The research on the simulation and verification technology of equipment maintenance characteristics based on big data is aimed at the possible failures and their causes in the actual operation training process of the technology-intensive complex system integrating mechanical, electronic, hydraulic, information and control technologies. The failure mode and impact analysis method (FMEA) is adopted for large equipment. This paper studies the electromechanical and electro-hydraulic faults in large weapon equipment, analyzes the basic properties and propagation characteristics of the faults, establishes the fault attribute description model database, carries on the fault propagation simulation analysis by using neural network. This study makes full use of the characteristics of big data virtuality, increases the number of repeated verification, reduces unnecessary costs and saves costs. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.