At a Glance
Commercial Energy Efficiency, Data Analysis
Quantifying energy saving potential based on baseline energy consumption model.
The challenge for McDonald's is to make its restaurants more energy-efficient. The energy efficiency level of the equipment itself has been upgraded, but energy use behavior has a great influence on the actual energy efficiency. In order to exploit behavioral energy saving potential, energy data needs to be modeled and analyzed.
The fellow took three steps to meet the challenge:
· Determine the characteristic variables of the model. After a descriptive analysis of a variety of continuous and categorical variables, I think it is most appropriate to take into account the characteristics of the restaurant's total area, total sales, temperature, and the city in which the restaurant is located, etc.
· Determine the regression method of the model. Considering the four regression methods of ridge regression, multiple linear regression, elastic network regression and gradient enhancement regression, the cross-validation method was used to obtain R2, MSE and other indicators, so as to confirm that GBR has the best fitting effect for various devices and total power consumption.
· Determine the relative energy efficiency of each restaurant. The predict values of the model and true values are used to evaluate the energy efficiency of each restaurant and quantify the energy saving potential and cost saving.
These projects, strengthen the management of energy use behavior, when fully implemented in 3,000 restaurants, could result in 52,762,438 kWh of annual electricity savings, $4,717,534 of cost savings, and 14,351 metric tons of CO2 emissions reductions.