Longxikou Navigation-power Protection Embankment project protects the lives and property of the people along the coast and maintains the safe and normal operation of the Longxikou Junction project.Therefore,its construction quality is important.The traditional quality inspection and control of the rolling process of the navigation power protection embankment project uses the test pit experiment to detect the degree of compaction,and it is difficult to achieve real-time control of the whole process.In order to ensure the construction quality,this paper uses global navigation satellite systems (GNSS) and artificial intelligence technology to develop a rolling process monitoring system for the Longxikou Protection Embankment Project,which realizes real-time monitoring of rolling trajectory,rolling speed,exciting force state,and number of roller passes.This paper uses the kernel extreme learning machine (ELM) algorithm optimized by the gray wolf to propose a compaction quality evaluation model and analyze the compaction quality in real time.Combined with the actual engineering,by comparing the average absolute error,root mean square error,and correlation coefficient,the accuracy of the new compaction quality evaluation model is higher than that of the traditional back propagation (BP) neural network,support vector machine (SVR),and extreme learning machine (ELM) models.