Quality inspection is essential in preventing defective products from entering the market. Due to the typically low percentage of defective products, it is generally challenging to detect them using algorithms that aim for the overall classification accuracy. To help solve this problem, we propose an ensemble learning classification model, where we employ adaptive boosting (AdaBoost) to cascade multiple backpropagation (BP) neural networks. Furthermore, cost-sensitive (CS) learning is introduced to adjust the loss function of the basic classifier of the BP neural network. For clarity, this model is called a CS-AdaBoost-BP model. To empirically verify its effectiveness, we use data from home appliance production lines from Bosch. We carry out tenfold cross-validation to evaluate and compare the performance between the CS-AdaBoost-BP model and three existing models: BP neural network, BP neural network based on sampling, and AdaBoost-BP. The results show that our proposed model not only performs better than the other models but also significantly improves the ability to identify defective products. Furthermore, based on the mean value of the Youden index, our proposed model has the highest stability.
Single-crystal LiNixCoyMnzO2 (SC-NCM, x+y+z=1) cathodes are renowned for their high structural stability and reduced accumulation of adverse side products during long-term cycling. While advances have been made using SC-NCM cathode materials, careful studies of cathode degradation mechanisms are scarce. Herein, we employed quasi single-crystalline LiNi0.65Co0.15Mn0.20O2 (SC-NCM65) to test the relationship between cycling performance and material degradation for different charge cutoff potentials. The Li/SC-NCM65 cells showed >77 % capacity retention below 4.6 V vs. Li+/Li after 400 cycles and revealed a significant decay to 56 % for 4.7 V cutoff. We demonstrate that the SC-NCM65 degradation is due to accumulation of rock-salt (NiO) species at the particle surface rather than intragranular cracking or side reactions with the electrolyte. The NiO-type layer formation is also responsible for the strongly increased impedance and transition-metal dissolution. Notably, the capacity loss is found to have a linear relationship with the thickness of the rock-salt surface layer. Density functional theory and COMSOL Multiphysics modeling analysis further indicate that the charge-transfer kinetics is decisive, as the lower lithium diffusivity of the NiO phase hinders charge transport from the surface to the bulk. 相似文献
A simple and cost-effective electrochemical method synthesized platinum nanoparticles on graphene nanosheet (PtNPs@GNS) is reported, and the Pt loading of the PtNPs@GNS can be controlled by electrodeposition. The structure and element analysis of the PtNPs@GNS have been investigated by scanning electron microscopy (SEM), Raman spectrum, X-ray diffraction (XRD) and energy dispersive spectroscopy (EDS). The electrochemical measurement including electrochemical active surface area, current density, mass activity, oxidation peak potential,shows the PtNPs@GNS have more performance electrocatalytic properties for methanol oxidation reaction (MOR) compared to Vulcan XC-72 carbon (XC-72) supported PtNPs electrocatalysts. Probably, the cause which may be attributes to no aggregation of PtNPs and the well-dispersion on surface of GNS, so PtNPs@GNS show large electrochemically active surface area, highly electrocatalytic activity and stability in direct methanol fuel cells. 相似文献