Electronic noses have been widely used in industrial production, food preservation, agricultural product storage, environmental monitoring, and other fields. However, due to the cross-sensitivity of gas-sensing responses, accurately measuring the concentration of mixed gases remains challenging. To address this issue, this study attempts to determine the number of state variables that produce the cross-influence based on the experimental data, establish the state space model from the equivalent circuit model, and obtain model parameters through parameter correlation iterative algorithms and a Kalman filter. The sensor response model and the concentration measurement model of mixed gases are established accordingly. The simulation and experimental results show that these two models have high accuracy in predicting the sensor response and measuring the concentrations of mixed gases under the influence of mixed gases on the sensors.