Nowadays, the energy problem is becoming more and more serious. The high pollution and low storage of traditional energy make it unsuitable for a sustainable supply of energy. Solar energy is renewable clean energy is praised by countries all over the world because of its clean, easy-to-use, safe and reliable advantages. As photovoltaic power generation utilizing solar energy converted into electricity, low efficiency of photovoltaic power generation is a technical problem restricting the development of solar energy. How to effectively improve the efficiency of photovoltaic power generation has become an important research direction. Due to the nonlinear output characteristics of photovoltaic cells, when the environment changes, the output power also changes. Therefore, in order to increase the efficiency of solar energy utilization and make the system work at its optimum, the Maximum Power Point Tracking (MPPT) technology to use. In MPPT, it can generally be divided into two categories: single-peak maximum power point tracking and multi-peak maximum power tracking. In the single-peak maximum power tracking, it is usually divided into constant voltage method, perturbation and observation method, incremental conductance method and optimal gradient method. In the multi-peak maximum power point tracking, it is usually divided into intelligent control group algorithm, neural network algorithm based on big data and compound algorithm. This paper briefly describes the improved particle swarm optimization. The article builds a foundation for more effective utilization of photovoltaic power generation in the future by analyzing the methods of single-peak multi-peak briefly.