Hourly Energy consumption in industrial site

The dataset contains an example of energy consumption, Functioning hours and Production KPI of different stages of the experimental open pit mine, mainly the destoning, the screening, and the train loading station. The Code is an example of the prediction algorithm, and the API can be used to apply the same algorithm used in this article.

In the proposed Dataset the energy consumption data for each station are collected from power meters and stored into a database that contains functioning hours and production.

In the data pre-processing stage, the gathered data are cleaned removing unwanted observations, such as energy consumed without equivalent timestamped production or functioning hours, dropping missing data. In the k-fold cross validation stage, the dataset is shuffled randomly, then split into groups, considering the group to be test data set or a holdout to compare. The remaining groups are considered training data set. Then comes fitting the model to the training data and testing it on the test data, keeping the evaluation score and discarding the model.