toPPT_2

分析了支持向量回归机在能源需求预测中的优势,确定了输入向量集合和输出向量集合,建立了基于Matlab技术的SVR能源需求预测模型.对我国1985-2008年能源需求相关数据进行模拟与仿真,并对中国2010年和2020年能源需求量进行预测.研究结果表明:一是中国未来对能源的需求量逐渐增加,从2010年的330400万吨标准煤上升到2020年418320万吨标准煤,年均增长率为2.39％；二是在解决我国能源系统小样本.非线性及高维模式识别问题中SVR比BP神经网络等方法有更高的预测精度.(Support vector regression analyzes advantage in energy demand forecast to determine the set of input vectors and output vector set, established SVR energy demand forecasting model based on Matlab technology. Energy demand for our 1985-2008 related data modeling and simulation , and China in 2010 and 2020 energy demand forecast results show that: First, China s increasing demand for energy in the future, from 3.304 billion tons of standard coal in 2010 rose to 4.1832 billion tons of standard coal in 2020, average annual growth rate of 2.39 second is to solve our energy system small sample nonlinear and high dimensional pattern recognition problem SVR higher prediction accuracy than the BP neural network method.)

龙格库塔法和fft转换计算非线性微分方程，十分好用。简单方便。(Runge kutta method and FFT transformation calculation of nonlinear differential equations, is very good.Easy and convenient.)

小波分析Matlab实现，可以直接编译运行(Matlab Wavelet Analysis realize, you can direct the compiler to run)

此程序为高斯白噪声AWGN信道的matlab描述(This program is Gaussian white noise AWGN channel matlab description)

程序内容是对信号的波束域检测，其中阵元数32，信号为宽带信号(Program content is the signal beam field testing, which array element number 32, the signal is wideband signals)

performance of channel capacity in wireless communication systems

粒子群和自适应混沌粒子群算法的程序，作了简单对比(Chaos particle swarm optimization and adaptive particle swarm optimization program, made ​ ​ a simple comparison)

echo cancellation adaptive filter

about filter bank non-uniform

文中的方法是把图像分块，小波分解得到低频分量、高频分量，然后计算每一块的对比度，把图像块划分为清晰块、模糊块，把清晰块和模糊块相邻的区域定义为边界区域，融合时，直接选取清晰块作为融合后的相应块，对于边界区域，在小波分解的基础上采用基于对比度的像素选取的方法进行处理。(Paper, the method is to image segmentation, wavelet decomposition are low frequency, high frequency components, then calculate the contrast of each piece, the image block is divided into clear blocks, fuzzy block, to clear blocks and fuzzy block is defined as the border region adjacent to area, integration time, a clear block directly select the corresponding block as a fusion, for the border region, the wavelet decomposition on the basis of the pixel-based contrast method selected for processing.)