Offshore platforms have complex structures, high construction costs and harsh operating environments. To prevent serious and fatal accidents, regular/irregular inspections and safety assessments of outdoor platform systems are required during work. Especially for external platforms in the middle and late stages of operation and platforms that are not running, it is more important to perform effective health detection. The purpose of this paper is to study the identification of structural parameters of offshore platforms based on the reality and virtual technology. This paper summarizes the feedback-based linear structure parameter identification methods at home and abroad in recent years, and makes a comprehensive analysis of its application origin, research importance, research status at home and abroad, existing research methods and existing problems. This paper points out that the main problem in the application of output response-based modal parameter identification technology in the field of water engineering is to eliminate the noise and null conditions of measurement signals and signal keys. Denoising and removing spurious noise is a logical model command. Experiments have shown that the recognition results are very close to the real values, and the recognition accuracy will decrease in the presence of noisy signals, and the use of virtual reality technology can improve the accuracy by about 10%.