homework

基于最小二乘支持向量机的NARX模型辨识，用于设备故障诊断(program based on LS-SVM NARX for diagnosis)

MATLAB实现单神经元PID控制，在智能控制算法中，运行结果较好。(MATLAB single-neuron PID control, intelligent control algorithm, the results better.)

Space Vector Modulation (SVM) Technique has become the important PWM technique for three phase Voltage Source Inverters for the control of AC Induction, Brushless DC, Switched Reluctance and Permanent Magnet Synchronous Motors.

独立成份分析问题，应用于声音的分离。此程序可以对两个混合的声音进行分离。(Independent component analysis problem, applied to the separation of sound. This program can be two separate mixed voices.)

用PTS降低PAPA的仿真源代码,试过了不错,是转发的(PTS reduce the simulation with the source code PAPA)

用matlab实现了对于离散傅里叶变换求其循环卷积(Achieved with matlab discrete Fourier transform for the cyclic convolution)

这是一个纯VML的例子，VML用于统计，有柱状图、饼状图、折线图。(This is a pure example of VML, VML for statistics, histograms, pie charts, line charts.)

explicit FDM for 1_2D_parabolic partial differential equation

Among the high-complexity methods, two methods dominate the literature Kalman filtering and Mixture of Gaussians (MoG). Both have their advantages, but Kalman filtering gets slammed in every paper for leaving object trails that can t be eliminated. As this seems like a possible deal breaker for many applications, I went with MoG. Also, MoG is more robust, as it can handle multi-modal distributions. For instance, a leaf waving against a blue sky has two modes—leaf and sky. MoG can filter out both. Kalman filters effectively track a single Gaussian, and are therefore unimodal: they can filter out only leaf or sky, but usually not both.

DO for all points in the minutia vector