In the aircraft development process, requirement is the most important part that must be identified and traced carefully. Since the aircraft is a typical complicated system, thousands of requirements are involved in the development process. To enable fast and accurate extraction of requirements from requirement documents, we apply an automated approach to extract requirements for aircraft development process from the text, such as design requirement documents and airworthiness certification regulations concerned with design. The method uses the python regular expressions to extract the statements that match the defined patterns from the document as requirement candidates. And then, topic modeling technique is performed to group the different requirement candidates which reduces the labor consumption of later reviewing. The result shows 81–91% of the requirements in the documents can be extracted correctly.