基于IGS和SVM的烧结返矿量智能集成预测模型

针对烧结返矿量难以进行有效预测的问题,提出一种智能集成预测模型.首先利用改进灰色系统和支持向量机两个单一模型分别对返矿量进行预测;然后基于预测精度的数学期望和标准差,通过求取最优加权系数,建立烧结返矿量智能集成预测模型进行返矿量集成预测.运行结果表明,该集成模型的预测精度高于单一模型,能有效地对返矿量进行预测.     

电离层延迟模型改进研究

由于电离层延迟会对卫星的定位精度产生很大的影响,传统的电离层延迟Klobuchar模型的预报精度不高,而且其对夜间电离层延迟以及对延迟峰值出现时刻的预报存在明显的不合理性。因此必须进行修正,根据实测电离层延迟数据的特点对传统的Klobuchar模型进行了改进,并利用非线性最小二乘迭代算法对两种模型进行参数拟合,采用拟合参数计算两种模型的拟合值。通过比较两种拟合值相对于实际测量值的精度,可以得出:改进模型比传统模型提高了电离层延迟预报的精度,并且改善了传统延迟模型预报中存在的不合理性。     

基于灰色预测的氯化聚乙烯生产过程控制

氯化聚乙烯作为一种新型的高分子材料,被广泛地应用于塑料和橡胶的生产中。反应釜温度是其过程控制的主要被控参数,但是由于其存在纯滞后和大惯性等特点,普通的串级控制方式很难得到良好的控制效果。本文基于灰色模型的原理,设计了灰色PID控制器。仿真表明,该控制器具有良好的抗干扰性和稳定的鲁棒性,可有效改善系统的控制效果。     

基于PCC的神经网络PID控制器设计

提出了一种基于可编程计算机控制器（PCC）的神经网络PID控制器,以PCC为硬件,利用神经网络逼近任意非线性函数的功能,实现了PID控制算法中三个参数Kp、Ki、Kd的在线调整。并利用神经网络模型对被控对象的输出值进行预测,根据预测值对神经网络各层中的加权系数进行在线修正,同时引入了带死区的控制算法。该设计方案具有调节速度快、适应能力强、可靠性高等优点。实验结果表明,该控制器具有强抗扰、响应快、鲁棒性好等特点。     

A class of new transforms tailored for the hypergeometric series

A new class of transforms, tailored for the hypergeometric series, is proposed and tested on a number of series. From a knowledge of a given number of terms of a sequence, these transforms reproduce the functions with a better accuracy than the Levin-like transforms. Though there exists a correlation between the approximative power of a rational approximant and its ability to predict the leading terms of a series, there are exceptions to this, especially in the case of divergent series. The new transforms can, in most cases, predict a number of extra terms not used in the construction of the approximants.     

支持向量回归机在物价预测上的应用

把时间序列SVM预测模型应用于物价指数的预测,采用逐步递归的方法进行,同时注意尽量减少训练样本的浪费和充分挖掘SVM模型适合短期预测的潜力.分析结果表明,无论是拟合情况,还是预测值的检验和物价指数的实际规律来看,都有很高的精度,可以作为物价指数预测的一种行之有效的方法.     

基于 GBDT 的铁路事故类型预测及成因分析

运用数据挖掘技术进行铁路事故类型预测及成因分析, 对于建立铁路事故预警机制具有重要意义. 为此, 本文提出一种基于梯度提升决策树(Grandient boosting decision tree, GBDT)的铁路事故类型预测及成因分析算法. 针对铁路事故记录数据缺失的问题, 提出一种基于属性分布概率的补全算法, 最大程度保持原有数据分布, 从而降低数据缺失对事故类型预测造成的影响. 针对铁路事故记录数据类别失衡的问题, 提出一种集成的GBDT模型, 完成对事故类型的鲁棒性预测. 在此基础上, 根据GBDT预测模型中特征重要度排序, 实现事故成因分析. 通过在开放数据库上进行实验, 验证了本文模型的有效性.     

Improved inter-view correlations for low complexity MV-HEVC

The more advanced multi-view extension, MV-HEVC, effectively exploits visual similarities between multi-view videos and enables high compression efficiency. Each view in the multi-view sequence depends on the captured scene, the distance between cameras and recording angles. Increasing the distance between dependent viewpoints generates an inter-view disparity. This impacts the inter-view similarities, affects the disparity estimation and further increases the computational complexity of the MV-HEVC encoder. In this paper, an efficient earlier disparity estimation is proposed for low complexity MV-HEVC. This algorithm is based on reducing the complexity of disparity estimation by eliminating the inter-view offset. Moreover, the inter-view similarities are controlled by considering the reliability of each coding unit size in the search range. This reliability is estimated by reducing the number of searching points within a new limited window. For reliable motion estimation, we further proposed an earlier decision of coding units splitting in the dependent views according to those in the reference views. Experimental results show that the proposed algorithm can achieve an average encoding time saving of 20.37%–40,61% with marginal performance degradation.     

A modified θ projection model for constant load creep curves-II. Application of creep life prediction

To minimize the deviation of the predicted creep curves obtained under constant load conditions by the original θ projection model, a new modified version that can be expressed by $\epsilon ={\theta }_1\left(1-e^{-{\theta }_{2}t}\right)+{\theta }_3\left(e^{{\theta }_4{e}^{{\theta }_5\epsilon }t}-1\right)$, was derived and experimentally validated in our last study. In the present study, the predictive capability of the modified θ projection model was investigated by comparing the simulated and experimentally determined creep curves of K465 and DZ125 superalloys over a range of temperatures and stresses. Furthermore, the linear relationship between creep temperature and initial stress was extended to the 5-parameter model. The results indicated that the modified model could be used as a creep life prediction method, as it described the creep curve shape and resulted in predictions that fall within a specified error interval. Meanwhile, this modified model provides a more accurate way of describing creep curves under constant load conditions. The limitations and future direction of the modified model were also discussed. In addition, this modified θ projection model shows great potential for the evaluation and assessment of the service safety of structural materials used in components governed by creep deformation.     

Multi-Step H∞ Generalized Predictive Control

An H generalized control law is derived which has properties in common with both the GH and GPC control laws. The stability and robustness properties are the same as for a GH optimal controller but a cost on future predictive error and control action is dealt with in the same manner as for the GPC or model based predictive control laws. This type of long range predictive control algorithm has stability and robustness problems which are much improved through the H formalism. The general two-block problem solution is considered first and this is then reduced to a novel one-block multi-step GH control law which is very attractive computationally. This latter algorithm has all the benefits of one-block H controllers and yet the equations can be written in a form where GPC type constrained solutions can be obtained.