Research On Forest Fire Estimation And Monitoring System Based On Multi-factor Analysis

In recent years,various reasons like climate warming have led to the increase of,extreme weather and higher frequency of various natural disasters,among which the risk of forest fire has increased.In addition to bringing serious economic losses and,endangering our life safety,forest fire can also cause damage to the ecological environment.Therefore,the prevention work of forest fire is of great significance.Through investigation data review and other channels,it can be seen,that there are few studies on fire prevention in forest scenic areas and forest parks in China.Since state-owned forest farms and large forest areas are generally inaccessible,there are few traces of human activities,resulting in low possibility of man-made fire Generally,the forest fires in these areas are caused by the natural environment,which proposes relatively lower real-time monitoring requirements Thus,simple equipment like satellites and drones can meet the requirements.This dissertation aims to study those areas with large passenger flows and relatively small forest coverage,such as forest parks and forest scenic areas.Since these areas with large passenger flows and relatively complex environment have limited equipment purchase funds,a forest fire monitoring and early warning system applying wireless sensors is proposed to solve the above problems.Compared with other sensor-based forest fire monitoring systems,this system incorporates a fire early warning model.The system design first analyzes the interrelationship of various factors in the fire system through the DEMATEL algorithm,so as to determine the amount of information collected by the sensor.At the same time,it collects the meteorological data when forest fires occurs,to find the central value by performing K-Means cluster analysis,and use it in Vague fuzzy decision-making.When each collection factor satisfies the alarm condition,a pre-alarm is performed,it will verify whether a fire has occurred by using a smoke sensor and other comprehensive information.Then,there will has a fire alarm prompt after the fire occurs.The user can see the real-time and historical data of the collected factors on the management end.This design may still have some deficiencies,which will be further improved in future research.