An improved Event air collision model is established to calculate CA(Conventional Aircraft)/RPA(Remotely Piloted Aircraft) integrated operation intervals. This model considers the data link delay of RPA, communication, navigation and surveillance errors. In order to investigate the problem of integrated operation of CA/RPA terminal area, a method of arrival sequencing and scheduling in terminal area is established. A Non-dominated Sorting Genetic Algorithms-II based on RHC(Receding Horizontal Control) with the elite retention strategy is then presented by introducing the adaptive strategy to calculate the optimal solution for approach flight sequencing and scheduling. Light, medium and heavy aircrafts are selected as combinations for the research object to solve the longitudinal safe operation intervals that satisfy the safety target level proposed by ICAO. Finally, the actual operation dataset of an airport is used to implement case simulation and verification, and comparison is made with the primitive NSGA-II. The simulation results show that the improved NSGA-II method has superior algorithmic performance and more efficient solution compared to the primitive NSGA-II algorithm; for the large-scale dataset, comparing the manned/unmanned operation with 25% mixing degree and the CA/RPA operation with 50% mixing degree, the total delay time and the average delay time are both improved by 0.6% compared to the traditional manned operation, and the total flight time and the maximum flight and maximum flight time were reduced by 1.32% and 0.96%, respectively.