learning theory and pre-training approach in Visual Geometry Group (VGG) model, we proposed the optimized VGG model to improve the efficiency of defect classification. Our model was applied on an industrial computer with images taken from a battery manufacturing production line and achieved a testing accuracy of 99.87%. The main contributions of this study are as follows: (1) Proved that the optimized VGG model, which was trained on a large image database, can be used for the defect classification of laser welding. (2) Demonstrated that the pre-trained VGG model has small model size, lower fault positive rate, shorter training time, and prediction time; so, it is more suitable for quality inspection in an industrial environment. Additionally, we visualized the convolutional layer and max-pooling layer to make it easy to view and optimize the model.
In this study, we have investigated the evolution process and dynamic characteristics of a multi-pulse regime in an erbium-doped fiber ring laser based on a single-mode fiber–graded-index multimode fiber–single-mode fiber (SMF-GIMF-SMF) structure as an optical modulator. By utilizing the excellent nonlinear optical absorption of the SMF-GIMF-SMF (SMS) device with a modulation depth of ~8.68%, stable single-pulse mode locking at the frequency of 9.84 MHz can be readily observed at low pump power. In addition, the single-pulse operation can evolve into a multiple-pulse regime on account of the peak-power-clamping effect via suitably raising the pump power and carefully regulating the polarization state. Further, the single-shot temporal evolution of multiple pulses is monitored, indicating that this state shows unique and interesting temporal characteristics with variable pulse separations and inconsistent pulse intensities, which, as far as we know, is the first such observation in ultrafast fiber lasers. Additionally, this study, based on the time-stretch dispersive Fourier transformation method, suggests that these multiple pulses consist of chaotic wave envelopes with erratic intensities and changeable pulse energy. We believe that these findings have profound implications for revealing fascinating nonlinear pulse dynamics in ultrafast fiber optics.