Power load forecasting based on wireless sensor networks

Abstract

At present, wireless sensor networks (WSN) technology is a field of much research interest in the area of information technology. The application of wireless sensor networks is expected to be very broad. With the development of communication protocol and their corresponding components, wireless sensor networks technology plays an increasingly important role in the power industry. The analysis and forecast of electric power data is essential to the construction and operation of the power network. Through the wireless sensor networks, we can obtain comprehensive power data. Then we can better forecast the power load through the data we obtain from the wireless sensor networks. In this paper, we propose an improved LSSVM method. We collect the power data by the wireless sensor networks and use the improved LSSVM method to forecast the power load. Experimental results demonstrate the effectiveness of the proposed method.