Research On The Methods Of Acquisition And Reliability Evaluation For Environmental Information Based On Internet Of Things

China is not only an important coal-producing country,but also one of the countries that most seriously affected by coal mine disasters.China has gradually set up the safety monitoring system in the coal mine based on the internet of things with the purpose of preventing safety accidents such as gas outburst.However,the system often leads to false and missed alarm due to the aging of sensors,failure or artificial occlusion damage of sensors and other problems,which greatly reduces the effectiveness of the safety monitoring system and lays a hidden danger for safety production.The kind of problem is common difficulty in the security monitoring system of internet of things,so how to evaluate the reliability of sensor information in the security monitoring system based on the internet of things is the key to improve the effectiveness of the monitoring system.In response to the problem that the traditional security monitoring system can't identify the reliability of the sensor,which affects the performance of the system.Then the following work is carried out from two aspects of environmental information acquisition and sensor information reliability evaluation:I.Considering the problems that the existing information acquisition system based on the internet of things is difficult to support the research of information reliability because of the slow response of sensors,unstable data transmission and other factors.Then a set of information acquisition system based on the internet of things is developed.On the software level,a high reliable MAC protocol based on TDMA is proposed,which guarantees the reliability of transmission by getting rid of competition and multi-level retransmission.On the hardware level,the response time of gas sensor effectively reduced by optimizing sensor selection and designing active airflow path.The construction of the acquisition system provides a good data support for the research on reliability evaluation of sensor information.II.Considering the reliability evaluation of sensor information under the condition of single sensor,a reliability evaluation method based on Kalman filtering is proposed.Aiming at the special situation that a large number of outliers existed in the original information,then reliability evaluation method based on iterative fitting is proposed,and the performance of the two algorithms is simulated and compared.III.Considering the reliability evaluation of sensor information under the condition of multi-sensor.Firstly,the spatiotemporal correlation of sensor information in the internet of things is analyzed,and then the spatiotemporal correlation is used to raise reliability evaluation method based on the neighbor.According to the problem of neighbor misjudgment caused by fault sensor,a new reliability evaluation method based on distance-reliability with double-weighted is proposed.Meanwhile,the performance of the two algorithms is simulated and compared.In the thesis,the acquisition system has passed the test verification,and reliability evaluation method has passed the simulation verification and has been tested by the acquisition system.The results show that the acquisition system enables to collect environment information stably and reliably,and reliability evaluation algorithm can effectively identify abnormal data under the conditions of simulation and experiment.Finally,the research content of the thesis provides a theoretical and engineering basis for improving the performance of various security monitoring systems in internet of things,such as coal mines,which has an important significance.