Solar wireless sensor node energy prediction based on long-short term memory

The energy supply problem of wireless sensor node is one of the important factors affecting the long-term and accurate measurement of the Internet of Things. Converting solar energy into electrical energy for use by the wireless sensor node is a way to solve the long-term operation of the wireless sensor node. Aiming at the characteristics that solar energy is greatly affected by environmental factor, this paper proposes a solar wireless sensor node energy prediction method based on long-short term memory recurrent neural network. The energy usage of the wireless sensor node is reasonably planned through the predicted energy harvesting information to ensure the energy supply stability of the wireless sensor node and the accuracy and reliability of measurement information. The experimental results show that the energy prediction method of the solar wireless sensor node based on LSTM-RNN can use the long-span solar energy to collect historical data, provide accurate wireless sensor node energy prediction result, and ensure the energy supply stability of the wireless sensor node, and the accuracy and reliability of the measurement information. © 2018, Science Press. All right reserved.