In recent years, nitrogen oxide emissions are growing rapidly. It is imminent to study and develop a new tenology of gas flue denitrificatio.This article shows that under laboratory condition, simulated flue gas system and used hydrogen peroxide (H2O2) solution to conduct de-NOx experiment, as well as investigated influence of many factors on the de-NOx efficiency and arrived at an optimal de-NOx system. At the same the machanisms were analysed. On this basis, the reseach on dynamics and thermodynamics were carried on to explore the dynamics and thermodynamics theory of this de-NOx system. H2O2 solution has a certain effect on the flue gas denitrification, the removal rate with time is not significant. The results show that the removal of NO by H2O2 / UV system is feasible, and the removal rate of NO is about 70%, and the removal efficiency is much better than that of H2O2 / UV system. Stable. Response surface method optimization experiments obtained regression equation: R=-10.10203*X1- 0.12647*X2-27.68087*X3+0.62404*X4-28071.60614*X5+0.011653*X6+1.66373*X1X3+0.021411*X2X3+2513.87184*X3X5+214.3005, predicted the nitric oxide removal efficiency well. The impact of various factors on the results for Fe catalyst descending> H2O2 concentration> NO initial concentration> H2O2 solution temperature> O2 concentration> flue gas flow. Under optimum conditions the theoretical NO removal efficiency reached up to 89.91%. Mechanics studies shows the apparent reaction order was 2 and thereby established kinetic model was consist with experiment data. The H2O2 / UV system used in low temperature flue gas NO removal has a significant effect. This study provides data and some quotable experience for development of efficient and cleaner flue gas denitration technology.