Aiming at the development and research status of the Maximum Power Point Tracking (MPPT) technology of solar photovoltaic power, a MPPT algorithm based on Radical Basis Function (RBF) neural network is proposed. Firstly, this paper establishes a parameter model of the solar cell and analyzes the main influencing factors of photovoltaic power generation, selects the voltage and current of the panel as the input layer of the RBF neural network, and directly adjusts the duty of the Boost converter to achieve the maximum power point tracking. Compared with the traditional Perturbation and Observation method (P＆O), this method avoids the difficulty of setting the step size. The duty cycle of the Boost converter is adjusted directly for maximum power point tracking through the RBF neural network. Simulation and experimental results show that the provided MPPT algorithm can track quickly and more efficient than the traditional P&O algorithm. This work is supported by National Natural Science Foundation of China (No. 41727801) and Graduates Innovation Fund of Wuhan University of Engineering (No. CX2018116).