Assessing the comprehensive cognitive state (CS) levels of pilots during aviation operational missions is crucial for ensuring flight safety and optimizing task performance. Traditional questionnaire-based assessment methods have limitations in capturing real-time multifaceted cognitive processes, leading to granularity in classification levels and fail to accommodate individual differences. Accordingly, this study quantifies pilots’ cognitions into a comprehensive five-level state based on cognitive processes (i.e., information filtering, cognitive matching, and decision-making). More specifically, the proposed assessment model incorporates the NASA-TLX scale and cognitive behavior patterns, while considering cognitive and time resources parameters tailored for aviation operational missions. Extensive operational experiments under aviation context are conducted among 10 subjects through the proposed simulation platform, reaching an accuracy of 86.5% through the back-propagation network. Results indicate the proposed CS assessment model can effectively measure the CS of different pilots in a real-time base, thereby enhancing the efficiency and safety of human-machine collaboration in aviation operational missions.