The Research Of The Key Technologies For Wireless Sensor Networks In CPS Of Power Grid

In order to cope with the increasingly complex applications of power grid,various measurement and control technologies need to be highly integrated with physical power grid to form smart grid.Cyber Physical System(CPS)is an important approach to meet the requirements of smart grid.Although wireless sensor networks is a powerful tool to realize the CPS in smart grid,the complexity of power grid often makes the application of wireless sensor network facing various problems.Therefore,the study of power CPS wireless sensor network will present significant value both in theory and practice on the realization of the real-time information acquisition and online monitor for all aspects of power grid.Aiming at the characteristics of wireless sensor network in power CPS and taking,this paper studies power CPS field sensor data acquisition technology,the optimizaion of the network load maximization in wireless sensor network,extension of wireless sensor network life cycle,and geographic information localization using wireless sensor network.Due to the large volume of data during the power equipment monitoring,this thesis explores the current data acquisition protocol and data collection methods in order to improve the low sensor data acquisition rate in wirless sensor network for electric power CPS.This thesis view the nuclear power plant turbine testing tomography system as the applicaiton backgrounds.Based on the analysis of the scanning resolution in that system from 10 mm to 2 mm under the requirements of data acquisition,this thesis puts forward the clock pulse rising and falling along the reuse of bilateral reading methods to build a new data acquisition system.This method can shorten the scanning period and effectively improve the sensor data acquisition rate,which has universal application value.To solve the limited energy of the sensor node and the limited lifetime of the energy harvestting wireless sensor network due to battery power,this thesis proposed a kind of periodical max flow routing(PMRF)algorithms.It can be used for the monitoring the operations of the power generation equipments in the first phase of Datang west bridge wind farm.The proposed algorithm is based on Ford-Fulkson algorithm(FF)and graph theory.For the further improvement of PMRF,thesis proposed periodical energy aware max flow routing(PEAMRF)by introducing the residual energy of the sensor nodes.Simulation results showed that PEAMFR algorithm is more suitable for wireless sensor networks based on small world model when the number of the sensor nodes is small and PEAMFR algorithm is more suitable for wireless sensor networks based on scale-free model when the number of the sensor nodes is large.The unbalanced energy consumption of wireless sensor network for monitoring micro-grid,named “hot spot” of the monitoring area,will lead to the poor monitoring quality and the short lifetime of the monitoring network.This thesis viewed power generation equipment monitoring network for the second phase of Datang west bridge wind in Chongli as a specific application scenario.A kind of real-time energy balanced uneven clustering algorithm for wireless sensor netowk(REBUCA)is proposed.The proposed algorithm takes distance and energy into account to simplify the process of the clustering,which results in the reduction of communication overhead.Compared with the classical LEACH algorithm and the non-uniform clustering EEUC algorithm,the proposed clustering algorithm is more effective in multiple aspects such as the death time of the first node,the network life cycle and the distribution of network dead nodes.Consequently,it can be used to improve the quality of monitoring using wireless sensor network.Wireless sensor networks can be used in localization of CPS for power gird.With the photovoltaic power plant of photovoltaic array as research background,thesis puts forward an improved PA iDV-Hop localization algorithm based on the classic DV-Hop algorithm.The proposed algorithm adopts the minimum mean square error to calculate average distance among the anchor nodes by the introduction of error factor modification.when calculating the distance bwtween the unknown nodes and the anchor nodes,the proposed algorithm also introduces a dynamic weight related to its minimum hop count to modify the average distance of the unknown node.The two-dimensional hyperbolic method instead of the three-side ranging method is used to calculate the unknown node coordinates and update them by introducing parameter factors.The proposed algorithm can effectively improve the positioning error,and it present significant value both in theory and practice in the aspect of geographic information aided positioning in power CPS.