{"title":"Big Data in Power Systems: An Introduction to Julia Linear Models using Tensor Flow","authors":"Ebby Thomas","doi":"10.1109/ISGT-Asia.2019.8881136","DOIUrl":null,"url":null,"abstract":"This paper proposes a set of tools to deal with high dimensional data that is often encountered in power systems. The tool set seems promising to network utilities, energy companies, data enthusiasts and others who are involved with the future planning, development, maintenance and operation of the power systems. The tool facilitates modelling and prediction based on Linear Model topology as well as variable screening based on user-discretion. Though the same tool set can be applied in a wide range of applications, here, as an example, the most significant variables that contribute to the energy consumption of a customer is obtained from among a pool of variables. Here, we put use to the immense data available in the power system paradigm, which, at the moment is not utilised to its full potential. The novelty of the paper is in using Julia and TensorFlow framework together in this dimensionality reduction. The training set input to the TensorFlow algorithm is utilised to establish a Linear Model and is later optimised to reduce the error through Gradient Descent optimisation. The final model is used to predict the energy usage, R-Squared values after every iteration is observed to give a flavor of variable screening process.","PeriodicalId":257974,"journal":{"name":"2019 IEEE Innovative Smart Grid Technologies - Asia (ISGT Asia)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 IEEE Innovative Smart Grid Technologies - Asia (ISGT Asia)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISGT-Asia.2019.8881136","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
This paper proposes a set of tools to deal with high dimensional data that is often encountered in power systems. The tool set seems promising to network utilities, energy companies, data enthusiasts and others who are involved with the future planning, development, maintenance and operation of the power systems. The tool facilitates modelling and prediction based on Linear Model topology as well as variable screening based on user-discretion. Though the same tool set can be applied in a wide range of applications, here, as an example, the most significant variables that contribute to the energy consumption of a customer is obtained from among a pool of variables. Here, we put use to the immense data available in the power system paradigm, which, at the moment is not utilised to its full potential. The novelty of the paper is in using Julia and TensorFlow framework together in this dimensionality reduction. The training set input to the TensorFlow algorithm is utilised to establish a Linear Model and is later optimised to reduce the error through Gradient Descent optimisation. The final model is used to predict the energy usage, R-Squared values after every iteration is observed to give a flavor of variable screening process.