Based on the semiconductor gas sensor using ZnO

2

as the sensitive material

2 × 2 and 2 × 3 hazardous chemical gas sensor arrays are constructed

and the hazardous chemical gas detection platform is built. The mixed gas composed of CO and CH

4

is detected

and the fusion model is established for quantitative analysis of the mixed gas. By calibrating the gas sensor

it is found that there is a nonlinear relationship between the response of the gas sensor and the gas concentration. The generalized dynamic fuzzy neural network（ GD-FNN） which can determine the fuzzy rules by self-learning is selected as the quantitative analysis tool. GD-F

NN algorithm is used to construct the data fusion model of 2 × 2 array and 2 × 3 array

set the initial parameters and conduct network training.Finally

GD-FNN algorithm fusion model is used to conduct quantitative analysis on the test sample data of mixed hazardous chemical gases. The results show that the fusion accuracy of the two array data fusion models is similar

the maximum relative error is 10. 63 %

the average precision can reach more than 90 %

the reliability is strong

and it has a good reference value.