| Forest fire monitoring is a key component in the ’earlier detection’ of forest fires,and forest fire information sensing is one of the four capabilities developed for forest fire monitoring.In the current field of forest fire risk assessment and environmental awareness,the conventional solutions include the use of 3S-based environmental meteorological factor inversion to assess forest fire risk,but there is a problem that the spatial resolution of satellites is not balanced with their revisit period,resulting in poor accuracy of forest fire risk assessment;Video surveillance using a combination of visible light and thermal infrared imaging technology for fire monitoring,but camera surveillance is susceptible to cloud and fog,and there is a significant time lag,preventing ’earlier detection’;The Zig Bee-based WSN ground-based system provides awareness of the forest fire environment,but has significant drawbacks such as short communication distance and poor interference immunity in the forest,as well as huge costs if deployed on a large scale.Therefore,it is important to build a new type of real-time forest fire environment sensing system with long range,long duration and high accuracy.This thesis uses LoRa RF and NB-IoT communication technology to carry out the integrated design of a wireless sensor network for fire risk environment sensing in complex forest areas.Through the design of node hardware circuits and embedded software control flow algorithms,a fire risk environment sensing solution is constructed with the advantages of intelligent sensing,information gathering,local decision making,collaborative regulation and control,which can achieve the purpose of sense-transmitcontrol applications.The work and results of this thesis can be summarised as follows:(1)The design requirements for WSN-based fire risk environment sensing are analysed and explained,detailing the main challenges faced in implementing this sensing system: the types of environmental factors to be collected and the choice of RF/Io T communication methods,and providing specific answers.The design principles and general framework are then elaborated accordingly,and finally the design of this solution is detailed both in terms of the hardware platform and the embedded software platform.(2)A specific and comprehensive analysis of the power consumption of different types of nodes is presented,and the methods for calculating the power consumption of different circuits are given.Various solutions are also proposed to extend the lifetime of the nodes and to optimise their power consumption: in terms of hardware circuits,controlling the power consumption through programmable switches,optimising the response time of the network by using interrupt-trigger wake-up circuits,introducing radio wake-up technology to solve the network wake-up in unexpected conditions,etc.In terms of software algorithms,energy consumption control solutions such as adjusting the sampling rate of environmental factors using fire risk prediction models and dynamically adjusting the node sleep time and RF signal transceiver power according to the principle of temporal and spatial adaptation are being developed.In addition,a variety of node security protection initiatives are provided in the field environment to ensure the stability and durability of node operation.(3)Relevant tests are carried out in three aspects: functionality,performance and overall stability,and the main indicators of the network are described qualitatively or quantitatively by means of indoor and outdoor experiments.The WSN system was tested for its effectiveness in terms of intelligent sensing and collection of environmental factors in the forest,self-grouping of nodes,area localisation and uploading of data to the cloud server,as well as node power control and system stability after introducing optimisation strategies in the system.After analysing and evaluating these test results,some suggestions are also made to improve the performance and efficient deployment of the nodes,such as the problem of RF antenna orientation,the problem of calculating the number of nodes in the area,and the problem of predicting the lifetime of the network. |
