Research On Several Key Techniques Of Wireless Sensor Network In Smart Grid

As it is a national requirement to implement smart grid technology to the expending electricpower systems that nowadays extends down to the remote rural areas and integrates emergingrenewable energy resources, an economic way such as the wireless sensor networks is worth to bestudied for building communication links to the numerous nodes at the lowest level of the powersystem instead of expensive communication cables. This dissertation studied some key problems ofwireless sensor network applied in the smart grid of a local power distribution system, such as thegraph model of smart network topology, QoS routing algorithm, and fault positioning andmonitoring etc. The above mention theoretic results were applied to and verified by a local powerdistribution system.The main work with its contribution of this dissertation may be summarized as following:1. The graph model that meets the wiring configuration and the terminal distribution isestablished for power distribution network. An improvement is made to the Dijstra routing algorithmto find the shortest paths. It is verified by simulation that the reasonable allocation of wirelesssensors, the optimized topology of wireless sensor networks, and higher efficiency of data transfermay be realized by using this improved algorithm.2. A real-time performance and reliability based QoS routing mechanism is proposed for smartpower distribution grid in order to obtain high-performance communication wireless sensornetworks. By analyzing the failure mode of the links, power-assigning to the delay and the reliabilityof the network nodes, a mathematical model that optimizes reliability and satisfies real-timeperformance is established for the path selection. It is verified by example analysis and simulationthat the propose model and algorithm may synthetically deal with and improve both the real-timeperformance and the reliability of data transmission.3. It is proposed to apply the locating algorithm of wireless sensor networks to monitor the faultlocation on power transmission lines. By analyzing the particularity of wireless sensor nodes ontransmission lines, a model is established for the relationship between sensor nodes and thetransmission line fault location. By introducing the straight line/poly line ratio and the hops betweenthe anchors into the error calculation, an improved DV-Distance algorithm can reducing thecumulated distance error and the location error. To overcome the shortcomings of DV-Hop algorithm,three improved algorithms are proposed based respectively on secondary reference node, the node oflast hop, and the Max-Min method and their effectiveness are tested, compared and validated on asame platform. 4. The above mentioned theoretical results were applied to an engineering project of HuangshanPower Company. The smart distribution network with wireless sensor networks was established. TheWSNs terminals, routers, coordinators and associate software are developed with embedded QoSrouting protocol, gateway management. The operation of the project indicates the application ofwireless sensor networks in a smart power distribution system is realized and expected goals of ahigh performance data communication is achieved.