Search

Accurate medium and long-term load forecasting plays an important role in the planning, construction and operation of power system. In this paper, a medium and long term electricity demand forecasting method of different industries based on deep belief network is put forward. In order to cope with many complicated factors in medium and long term electricity demand forecasting, the proposed method takes advantage of powerful function modeling ability of deep belief network and selects a number of quantifiable main influencing factors as input variables, and establishes a medium and long term electricity demand forecasting model based on deep belief network for each load types classified of different industries. Simulation example shows that the proposed method has higher prediction accuracy than the traditional neural network algorithm.

Deep learning method has been considered as a breakthrough in computer vision, successfully aplied in many domains, including biometrics. Palmprint recognition has been accepted with high acceptability and low intrusion. In this study, deep learning was introduced into palmprint recognition for a better performance. Three concrete steps were involved in the application. First, a deep belief net was built by top-to-down unsupervised training with training samples. Second, the optimum parameters were chosen to adapt the model for a robust performance. Third, the testing samples were labeled by employing the deep learning models. Compared with traditional recognition methods, such as PCA, LBP, the experimental results show that deep learning method has a higher recognition rate for palmprint recognition.