Abstract
In recent years, accidents occurred in the chemical plant are linked to the failure of the safety valve or the excessive mechanical stress, and the reasons are mainly related to the safety of the safety valve, piping, support selection and the stress overload of the safety valve and the attached pipelines. The status of the safety valve cannot be effectively monitored, so that blind spots exist in the industrial production process, which has taken great operational risk. Voice recognition is an important subject of pattern recognition, and is the most important part of the security monitoring system. In this paper, the state monitoring of safety valve in industrial production is studied with the latest acoustic identification localization algorithm as well as the advanced wireless sensor network. Firstly, the working principle and the pressure mode of the safety valve are studied. Secondly, the key algorithm is analyzed for the audio frequency detection. As for the problems faced in industrial production, this paper proposes a new wireless sensor network application model that is based on the new audio detection algorithm. When this technology is applied, the problem that the state of the safety valve cannot be monitored can be effectively solved. Finally, on the basis of the above research, this paper designs and implements a set of safety valve take-off positioning system which is based on audio sensor, and some relevant experiments are designed to verify it. It has the characteristics of accurate location, small disturbance, low cost and little robustness, which is a successful exploration and attempt of industrial production by using artificial intelligence and sensor technology.
