改进的运动模糊矿石图像复原方法

针对矿石粒度检测过程中相机与目标矿石之间相对运动造成的图像模糊问题,提出一种模糊矿石图像清晰化的方法。采用维纳滤波进行图像复原是常用方法,因此,采集模糊矿石图像,建立图像复原模型。首先通过傅里叶变换构造频谱图,几何分析法估计模糊角度和微分自相关法估计模糊长度;采用无参考图像清晰度算法校准并确定模糊角度及长度,构造出最为近似的点扩散函数(PSF)参数;再以快速K值估计法获得最佳噪信比值,最后用维纳滤波处理所求,实现运动模糊图像的复原。实验结果表明:该方法能估计并矫正模糊矿石图像的模糊长度和角度,其复原效果比传统方法更优,准确度高、稳定性强,具有一定的实际工程意义。

模糊聚类的马尔可夫方法及神经网络法在铁矿石需求预测中的应用

以徐州市生铁产量及其与铁矿石需求量之间的关系为基础,通过模糊聚类的马尔可夫方法、神经网络法进行预测,发现两种方法所得结果基本吻合,旨在为决策者提供参考依据。

Identification on rock and soil parameters for vibration drilling rock in metal mine based on fuzzy least square support vector machine

A single freedom degree model of drilling bit-rock was established according to the vibration mechanism and its dynamic characteristics. Moreover, a novel identification method of rock and soil parameters for vibration drilling based on the fuzzy least squares(FLS)-support vector machine(SVM) was developed, in which the fuzzy membership function was set by using linear distance, and its parameters, such as penalty factor and kernel parameter, were optimized by using adaptive genetic algorithm. And FLS-SVM identification on rock and soil parameters for vibration drilling was made by changing the input/output data from single freedom degree model of drilling bit-rock. The results of identification simulation and resonance column experiment show that relative error of natural frequency for some hard sand from identification simulation and resonance column experiment is 1.1% and the identification precision based on the fuzzy least squares-support vector machine is high.

THE CLASSIFICATION OF THE SURROUNDINGS OF COAL MINING ROADWAYS

This paper introduces the calculation of the deformation of the surroundings of roadways and the division of surroundings into 5 levels by means of fuzzy integral assess matrix, which serves as the scientific basis for selecting supporting pattern of roadways and determining the parameters of support.