Polynomial Feedforward Artificial Neural Network

A new artificial neural network,i.e.,polynomial feedforward artificial neural network(PFANN), which has three layers(input layer,hidden layer and output layer) is presented. The neural activation functions of hidden layer and output layer are f(x)=x p and linear function, respectively. The learning method of hidden-output layer weights is the steepest descent method and the one of input-hidden layer weights is genetic algorithm(GA) . During the learning process, the error function is decreased monotonely. So the learning algorithm is convergent and the network ,which can approximate to arbitrary continuous function , is stable. Some applying samples of PFANN, which reveales the remarkable quality, are proposed,too.     

水泵性能曲线的正交多项式拟合

采用不规则点的正交多项式来拟合水泵性能曲线，可避免采用一般多项式拟合时可能出现病态方程组的弊端，文中采用Forsythe递推法生成正交多项式，根据显著性检验来确定拟合的多茂式次数，并在计算中佐以作图程序来进行直观分析。文中还结合具体的算例证明了壕种方法的实用性。     

金属带式无级变速传动研究中的发动机建模

提出以试验数据为依据，通过三次多项式曲线拟合建立发动机模型的思想。介绍了发动机模型的具体建立方法；利用VB编程语言编写了相应的数据处理程序，给出了程序流程图；建立了发动机的转矩模型、油耗率模型和转速调节模型，所建立的模型能较准确地反映发动机的工作状况，为进一步研究CVT传动系统的工作过程打下基础。     

基于状态转移算法优化的投影寻踪病虫害预测模型

为解决新疆加工番茄病虫害预测问题中样本数据的非线性和高维性等问题,采用投影寻踪回归模型对加工番茄病虫害预测进行研究。根据新疆某种植基地的样本数据,将投影寻踪回归模型与改进状态转移算法结合,建立了改进状态转移算法优化的基于Hermite多项式的投影寻踪病虫害预测模型。投影寻踪病虫害预测模型将高维的数据投影到低维空间,利用加入正交变换的状态转移算法优化得到投影方向和多项式系数。试验结果表明,利用该模型对新疆某种植基地2003—2008年的样本数据训练效果误差0.2,等级预测达到完全正确;对2009—2011年的病虫害等级预测准确率95%。基于改进状态转移算法的Hermite投影寻踪回归模型可靠性及预测精度很高,能有效的解决病虫害预测中存在的数据非线性、高维性等实际难题。该模型应用于加工番茄病虫害的预测具有一定的可行性和实用性。     

Optimum dietary protein level for maximum growth of juvenile olive flounder Paralichthys olivaceus (Temminck et Schlegel)

This experiment was conducted to determine the optimum dietary protein level for juvenile olive flounder Paralichthys olivaceus (Temminck et Schlegel) fed a white fish meal and casein‐based diets for 8 weeks. Olive flounder with an initial body weight of 4.1 ± 0.02 g (mean ± SD) were fed one of the six isocaloric diets containing 35%, 45%, 50%, 55% and 65% crude protein (CP) at a feeding rate of 4–5% of wet body weight on a dry‐matter basis to triplicate groups of 20 fish per aquarium. After 8 weeks of feeding, per cent weight gain (WG) and feed efficiency ratios of fish fed the 55% CP diet were not significantly higher than those from fish fed the 50% and 65% CP diets, but significantly higher than those from fish fed the 35% and 45% CP diets. Fish fed the 50%, 55% and 65% CP diets had significant higher specific growth rates than did fish fed the 35% and 45% CP diets; however, there was no significant difference among fish fed the 50%, 55% and 65% CP diets. The protein efficiency ratio was inversely related to the dietary protein level; that is, maximum efficiency occurred at the lowest dietary protein level. Broken‐line model analysis indicated that the optimum dietary protein level was 51.2 ± 1.8% for maximum weight gain in juvenile olive flounder. The second‐order polynomial regression analysis showed that the maximum WG occurred at 57.7% and it revealed that the minimum range of protein requirement was between 44.2% and 46.4%. These findings suggest that the optimum dietary protein level for maximum growth could be greater than 46.4%, but less than 51.2% CP in fish meal and casein‐based diets containing 17.0 kJ g^?1 energy for juvenile olive flounder.     

Relationship between corn leaf composition and high and low yield levels

Abstract

Corn yields and leaf samples were obtained from experimental plots receiving varying rates of N, P, and K. Yields were placed in two groups. One group consisted of yields < 134 bu/A, and a second group consisted of yields ≥ 134 bu/A. Within each group yields were regressed on leaf levels of N, P, K, Ca, Mg, B, Cu, Fe, Mn and Zn as independent variables in a second order polynomial. Using stepwise regression procedures the fit of the polynomial was determined for each set of observations, and statistical significance of various effects was determined.

The regression accounted for 71.5 percent of observed variability in yields in the < 134 bu/A category, and 53.2 percent of observed variability in yields in the ≥ 134 bu/A category. With statistical criteria used, twelve interactions were observed in both groupings, but the PFe interaction was the only one common to both analyses. Seven interactions with Mn were observed in the lower yield category, but only one in the upper category. One linear effect and three quadratic effects were observed in the lower yield category, and three linear effects and four quadratic effects were observed in the upper yield category.     

用多项式模拟烤烟赤星病病指趋势值的研究

 用多项式模式模拟预测了2001年玉溪红塔区烤烟赤星病病指的趋势值,并与实测值进行了误差分析;同时已模拟预测了2002年烤烟赤星病病指的发生趋势值,为烤烟生产的综合防治工作服务。     

关于微分多项式的正规族

证明了关于微分多项式的一个正规定则：设F为区域D上的亚纯函数族,a为非零有限值,k为正整数,如果对任意f∈F,满足：（i）f与L（f）IM分担a,（ii）f的零点的重级不小于k＋1,则F在D上正规.     

基于五次多项式的四足机器人轨迹规划

为达到四足机器人在地面上稳定、连续行走,不出现打滑、侧翻等现象的目的,提出一种基于walk静步态的五次多项式足端轨迹规划方法。首先利用D-H坐标法推导四足机器人单腿的运动学方程,由运动学逆解得到四足机器人足端轨迹和关节角之间的关系。将水平和竖直方向的约束方程代入五次多项式,分别求出支撑相和摆动相的足端轨迹方程。最后通过将轨迹方程代入MATLAB和ADAMS中进行仿真验证,实验结果验证了该轨迹规划方法的正确性。