教育学论文金融学论文 国际商务管理论文 市场营销论文 会计学论文 社会学论文 旅游管理论文 计算机论文 人力资源管理论文 广告媒体学 物流学论文 Engineering Thesis sociology Thesis MBA论文 linguistic Thesis Journalism and Mass Communicat Law Thesis Economics Thesis
返回首页

制定一篇金融专业的留学生论文-The Depth Analysis by New Model

时间:2013-12-18 13:22来源:www.ukthesis.org 作者:英国论文网 点击联系客服: 客服:Damien
制定一篇金融专业的留学生论文-The Depth Analysis by New Model
 
Considering the wind

 

As a commodity, the electricity price is affected by the flow and distribution of all kinds of resources in the entire market. With the development of renewable energy generation technology and the increase of its share, such as the wind power, electricity price curve is more complex than before. The tariff fluctuations are very sensitive to any changes in grid-connected wind power, which will lead to rose or fell of the electricity price. Due to the limitation of GBM estimates, the neural network model can effectively deal with the multi-variable and nonlinear problems, which is qualified for the analysis of grid-connected wind power system.作为一种商品,电力价格是由各种资源在整个市场的流动和分布的影响。随着可再生能源发电技术的发展和增加其份额,如风力发电,电价曲线比以前更加复杂。电价波动对并网风电的任何变化,这将导致大涨或大跌的电价非常敏感。由于GBM估计的局限性,神经网络模型可以有效地处理多变量,非线性问题,这是合格的并网风电系统的分析。
 
The analysis main consider the factors like grid-connected wind power, load and historical clearing price. Taking into account a different variation of the tariff at different time period, it is divided into 48 sub-time tariff sequences. 
Prior to electricity price forecasting, we need to make clear of the relevance between the factor and tariff. The correlation coefficient can be used to measure the degree of association, which is defined as follows:  .Where, ρ is the correlation coefficient, f for a certain influence factor, σf for factors of standard deviation, p for electricity price, σp is the standard deviation of the tariff, Cov (f, p) is the covariance between factor f and tariff p. 该分析主要考虑像电网风电,负载和历史清算价格的因素。考虑到资费在不同的时间段的一个不同的实施方案中,它被划分成48个子时间资费序列。
此前电价预测,我们需要明确的因素和关税之间的相关性。相关系数可以用来测量关联程度,其定义如下:式中, ρ为相关系数,F为一定的影响因子, ΣF为标准偏差,对电力价格的因素, ΣP是资费标准差,冠状病毒( F,P )为系数f和关税P间的协方差。
 
According to the British electricity market data, we can calculate the correlation coefficient between the wind power and related load ratio (ρ1) and the correlation with historical clearing price (ρ2). Based on the strong correlation with tariff, we consider the equivalent load history clearing price, load, wind power and the load ratios, historical clearing price as the neural network input factor to forecast market sub-period clearing price.
 
Neural network is a parallel, distributed information processing structure consisting by the processing unit. Artificial neuron is to simulate the basic characteristics of the neurons in the brain, which has multi-input / single output nonlinear unit with a certain internal state and the threshold.
 
Radial basis function network (RBF) has only one hidden layer, the output is a weighted sum of the hidden layer [12-13]. The most commonly used radial basis function is a Gaussian function. Generalized regression neural network (GRNN) is a variation of the RBF network [14-15]. The theoretical basis is non-linear regression analysis. We can obtain the non-independent variable y with respect to the regression analysis of independent variable x. As a forward-feed neural network, GRNN has input layer, a hidden layer and output layer. Hidden layer used Gaussian transformation to control the output, thereby inhibiting the activation of the output unit. Gaussian function belongs to accepted domain in the input space. The influence input neuron was attenuation because of the distance between the input vector and the network output.
 
To calculate the evaluation error, the traditional method of mean absolute percentage error (MAPE) is not enough. We substitute the predictive value with the mean price to decrease the error caused by the fraction. 
 .
However, we need the probability index to evaluate the credibility of the forecast results [13]. And the error distribution function F(ε<ε(n))can be fitted by S(n)(ε) as the following.
 
Where ε represents the errors. 
 
Considering the storage
The generation, transmission, distribution and use of traditional electricity production is almost preceded at the same time, which significantly influenced the planning, construction, scheduling, operation and control of power system. The application of large-capacity storage technology will break the limitation of the real-time power supply and demand balance. The development of storage has become an inevitable trend of energy storage technologies in power system [1.5].
Currently, there are several storage devices. 
 
Pumped storage reserves the pumped water at the upstream during low load hours, and generates electricity using the stored water during the peak load hours. The efficiency is about 75 percent. Limited by location places and construction period, it is a bit difficult for large-scale application. 
 
The flywheel energy storage device combines the motor with the flywheel, which can store energy into a high speed rotation, and transfer into the electricity when necessary. The efficiency can reach 85percent to 90percent[7]. The series of commercial products practice this storage style. Benefited by the rapid response performance, it can be used as small capacity, short discharge time situation. [1](责任编辑:anne)


------分隔符-------------------------------------
UK Thesis Base Contacts
推荐内容