Research On Odor Detection System Based On Multi-sensor Information Fusion Algorithm

Odor pollution as a special form of air pollution,has attracted enough attention.At present,the detection of odor gas is mainly through the subjective test method of traditional artificial olfaction.With the requirements of scientific law enforcement of environmental supervision,the purpose of odor detection cannot be achieved only through artificial olfaction.Therefore,it is meaningful to design a odor detection instrument that conforms to the human sensory evaluation to replace the artificial olfactory discrimination.This paper first introduces the research status of odor gas detection and the research progress of multi-sensor information fusion technology.Based on the requirements of odor detection,three types of gas sensors,namely electrochemical,photoion and metal oxide,are selected,and temperature sensors are added to monitor the working temperature of the whole system.According to the working principle of different kinds of sensors,the corresponding sensor signal conditioning circuits are designed.According to the system detection requirements,the corresponding hardware circuit and software system are designed.Cross sensitivity is a common characteristic of gas sensors.In this paper,the representative BP neural network in the information fusion algorithm is used to conduct pattern recognition and quantitative analysis on the data collected by the sensor,and the odor activity value addition model is used to convert the odor concentration obtained with the human sensory intensity,and the odor grade evaluation and analysis method is established by combining the sensory evaluation and the odor concentration intensity relationship model.Finally,the performance of the system is tested and analyzed.The experimental results show that the maximum standard deviation of the sensor output signal is 0.0037 V,and the maximum relative standard deviation is 1.12%,indicating that the system has good stability and repeatability.In the experiment,the mixture of hydrogen sulfide,sulfur dioxide and ammonia with different concentrations was used for the experiment.The odor intensity obtained from the five experiments was compared with the results of artificial olfaction.The correlation between the two was 0.953,indicating that the system has high prediction accuracy and can meet the requirements of odor detection.