Composite iterative learning controller design for gradually varying references with applications in an AFM system简

Learning control for gradually varying references in iteration domain was considered in this research, and a composite iterative learning control strategy was proposed to enable a plant to track unknown iteration-dependent trajectories. Specifically, by decoupling the current reference into the desired trajectory of the last trial and a disturbance signal with small magnitude, the learning and feedback parts were designed respectively to ensure fine tracking performance. After some theoretical analysis, the judging condition on whether the composite iterative learning control approach achieves better control results than pure feedback contro! was obtained for varying references. The convergence property of the closed-loop system was rigorously studied and the saturation problem was also addressed in the controller. The designed composite iterative learning control strategy is successfully employed in an atomic force microscope system, with both simulation and experimental results clearly demonstrating its superior performance.     

Algorithm design and application of novel GM-AGC based on mill stretch characteristic curve简

As the spring equation is limited to the accuracy of mill stiffness and the linearity of the mill spring curve, the traditional gaugemeter automatic gauge control （GM-AGC） system based on spring equation cannot meet the requirements of practical production. In allusion to this problem, a kind of novel GM-AGC system based on mill stretch characteristic curve was proposed. The error existing in calculating strip thickness by spring equation were analyzed first. And then the mill stretch characteristic curve which could effectively eliminate the influence of mill stiffness was described. The novel GM-AGC system has been applied successfully in a hot strip mill, the application results show that the thickness control precision is improved significantly, with the novel GM-AGC system, over 98.6% of the strip thickness deviation of 3.0 mm class can be controlled within the target tolerances of ±20 μm.     

Two-way Markov random walk transductive learning algorithm简

Researchers face many class prediction challenges stemming from a small size of training data vis-a-vis a large number of unlabeled samples to be predicted. Transductive learning is proposed to utilize information about unlabeled data to estimate labels of the unlabeled data for this condition. This work presents a new transductive learning method called two-way Markov random walk （TMRW） algorithm. The algorithm uses information about labeled and unlabeled data to predict the labels of the unlabeled data by taking random walks between the labeled and unlabeled data where data points are viewed as nodes of a graph. The labeled points correlate to unlabeled points and vice versa according to a transition probability matrix. We can get the predicted labels of unlabeled samples by combining the results of the two-way walks. Finally, ensemble learning is combined with transductive learning, and Adboost.MH is taken as the study framework to improve the performance of TMRW, which is the basic learner. Experiments show that this algorithm can predict labels of unlabeled data well.     

Stability analysis for natural slope by kinematical approach

The stability of natural slope was analyzed on the basis of limit analysis. The sliding model of a kind of natural slope was presented. A new kinematically admissible velocity field for the new sliding model was constructed. The stability factor formulation by the upper bound theorem leads to a classical nonlinear programming problem, when the external work rate and internal energy dissipation were solved, and the constraint condition of the programming problem was given. The upper bound optimization problem can be solved efficiently by applying a nonlinear SQP algorithm, and stability factor was obtained, which agrees well with previous achievements.     

A hardware-based algorithm for virtual machine provisioning in a private cloud

Cloud computing is becoming a key factor in the market day by day. Therefore, many companies are investing or going to invest in this sector for development of large data centers. These data centers not only consume more energy but also produce greenhouse gases. Because of large amount of power consumption, data center providers go for different types of power generator to increase the profit margin which indirectly affects the environment. Several studies are carried out to reduce the power consumption of a data center. One of the techniques to reduce power consumption is virtualization. After several studies, it is stated that hardware plays a very important role. As the load increases, the power consumption of the CPU is also increased. Therefore, by extending the study of virtualization to reduce the power consumption, a hardware-based algorithm for virtual machine provisioning in a private cloud can significantly improve the performance by considering hardware as one of the important factors.     

Comparative analysis of different methods for image enhancement

Image enhancement technology plays a very important role to improve image quality in image processing. By enhancing some information and restraining other information selectively, it can improve image visual effect. The objective of this work is to implement the image enhancement to gray scale images using different techniques. After the fundamental methods of image enhancement processing are demonstrated, image enhancement algorithms based on space and frequency domains are systematically investigated and compared. The advantage and defect of the above-mentioned algorithms are analyzed. The algorithms of wavelet based image enhancement are also deduced and generalized. Wavelet transform modulus maxima（WTMM） is a method for detecting the fractal dimension of a signal, it is well used for image enhancement. The image techniques are compared by using the mean（μ）,standard deviation（？）, mean square error（MSE） and PSNR（peak signal to noise ratio）. A group of experimental results demonstrate that the image enhancement algorithm based on wavelet transform is effective for image de-noising and enhancement. Wavelet transform modulus maxima method is one of the best methods for image enhancement.     

基于并行模板的手写体汉字串行细化算法

针对并行细化算法中细化模板较多,细化不全和串行算法的骨架非对称现象,提出了一种基于并行模板的手写体汉字串行细化算法。该算法改变了剥离边界的方向,引入了4个新的并行细化模板,根据串行算法特点和书写运笔规律,采用逆向扫描法遍历整个图像,通过与其他的模板细化算法进行实验比较,得出该算法在细化手写体汉字时,模板较少,速度较快,细化完全,骨架对称性较好的结论。     

线性互补约束问题的一个SQP算法

针对非线性规划一些经典的算法一般不能直接应用到均衡问题上来的缺点,通过一个处处连续可微的互补函数以及光滑近似法的思想,把互补约束均衡问题转化为一光滑非线性规划问题,提出一个SQP算法求解该光滑非线性规划问题,并给出了算法的数值解,数值实验结果表明算法具有很好的收敛速度和寻优结果.     

SADE-ELM电容层析成像流型辨识算法

针对电容层析成像反问题流型识别较难的问题,提出了一种新的ECT流型辨识算法——差分演化优化极端学习机算法,进而提出了基于自适应差分演化优化极端学习机(Sa DEEML)的ECT辨识算法.在论述极端学习机算法的基础上,结合差分演化算法对极端学习机算法进行优化,自适应差分演化算法中的关键参数,通过训练得到各类流型的分类器的参数,构造分类器进行精准与快速分类.实验结果表明:该算法能有效克服极端学习机算法的缺点并提高了局部与全局收敛能力,通过与BP、SVM算法比较,该算法具有竞争力,并为电容层析成像流型辨识的研究提供了新算法.     

基于约束优化问题的人工鱼群算法及其改进

在人工鱼群算法基础上,对人工鱼群算法进行改进,结合遗传算法提出的适应度函数来解决约束优化问题.具体表现在改进了人工鱼的觅食行为,另外引入了吞噬行为以便加快收敛速度,得到更优的适应度值.仿真结果表明改进的人工鱼群算法在解决约束优化问题时,具有收敛速度快、适应度值优、全局寻优性能强等优点.改进的人工鱼群算法较之基本人工鱼群算法具有更好的性能.