2012 HiMCM B题特等奖学生论文下载3594

2012 HiMCM B题特等奖学生论文下载3594

 

下载方式见文末

 

论文摘要如下：

 

15th Annual High School Mathematical Contest in Modeling (HiMCM)

Summary Sheet (Please attach a copy of this page to each copy of your Solution Paper.)

Team Control Number: 3594

Problem Chosen: B

Please type a summary of your results on this page. Please remember not to include the name of your school, advisor, or team members on this page.

 

Executive Summary:

 

Given the constantly fluctuating price of gas, as consumers, we face a challenging decision every time we pull up to the pump: to fill or not to the fill. Without knowledge of what the gas prices will do in the coming weeks, we risk spending our money only to watch the gas prices drop later. Our task was to solve this problem by developing a model to instruct a consumer on when to buy gas so as to minimize spending. Specifically, we looked at consumers who use a half tank in a week and those who use a quarter tank in a week. We developed a model that could predict future gas prices and determine how much gas a consumer should buy in a given week based on whether the price would go up or down in the next couple of weeks. Our future gas price predictions were based on the correlation between crude oil price and gas price. Since the process of refining and transporting crude oil to produce gas takes time, the current trend of the crude oil prices forecasts the trend in future gas prices.

 

Once we had a model for predicting prices, we used a decision tree to develop an algorithm for how much gas a consumer should buy on a given week in order to be cost efficient. Therefore, the consumer only needs to input the crude oil prices for the past three weeks, the current gas price, and the current amount of gas in their tank, and our model will tell them whether to buy a full tank, half tank, or no gas. Our model is applicable to large cities across the United States, predicts future gas prices with only 1.63% error, and recommends a buying pattern that differs only 1% from the ideal spending that would occur if consumers actually knew future gas prices. Therefore, our model is useful as a way for drivers to save money.

 

2012 HiMCM特等奖论文完整版 下载链接