PurposeThe soft recoil firing technology is an important way to reduce the recoil force of a gun. To further improve the effect of the recoil mechanism and provide a test prediction method for soft recoil firing, a new multi-port structure of the recoil mechanism with a central port and variable depth wall grooves is proposed to research the working characteristics of a soft recoil gun. MethodsTo describe the firing law of a soft recoil gun, this paper establishes its differential equations of firing motion, constructs the analytical model of each subsystem in the system, and acquires a relatively comprehensive way to describe the firing motion process for this soft recoil system. The flow distribution Bernoulli equation is established with CFD fluid simulation to improve the model accuracy. On this basis, the interval uncertainty theory and genetic algorithm are applied to establish the parameter identification model of the firing process. The typical parameters of the soft recoil gun are identified and obtained with the test data of the first round. ResultsThe similarity rate between the simulation and test data can reach up to 92.78%. Moreover, the corresponding simulation results match the test data under the second and third rounds. It is proved that the correct modeling and effective identification results provide theoretical guidance for the design of a soft recoil gun.