Abstract: Objective: The realtime performance evaluation of MLM (mediumlow speed maglev) train levitation system based on the data uploaded by levitation controller is not timely enough. To achieve the realtime evaluation of MLM train levitation performance, the development and research work of a MLM train performance testing system is carried out.Method: The development and research work is divided into two parts: hardware development and software design. The hardware part uses a highspeed A/D (analog/digital) conversion chip to collect data from multiroute sensors, and the data is flexibly configured and collected at a sampling frequency through a FPGA (fieldprogrammable gate array). The Kalman filtering method is used to filter the sampled data, and a Wifi module is embedded to perform highspeed packaging and sending of the data collected and processed by FPGA to the software part, so as to meet the timeliness requirements. The software part is developed on the Labview platform, which receives the sensor data uploaded by the Wifi module in realtime, calibrates and filters the data, calculates various types of original sensor data, and finally displays the levitation performance index curves in the interface.Result and Conclusion: The testing results show that the developed MLM train testing system can carry out realtime evaluation of the levitation system performance timely and meet the requirements of MLM system.