没有合适的资源?快使用搜索试试~ 我知道了~
亚太数学建模的c题论文,代码使用python实现
需积分: 0 5 下载量 118 浏览量
2022-12-03
10:39:22
上传
评论
收藏 1.12MB PDF 举报
温馨提示
试读
30页
亚太数学建模的c题论文,代码使用python实现
资源推荐
资源详情
资源评论
Team # apmcm2202706
Team Number :
apmcm2202706
Problem Chosen :
C
2022 APMCM summary sheet
The main work in this paper is to build a model to fit the global average
temperature as a function of temperature based on a small amount of data, and to
analyze the degree of influence of other factors on global temperature using
correlation tests. Finally, a proven method for alleviating global warming is proposed
based on our findings.
In the first part, to solve the problem of relatively small amount of data relative to
the global area, we propose a superposition prediction model based on Gaussian
distribution, which can use only a small amount of data to fit the global temperature
distribution and finally derive the global average temperature variation over time
based on the global temperature distribution in latitude and longitude. Next, our study
relies on the obtained global temperature data to train the BP and LSTM neural
networks separately, and the obtained model fits the historical temperature data
relatively accurately and predicts the global temperature change in the next 80 years.
After comparing the results, we believe that the results obtained from the LSTM
model are more realistic, while the results obtained from the BP model tend to reflect
the trend of global temperature change.
In the second part, we use the annual average national temperature, which is more
important and representative than other factors, as the independent variable and the
global average temperature as the dependent variable, to measure the sensitivity of
global temperature to regional temperature and time through multiple regression
analysis. Fires, novel coronavirus outbreaks and forest fires were selected as proxies
to examine the impact of regional temperatures and natural hazards. Analysis of the
observations shows that natural hazards do not have a significant impact on regional
temperatures or global temperatures in the short or long term.
Next, a multiple regression analysis was conducted to study the variables that
have a significant impact on global temperature, in order to find the factors that have
the greatest impact on global temperature, so as to achieve the purpose of observation
and control of global temperature change.
In the end we wrote a non-technical paper to be submitted to the organizing
committee to report on the general results of our modelling and the recommendations
we had made.
Team # apmcm2202706
Contents
1 Introduction .................................................................................................................................... 1
1.1 Problem Background ........................................................................................................... 1
1.2 Problem Restatement .......................................................................................................... 1
1.3 Our Work ............................................................................................................................ 2
2 Assumptions and Justifications ...................................................................................................... 2
3 Notations ........................................................................................................................................ 3
4 Data Restoration ............................................................................................................................. 3
5 Models for question 1 .................................................................................................................... 4
5.1 The solution for part a) ........................................................................................................ 4
5.2 Neural Network model ........................................................................................................ 5
5.2.1 BP Neural Network .................................................................................................. 5
5.2.2 LSTM Neural Model ................................................................................................. 9
5.3 Gaussian Distribution model ............................................................................................. 11
5.4 Analysis of results ............................................................................................................. 14
6 Models for Question 2.................................................................................................................. 15
6.1 Model to Connect Locations to Temperature .................................................................... 15
6.2 The impact of various factors on global warming ............................................................. 16
6.3 Analysis of results ............................................................................................................. 17
7 Sensitivity Analysis ...................................................................................................................... 17
7.1 Sensitivity of average annual temperature in the region ................................................... 17
7.2 Sensitivity of other factors ................................................................................................ 18
8 A non-technical article to APMCM organizing committee .......................................................... 18
9 Model Evaluation and Further Discussion ................................................................................... 19
9.1 Strengths ........................................................................................................................... 19
9.2 Weaknesses ....................................................................................................................... 19
9.3 Further Discussion ............................................................................................................ 19
10 Conclusion ................................................................................................................................. 20
References ....................................................................................................................................... 20
Appendices ...................................................................................................................................... 20
Team # apmcm2202706 Page 1 of 28
1 Introduction
1.1 Problem Background
From the Canada’s unbelievable temperature to the general temperature of over
50℃ in the Middle East, the reality that the earth is burning is beyond doubt.
This phenomenon, known as world warming, is a consequence of continuous
accumulation of greenhouse effect, which leads to an imbalance between the energy
absorbed and emitted by the Earth's atmospheric system, and the accumulation of
energy in the Earth's atmospheric system, leading to an increase in temperature and
global warming [1].
Nevertheless, there are still some people who haven’t realized how serious the
situation has been, and the need for an exact description of the global warming still
remains to be satisfied. Moreover, even if the seriousness has been elaborated, the
problem cannot be addressed yet until the underlying causes are discovered and
countermeasures are proposed and implemented.
1.2 Problem Restatement
The first problem requests us to make an analysis of the global temperature
changes. At the beginning, we have to make a judgement about whether the increase
of global temperature in March 2022 resulted in a larger increase than observed over
any previous 10-year period. Next, to realize how serious the dilemma has been, we
need to build several mathematical models which can not only reflect and demonstrate
the conditions of the past years and will also have the ability to predict and tackle the
future. After that, when it comes to the issue about which model work best, we shall
test our models through a prediction of whether the average global temperature will
reach 20℃ in 2050 or 2100.
Then we need to utilize the results from the first problem to construct a module to
evaluate some possible indicators to determine whether they are the causes of global
warming or not. The indicators would include but not limit to the time, location,
volcanic eruptions, forest fires and the COVID-19. After the analysis, we shall find
Team # apmcm2202706 Page 2 of 28
the main reasons, and then develop some methods to curb or slow down global
warming.
1.3 Our Work
Figure 1: Flowchart of our modeling work
2 Assumptions and Justifications
1.The sensitivity of global warming to an area is only related to the average
annual temperature at that location, ignoring the remaining factors that are of little
relevance. What’s more, through subsequent model sensitivity analysis, we can
demonstrate that the effects of the remaining factors are white noise compared to the
mean temperature of the site and can be ignored.
Team # apmcm2202706 Page 3 of 28
3 Notations
Table 1: Notations used in this paper
Symbol
Description
Unit
o / h
order of the output and hidden layer neurons
\
x
input data
\
inputs and outputs of the hidden layer neurons
\
inputs and outputs of the output neurons
\
f(x)
activation function
\
w
connection weight parameter
\
b
bias parameter
\
η
learning rate
\
the state of renewal of memory cells at moment t
\
input Gate
\
forgotten Gate
\
output Gate
\
output of the hidden layer at time t
\
memory cells with input and the weight matrix of the
hidden layer
\
input gate with weight matrix of hidden layers, memory
cells
\
oblivion gate with output layer, weight matrix of memory
cells
\
memory cells with output layer, weight matrix of
memory cells
\
⊗
point multiply
\
𝜎
sigmoid activation function
\
𝑏
𝑐
,𝑏
𝑖
,𝑏
𝑟
𝑏
𝑜
deviation
\
Gaussian distribution
\
D
dimension
\
covariance matrix
\
mean vector
\
exp
natural logarithmic function
\
Latitude and longitude of the city
\
temperature at coordinates (x,y)
\
4 Data Restoration
Considering that some of the data appeared to be missing, we have to carry out
剩余29页未读,继续阅读
资源评论
「已注销」
- 粉丝: 1
- 资源: 1
上传资源 快速赚钱
- 我的内容管理 展开
- 我的资源 快来上传第一个资源
- 我的收益 登录查看自己的收益
- 我的积分 登录查看自己的积分
- 我的C币 登录后查看C币余额
- 我的收藏
- 我的下载
- 下载帮助
安全验证
文档复制为VIP权益,开通VIP直接复制
信息提交成功