Team Control Number
C
Summary Sheet
A Setting System of Interstate Energy Cooperation
Goals Based on Data Insight
Summary
After performing data analysis and modeling, we finally determine a set of develop-
ment goals for the new four-state energy compact.
First, we preprocess the data provided, which includes default value processing, ab-
normal value process and data classification. For the sake of analysis, we divide various
energy into two broad categories. One is cleaner renewable energy (CRE), the other is
traditional fossil energy (TFE). After that, we select 11 important variables from the given
data to create the energy profile for each of the four states. We call the 11 variables the
basic variables
Next, we apply the decoupling theory to characterize the dynamic relationship be-
tween economic development and energy utilization, which can reflect the evolution of
energy profile. We find that the four states differ in production and usage of various
energy significantly. To determine the underlying factors that lead to the differences, we
construct the simultaneous equations model. Combining natural environment informa-
tion further, we find out the factors and know the respective strengths of the four states
in CRE.
Then, we establish a multi-dimensional evaluation system to identify the state that
has the“best”energy profile on the whole. We introduce the index, comprehensive
utilization performance (CUP) to measure the energy profile. The CUP is composed of
three parts, energy performance, economic performance and environment performance.
And each of the three parts includes three indexes respectively, all of which are synthe-
sized by the basic variables. We use the PCA method to integrate the nine indexes into
an overall index, namely the CUP. Ranking CUP, we find that California is the“best”.
Finally, we construct BP neural network to predict the energy profile. Analogous
to Cobb-Douglas Production Function in economics, we define the CUP in a new way
for predicting. Through setting various development scenarios, we get the predictions
successfully. After that, we regard the four states as a whole to determine renewable
energy usage targets for 2025 and 2050. In this process, we use the BP neutral network
and previous models again. Wecollect real data from 2010 to 2015 to calculate the values
of CUP. Compare them to the predicted value, we test our predicting system. The result
shows that our predicting system works well.
精品数模资料,各类比赛优秀论文、学习教程、写作模板与经验技巧、matlab程序代码资料等,尽在淘宝店铺:闵大荒工科男的杂货铺!
评论0
最新资源