Fault-Level Grading of Photovoltaic Cells Employing Lightweight Deep Learning Models.

The deployment of photovoltaic (PV) cells as a renewable energy resource has been boosted recently, which enhanced the need to develop an automatic and swift fault detection system for PV cells. Prior to isolation for repair or replacement, it is critical to judge the level of the fault that occurred in the PV cell. The aim of this research study is the fault-level grading of PV cells employing deep neural network models. The experiment is carried out using a publically available dataset of 2,624 electroluminescence images of PV cells, which are labeled with four distinct defect probabilities defined as the defect levels. The deep architectures of the classical artificial neural networks are developed while employing hand-crafted texture features extracted from the EL image data. Moreover, optimized architectures of the convolutional neural network are developed with a specific emphasis on lightweight models for real-time processing. The experiments are performed for two-way binary classification and multiclass classification. For the first binary categorization, the proposed CNN model outperformed the state-of-the-art solution with a margin of 1.3% in accuracy with a significant 50% less computational complexity. In the second binary classification task, the CPU-based proposed model outperformed the GPU-based solution with a margin of 0.9% accuracy with an 8× lighter architecture. Finally, the multiclass categorization of PV cells is performed and the state-of-the-art results with 83.5% accuracy are achieved. The proposed models offer a lightweight, efficient, and computationally cheaper CPU-based solution for the real-time fault-level categorization of PV cells.

Fault-tolerant scheme for robotic manipulator-Nonlinear robust back-stepping control with friction compensation.

Emerging applications of autonomous robots requiring stability and reliability cannot afford component failure to achieve operational objectives. Hence, identification and countermeasure of a fault is of utmost importance in mechatronics community. This research proposes a Fault-tolerant control (FTC) for a robot manipulator, which is based on a hybrid control scheme that uses an observer as well as a hardware redundancy strategy to improve the performance and efficiency in the presence of actuator and sensor faults. Considering a five Degree of Freedom (DoF) robotic manipulator, a dynamic LuGre friction model is derived which forms the basis for design of control law. For actuator's and sensor's FTC, an adaptive back-stepping methodology is used for fault estimation and the nominal control law is used for the controller reconfiguration and observer is designed. Fault detection is accomplished by comparing the actual and observed states, pursued by fault tolerant method using redundant sensors. The results affirm the effectiveness of the proposed FTC strategy with model-based friction compensation. Improved tracking performance as well robustness in the presence of friction and fault demonstrate the efficiency of the proposed control approach.

Oral cancer stem cells drive tumourigenesis through activation of stromal fibroblasts.

BACKGROUND: Cancer stem cells are responsible for tumour progression and chemoresistance. Fibroblasts surrounding a tumour also promote progression and fibroblast "activation" is an independent prognostic marker in oral cancer. Cancer stem cells may therefore promote tumourigenesis through communication with stromal fibroblasts. METHODS: Cancer stem cells were isolated from oral cancer cell lines by adherence to fibronectin or cisplatin resistance. Fibroblasts were exposed to conditioned medium from these cells, and the activation markers, alpha smooth muscle actin and interleukin-6, were assessed using qPCR and immunofluorescence. Stem cell markers and smooth muscle actin were examined in oral cancer tissue using immunohistochemistry. RESULTS: Adherent and chemoresistant cells expressed increased levels of stem cell markers CD24, CD44 and CD29 compared with unsorted cells. Adherent cells exhibited lower growth rate, higher colony forming efficiency and increased cisplatin resistance than unsorted cells. Smooth muscle actin and Interleukin-6 expression were increased in fibroblasts exposed to conditioned medium. In oral cancer tissue, there was a positive correlation between expression of αSMA and stem cell markers. CONCLUSIONS: Adherence to fibronectin and chemoresistance isolates stem-like cells that can activate fibroblasts, which together with a correlation between markers of both in vivo, provides a mechanism by which such cells drive tumourigenesis.

Multipath estimation in urban environments from joint GNSS receivers and LiDAR sensors.

In this paper, multipath error on Global Navigation Satellite System (GNSS) signals in urban environments is characterized with the help of Light Detection and Ranging (LiDAR) measurements. For this purpose, LiDAR equipment and Global Positioning System (GPS) receiver implementing a multipath estimating architecture were used to collect data in an urban environment. This paper demonstrates how GPS and LiDAR measurements can be jointly used to model the environment and obtain robust receivers. Multipath amplitude and delay are estimated by means of LiDAR feature extraction and multipath mitigation architecture. The results show the feasibility of integrating the information provided by LiDAR sensors and GNSS receivers for multipath mitigation.