Low Power Design Of Wireless Sensor Networks Node For Mechanical Vibration

Mechanical vibration wireless sensor network has self-organization,strong expansibility,convenient deployment and flexible features,can effectively compensate for the limitations of the traditional wired monitoring system,it has great application value and potential application prospect in closed,rotation,large displacement,high temperature and other special extreme environments.Compared with the low frequency applications such as structural health monitoring,the performance of wireless sensor networks with higher sampling frequencies is more demanding.For example,the higher sampling frequency,higher sampling accuracy,higher precision,higher data rate and higher reliability,high performance and high energy consumption demand brings disadvantages,the energy limitation of wireless sensor networks is more severe.The existing mechanical vibration wireless sensor network nodes have the problems of excessive power consumption and short network life cycle,so it is difficult to realize long term mechanical vibration monitoring.There are many factors that affect the mechanical vibration of the wireless sensor network node energy consumption,including the vibration signal acquisition,data processing,data caching and transmission of multiple links,such as signal acquisition hardware,sampling interval,sampling frequency,sampling length and transmission process topology is balanced,transmit power,transmission path is optimal optimization,whether the raw data processing etc..This dissertation focuses on the following three key factors that affect the energy consumption of wireless sensor networks in mechanical vibration:(1)Aiming at the problems of high hardware energy consumption in mechanical vibration wireless sensor networks node,a hardware design method based on single processor dual core architecture is proposed.Firstly,the key factors that affect the energy consumption of mechanical vibration wireless sensor nodes are analyzed under the premise of guaranteeing the accuracy of data acquisition.The management scheme of low dissipation and multi power supply and the drive circuit of IEPE driver for self noise suppression of power supply are designed to avoid crosstalk between analog and digital circuits,to reduce the influence of noise on signal acquisition,and to ensure the minimum energy dissipation of the power circuit.The high precision oversampling A/D conversion module is adopted to avoid the influence on the accuracy of data acquisition from unstable time interval of the processor.By means of A/D internal digital filter and Sallen-Key six order active low pass analog filter,the bandpass filter with smooth passband and low power consumption is realized.In terms of overall architecture,an internal integrated dual core SOC is adopted as the core of the whole control system,which integrates data acquisition,storage and wireless transmission.While ensuring the collection performance of mechanical vibration signals,the energy consumption of nodes is greatly reduced.(2)Aiming at the problems of mechanical vibration wireless sensor networks node with limited storage capacity and narrow bandwidth should store,read,transmit the mass vibration data which generated under the high sample rate in blocks,while the data block size will affect the energy consumption of data storage,data reading and data transmit directly,a method of data block optimization in node data buffering and sending is proposed.In order to study the relationship between the data block size and energy consumption,Firstly,obtain the discrete relationships between operating current,operating time and the data block size in data storage,data reading and data transmit processes with experiment method.Then based on the discrete relationship,construct the energy consumption mathematical model of each process by least square method and obtain the optimal data block size which makes the energy consumption of each process lowest combined with limiting factors in storage capacity and physical layer payload.(3)Aiming at the problems of mechanical vibration wireless sensor networks node has low data transmission rate,high node idle rate,high energy dissipation,especially in mass raw data transmission requirements,a Multi-channel Data Transmission method based on tree-star Hybrid Topology(MDTHT_WJ)is proposed.Firstly,to avoid adjacent channel interference affect parallel communication,allocate channels for each wireless sensor network node with minimizing inter-tree communication interference,each node forms a tree-star hybrid topology network with allocated channel for data transmission after data acquisition.Then,handshake mechanism and preemptive priority mechanism is utilized to solve network partitions problem.Finally,the network short address of each data acquisition node is regarded as scheduling information,routers broadcast the scheduling information with beacons,and each data acquisition node decides to whether transmit data according to the scheduling information,ensure the time-scheduled inner-tree communication with minimizing data collision,realizing the network parallel transmission,improve the data transmission rate,reduce the node idle time and energy dissipation in the process of data gathering.(4)A mechanical vibration wireless sensor network monitoring system with low power is designed.The data monitoring center software is designed based on C#.NET platform,the function modules of wireless sensor network setting,acquisition parameters configuration,data gathering and display are realized.The performance of monitoring system is tested and the effectiveness of the proposed node hardware design and multi-channel data transmission method.At the end of the thesis,the summarization of the article and experctation of the relative technology development are presented.