Research On Privacy Preserving Data Mining Of Smart Grid

The development of Internet of Things technology,make the extension of information to any items to items,led to the world's third wave of information industry development after the computer and the Internet development.More and more traditional industries take steps with the development of Internet of Things technology,such as to smart home,intelligent medical,intelligent transportation,smart grid and other emerging industries,but the ensuing security problems are growing.Particularly in the smart grid,as an important part of the infrastructure of our country,the damage of smart grid system will easily lead to a series of disaster,even more worrying is that the lack of the smart grid information Privacy preserving mechanism,attacker can easily obtain the user's personal information from the electricity data,which are seriously harmful to the user privacy.In view of the above situation,this paper bases the background of the Internet of Things,selects the the smart grid which is an important application,analyzes its security problem,summarizes and puts forward the smart grid security reference model based on the security model of the Internet of Things.The model summarizes the security from multiple angles of the threat as basis,summarizes the smart grid security reference model from the three dimensions which are security architecture,security measures and security objects.On this basis,we use the smart grid as the research object,focusing on the smart grid preserving of user privacy.Specifically,the contribution of this article are as follows:The main contents of this paper are as follows:1.We recognize that smart grid security issues are the lack of systemic,structural security measures.On the basis of security reference model of the Internet of Things,the security reference model of the smart grid is summarized from the three dimensions of whick are security architecture,security measures and security objects.Summarize the technical means to ensure the security needs of the smart grid from the point of security architecture,summarize the security measures from a logical implement point,summarize the smart grid security targets from the point of security object.2.On the basis of the smart grid security reference model,we aim to the smart grid,focused on the issue of smart grid preserving of user privacy.In order to associate the privacy preserving of smart grid protection with the abnormal behavior pattern mining problem,we start from the different dimension defined by the traditional privacy,and correspond them to the smart grid privacy,then put forward the definition of the abnormal behavior.3.On the basis of the definition of abnormal behavior pattern,this paper studies and implements a "Internal-External Entropy Based Privacy Data Mining in Smart Grid" method.The natural ordering of this method also reflects the "abnormal degree ".4.On the basis of mining the abnormal data,the paper studies and implements a "HP Filtering with Regularization" method.This method use HP filter with regular terms which can not only ensures the change of privacy information,but also constrain the character of smart grid electricity data.So this method ensures data privacy as much as possible without compromising the availability of data.Finally,this article uses the public smart grid data set to test the effect of privacy mining and privacy preserving.Experiments show that the anomaly behavior mining method based on the entropy difference of the intrinsic-external model is significantly increased compared with the standard method,which means that the amount of privacy information can be preserved will be more,and leak of privacy information will be less;and we use multiple sets of parameters to find a more balanced comparison of the results,so the the purpose of privacy preserving can be achieved.