Research And Development Of Intelligent Question Answering System For Simulation Practice Teaching In Thermal Power Plant

With the rapid development of the Internet and artificial intelligence,intelligent question answering system has become a research hotspot because it can provide users with accurate answers and intelligent services.it is gradually entering the field of simulation practice teaching in thermal power plant.In teaching,there are many questions of students and poor efficiency of teachers’ answering questions,so the research on intelligent question answering system of simulation practice teaching in thermal power plant has great application demand and practical value.In this paper,a question answering system with voice interaction and video demonstration function is realized by using Visual C++.First of all,this paper analyzes the significance of intelligent question answering system in thermal power plant simulation training and its research status at home and abroad,and makes a further elaboration on the research significance and demand analysis of voice answering.On the basis of fully comparing the speech recognition technology solutions provided by Google,IFLYTEK and Microsoft,the voice development kit Microsoft Speech SDK suitable for this research system is selected.Focus on the design of SAPI,provided by Microsoft speech engine to realize the functions of speech recognition and speech synthesis.Among them,due to the professional practical knowledge of thermal power plant simulation,speech recognition adopts command recognition mode,and a XML file is written to define the tags that need to be determined by SDK and the words and phrases used to match.In addition,this research system realizes the function of playing demo video with the help of Windows Media Player player.The essence of the question answering system is to provide problem solutions for users.this paper makes an in-depth study of the keyword matching method,which provides a theoretical basis for the knowledge question and answer module.At the same time,some links are designed for the module,such as question analysis,information retrieval,answer extraction and so on.In the part of question analysis,we design the structure of the defined XML file,standardize the sentence patterns of questions,realize the classification of questions,and extract keywords;in the part of information retrieval,we use BM algorithm to retrieve the extracted keywords;because the question and answer database and video database are one-to-one storage of questions,answers and presentation video,so we can extract answers directly.Finally,a question-and-answer test is carried out on the research system.the test results show that when the standardized sentence pattern is used to ask the system,the corresponding answer can be obtained successfully by speech recognition;when the non-standard sentence pattern is used to ask the system,the system will extract the most similar voice information to the students,and the students can ask questions again according to the prompts in order to get a more accurate answer.