Receiving Power Prediction in Wireless Power Transfer Systems Using Deep Learning

Wireless power transmission (WPT)and its received power prediction can help improve the efficiency and productivity of wireless energy systems and reduce the challenges in the field of energy management. This study has developed and evaluated the received power prediction model based on LSTM (Long Short-Term Memory) neural network in wireless power transmission systems. Due to the nonlinear complexities and longtime dependencies of the data, it is necessary to use deep learning models, especially LSTM, to simulate and predict the received power in such systems. The proposed model is developed in MATLAB software and evaluated under different conditions including noise and environmental disturbances. The experimental results show that the LSTM model is able to accurately predict the received power. In this simulation, The Root Mean Square Error (RMSE) is 5.81 × 10<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">−3</sup> watts and the maximum error is 1.99 × 10<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">−2</sup> watts. in This high accuracy of the introduced model is able to simulate complex temporal patterns even in the presence of noise and environmental disturbances and provide accurate predictions. Also, optimization of LSTM neural network parameters, including the number of layers, hidden units, and data normalization, has significantly reduced the model error and increased its generalizability. These characteristics have made the LSTM model an effective tool for analyzing and managing received power in wireless power transmission systems. The results of this study can play a significant role in improving the efficiency and stability of wireless power transmission systems and precision-sensitive applications, such as energy management in smart grids and the Internet of Things (IoT).