The Application And Study Of Cat In The Virtual Maintenance Training System

This paper is about Computerized Adaptive Testing based on Virtual Maintenance Training, Computerized Adaptive Testing has become a hot research field in recent years. The system plays an important role in the self-examination on getting a comprehensive assessment of the ability of candidates. The research in this paper has been done is based on the existing development on the adaptive test. The research work of this paper can be divided into the following parts.1. Proposed a new method to estimate the ability which is called Difficulty ratio logarithmic method. Difficult coefficient is involved in this algorithm which is based on the original logarithmic method. This method can be used in the ability exploration stage of Computerized Adaptive Testing System to estimate the ability of the candidates. Compared with the original method, the new one has a higher precision, besides it needs fewer topics to be tested.2. Proposed a new method based on the original step method, we call it Dichotomous Indefinite Step Method. Compared with the fixed-step method this method is more flexible. The step size can be adjusted appropriately according to the number of questions, and we do some research compared to the fixed-step method. Then we design experiments to verify the feasibility of the algorithm also improved the defects of the fixed-step method.3. Architecture and code implementation of Computerized Adaptive Testing System based on Maintenance Training. We designed an Computerized adaptive testing system based on virtual maintenance training according to IRT theory and system requirements. We gives the key technology algorithms, database code development, each module development and implementation process, showing the front-end interface of the system realization in this part4. Establishment and maintenance of question bank of Maintenance Training System. We design experiments to estimate the project parameters of questions in question bank, including the difficult coefficient, discrimination, guess coefficient. Questions are stored in the database to build the question bank. And do dynamic maintenance on the question bank.5. We use the Adaptive Testing System during Virtual Maintenance Training process in our experiment, it proves that the Adaptive Testing System has a high reliability. During Virtual Maintenance Training process, we recorded the time that the subjects used in the experiment to complete the task and the adaptive testing results, then we compared the results of our experiment to verify the reliability of Adaptive Testing System. According to the results we recorded and organized from the subjects’subjective questionnaires, proving that the Adaptive Testing System will not bring psychological pressure to the subjects. In short, we achieved the desired effect.The improvements and additions to the original algorithm is that we proposed Difficulty ratio logarithmic method and Dichotomous Indefinite Step Method in the ability exploration stage. These two methods make the Computerized Adaptive Testing System more efficient and has higher accuracy in the initial ability estimate stage. Meanwhile, we introduce the difficult coefficient of the questions into the formula. Taking the influence of the questions with various difficult coefficient in to consideration to improve the accuracy of the test results.