Research On The Stability Of Vehicle-mounted Agricultural Fire Sensor

Gas detection technology is widely used in factory production,mining and other industries,and it also plays a great role in monitoring agricultural fires.Using this technology to analyze the carbon monoxide（CO）and carbon dioxide（CO₂）gases produced when combustibles are in the smoldering stage of the fire,the agricultural fire can be accurately predicted.When the system is placed on a vehicle and driving in a complex environment in the wild,the vibration of the driving vehicle and changes in ambient temperature will reduce the stability of the gas detection system,thereby causing errors in the measurement results.Therefore,this thesis aims to improve the stability of the vehicle-mounted agricultural fire sensor,and designs the system’s damping device;and studies the impact of temperature differences on the system,establishes a neural network structure for temperature compensation,and verifies the system through experiments The overall performance.1.The vibration reduction structure of the agricultural fire warning system is designed,the vibration amplitude of the system is calculated by Duhamel integral,and the vibration is simulated by MATLAB software.The integrated vibration sensor was used to conduct vibration tests on various outdoor roads,and gas experiments were carried out.2.This paper designs a temperature compensation algorithm based on neural network.Analyze the impact of ambient temperature on the system,establish a neural network,and verify it through gas experiments.When the outside temperature is 5℃,the CO agricultural fire early warning system is used to detect CO under the background of N₂,and the CO concentration data measured in the low temperature environment and the outdoor temperature are used as the input terminals of the neural network,and the laboratory room temperature is measured The CO concentration data of is used as the output end of the neural network,and the above-mentioned data is used to train the neural network structure.Then,a constant temperature box was used to carry out CO₂ temperature compensation experiments under multi-gradient temperature and multi-gradient concentration.The results showed that the neural network algorithm can effectively compensate for the deviation caused by temperature,thereby improving system stability.3.A simulated combustion experiment was carried out for the agricultural fire warning system.Firstly,the optical path structure and circuit structure of the system are introduced,the absorption peak of CO₂ gas is selected,and the performance indexes of the system are analyzed to verify the stability of the system and the Allan variance.Heating the cotton,paper,and wooden boards commonly found in farmland,it can be seen from a smoldering state to an open flame.Then the gas concentration detected by the system is analyzed,and various stages of the combustion process can be inferred,which verifies the feasibility of the system to prevent fires.Finally,two ways to improve the stability of the system are summarized,and the existing problems and the direction of continued optimization are analyzed.Innovations in this paper:1.Based on the Duhamel integral model and its application in solving the response problem of a vibration system,a vibration reduction device is designed.2.Using neural network algorithm to solve the problem of system temperature drift.