With the increase of economic globalization and international trade, waterway transportation is increasingly prosperous, and due to the extreme complexity and unknown of the maritime navigation environment, all kinds of ships are prone to collision avoidance accidents when sailing in limited waters, so the development of the ship intelligent collision avoidance system is necessary. Starting from the decision-making and key methods of ship intelligent collision avoidance system, the development status of ship intelligent collision avoidance technology is expounded, some achievements and shortcomings in this field in recent years are summarized, and the characteristics and existing defects of intelligent collision avoidance algorithms such as path planning and machine learning are analyzed. In the future, it is necessary to continuously improve and enhance the collision avoidance technology of ships, strengthen the self-learning ability of the collision avoidance system to reduce ship collision accidents from the source, protect the safety of life at sea and the clean marine environment, and promote the long-term sustainable development of the entire shipping industry.