Freight train positioning and axle safety monitoring are the key to railway traffic supervision, which can be achieved by sensors based on Internet of Things (IoT) technology, and the energy harvesting technology is the best solution to those sensors due to neither electrical power nor communication system is available on railway wagons. This paper proposes a self-powered railway wagon monitoring sensor (SpRWMS) solution based on the geophone, which has the characteristics of mature technology and simple structure, and can convert mechanical vibration of wagons into electrical energy. Since the vibration is so weak that the voltage induced by the geophone is very weak, hence, this paper proposes an electromagnetic generator (EMG) constructed by geophone matrix and an energy management interface (EMI) to convert the induced potential into the voltage that can drive the circuit system of sensors. In this paper, the implementation and performance of EMI will be focused on, and the energy harvesting system analysis prior to developing prototypes and conducting field trail. The design principles and prototypes of EMG and EMI are proposed based on the parameters of vibration profiles collected by a data logger, and the performance of the SpRWMS prototype is evaluated on a trial run freight train. Experimental results show that the proposed EMI can extract the energy generated by EMG more effectively and increase the voltage across the storage capacitor from 2V to 3.3V, that is, our proposed EMI and EMG can convert the railway wagon vibrational energy into the electrical energy form required by the circuitry of sensors.
The energy generated by a photovoltaic power station is affected by environmental factors, and the prediction of the generating energy would be helpful for power grid scheduling. Recently, many power generation prediction models (PGPM) based on machine learning have been proposed, but few existing methods use the attention mechanism to improve the prediction accuracy of generating energy. In the paper, a PGPM based on the Bi-LSTM model and attention mechanism was proposed. Firstly, the environmental factors with respect to the generating energy were selected through the Pearson coefficient, and then the principle and implementation of the proposed PGPM were detailed. Finally, the performance of the proposed PGPM was evaluated through an actual data set collected from a photovoltaic power station in Suzhou, China. The experimental results showed that the prediction error of proposed PGPM was only 8.6 kWh, and the fitting accuracy was more than 0.99, which is better than existing methods.
One-shot object detection has been a highly demanded yet challenging task since the early age of convolutional neural networks (CNNs). For some newly started projects, a handy network that can learn the target’s pattern using a single picture and automatically decide its architecture is needed. To specifically address a scenario in which a single or multiple targets are standing in relatively stable circumstances with hardly any training data, where the rough location of the target is required, we propose a one-shot simple target detection model that focuses on two main tasks: (1) deciding if the target is in the testing image, and (2) if yes, outputting the target’s location in the image. This model requires no pre-training and decides its architecture automatically; therefore, it could be applied to a newly started target detection project with unconventionally simple targets and few training examples. We also propose an architecture with a non-training parameter-gaining strategy and correlation coefficient-based feedforward and activation functions, as well as easy interpretability, which might provide a perspective on studies in neural networks. We tested this design on the data we collected in our project, the Brown–Yosemite dataset and part of the Mnist dataset. It successfully returned the target area in our project and obtained an IOU of up to 87.04%, reached 80.28% accuracy on the Brown–Yosemite dataset with disposable networks, and obtained an accuracy of up to 89.4% on part of the Mnist dataset in the detection task.