Intelligent testing is a comprehensive application of scientific methods such as computer technology, information technology, and management. The application of speech recognition technology in the field of information technology is becoming increasingly widespread. The main purpose of this system was to achieve the standard mandarin speech testing function. By collecting speech signals and using machine learning algorithms, the required data was calculated. Computer related technologies were used to enhance the recognition ability of standard mandarin, which had a demand advantage in the context of China's continuous development. This paper mainly used the methods of experiment and comparison to analyze the speech recognition of standard mandarin intelligent test system based on particle swarm optimization algorithm. The experimental data showed that in the particle swarm optimization, the recognition accuracy of the speech recognition system for the Shandong mandarin reached 92%. The error rate was 8%, and the time taken was 2.5 seconds.