The missing data filling method of the industrial internet platform based on rules and lightGBM

Abstract

Abstract As the core of the Industrial Internet platform, industrial big data plays an increasingly important role in intelligent production, operation and maintenance. Filling missing data has always been a challenge for the data analyst. In order to solve the missing problem in the data collection process of the industrial Internet platform, according to the operation mechanism of the industrial Internet platform, this paper proposes a rule-based and Light Gradient Boosting Machine (LightGBM) algorithm to fill the missing data of the Industrial Internet platform. For the input data with 1 missing interval, the rule of the model is taken to complete the filling. While, for the data with larger missing interval, the combination of rule and LightGBM is used to iteratively realize the filling. The details could be found in figure 1. Taking the data of the Industrial Internet platform of the electric power industry as an example, the comparison and verification show that the improved data filling method is faster, has less error and better performance, which is suitable to solve the missing data problem of the Industrial Internet platform.