MULTI SUPPORT VECTOR MACHINES DECISION MODEL AND ITS APPLICATION

IntroductionStatistical Learning Theory( SLT) is a small-sample statistical theory by Vapnik etal.SupportVector Machine( SVM) is a novel powerful ma-chine learning method developed from SLT.SVMis powerful for the problems will small sample,nonlinearity,high dimension and local minima.Currently,SVM has many applications in the pat-tern recognition,function estimation,signal pro-cession,control,and others field[1～ 3 ] .SVM en-hances generalization by principle of the structuralrisk min…     

Steam Pressure Control of 1000 MW Ultra-Supercritical Coal-Fired Power Unit Based on Multi-Model Predictive Control

Ultra-supercritical(USC) coal-fired unit is more and more popular in these years for its advantages.But the control of USC unit is a difficult issue for its characteristic of nonlinearity, large dead time and coupling among inputs and outputs. In this paper, model predictive control(MPC) method based on multi-model and double layered optimization is introduced for coordinated control of USC unit running in sliding pressure mode and fixed pressure mode. Three inputs(i.e. valve opening, coal flow and feedwater flow) are employed to control three outputs(i.e. output power, main steam temperature and main steam pressure). The step responses for the dynamic matrix control(DMC) are constructed using the three inputs by the three outputs under both pressure control mode. Piecewise models are built at selected operation points. In simulation, the output power follows load demand quickly and main steam temperature can be controlled around the setpoint closely in load tracking control. The simulation results show the effectiveness of the proposed methods.     

椭圆形排放口水下航行器热射流特征仿真对比

为研究排放口结构对水下航行器热射流的影响，以椭圆形排放口结构为重点，建立不同半径比的椭圆形排放口仿真模型，采用计算流体动力学方法对不同半径比椭圆形排放口水下航行器的热射流特征进行研究。计算结果表明：当椭圆形排放口的半径比发生变化时，其热射流温度衰减特征和水面红外特征变化明显，其中结构2椭圆形排放口的热射流浮升高度和水面最高峰值温度的下降幅度最大，水面纵向扩散最大距离抑制效果明显。