Estimation of froth flotation recovery and collision probability based on operational parameters using an artificial neural network

An artificial neural network and regression procedures were used to predict the recovery and collision probability of quartz flotation concentrate in different operational conditions. Flotation parameters, such as dimensionless numbers (Froude, Reynolds, and Weber), particle size, air flow rate, bubble diameter, and bubble rise velocity, were used as inputs to both methods. The linear regression method shows that the relationships between flotation parameters and the recovery and collision probability of fl...     

人工神经网络在水工结构优化中的应用

采用人工神经网络方法 ,为水工结构优化提供初始方案 .通过典型的混凝土重力坝断面优化的例子 ,说明该方法能形成很好的初始方案 ,并容易推广到其他结构优化中     

Prediction of operational parameters effect on coal flotation using artificial neural network

Artificial neural network procedures were used to predict the combustible value (i.e. 100-Ash) and combustible recovery of coal flotation concentrate in different operational conditions. The pulp density, pH, rotation rate, coal particle size, dosage of collector, frother and conditioner were used as inputs to the network. Feed-forward artificial neural networks with 5-30-2-1 and 7-10-3-1 arrangements were capable to estimate the combustible value and combustible recovery of coal flotation concentrate respectively as the outputs. Quite satisfactory correlations of 1 and 0.91 in training and testing stages for combustible value and of 1 and 0.95 in training and testing stages for combustible recovery prediction were achieved. The proposed neural network models can be used to determine the most advantageous operational conditions for the expected concentrate assay and recovery in the coal flotation process.     

水文预报的神经网络模式分类预报方法

基于时间序列的相空间扩维思想,利用神经网络的感知器分类原理,提出一种水文预报方法.直接从水文单因素时间序列滑动生成多维空间的一个相型分布中寻求水文现象变化的规律.并利用湖北省宜昌市沮漳河洪水资料对该方法做了初步验证.     

基于粗糙集-BP神经网络的浮选槽液位预测

针对反浮选过程中浮选槽液位指标难以建立精确的数学模型、常规检测方法不能有效控制问题,提出一种将粗糙集与BP神经网络理论相结合方法[1],建立反浮选液位软测量模型。从浮选过程积累的数据中获取过程知识,通过粗糙集属性约简对训练样本数据进行处理,根据结果确定BP网络的输入、输出、隐层神经元数,从得到的优化设定自动更新浮选槽液位控制回路的设定值,避免了人工控制的不稳定性和不精确性。此方法应用于某浮选厂,满足了液位预测要求的精度,在液位控制、经济指标提高及浮选过程稳定等方面取得了明显的效果。     

A fuzzy neural network evolved by particle swarm optimization

A cooperative system of a fuzzy logic model and a fuzzy neural network(CSFLMFNN)is proposed,in which a fuzzy logic model is acquired from domain experts and a fuzzy neural network is generated and prewired according to the model.Then PSO-CSFLMFNN is constructed by introducing particle swarm optimization(PSO)into the cooperative system instead of the commonly used evolutionary algorithms to evolve the prewired fuzzy neural network.The evolutionary fuzzy neural network implements accuracy fuzzy inference without rule matching.PSO-CSFLMFNN is applied to the intelligent fault diagnosis for a petrochemical engineering equipment,in which the cooperative system is proved to be effective.It is shown by the applied results that the performance of the evolutionary fuzzy neural network outperforms remarkably that of the one evolved by genetic algorithm in the convergence rate and the generalization precision.     

多子群粒子群集成神经网络

为了使参加神经网络集成的个体差异度较大,从而提高网络集成的泛化能力,本文提出一种新的基于多子群粒子群算法的神经网络集成方法.每个子群通过补充差异度独立训练出一批神经网络,从每个子群中选择一个最优个体参加网络集成,实验使用了UCI标准数据集.实验证明,该算法的识别能力要好于Boosting、Bagging等传统方法.     

基于神经优化的最大熵图像重建算法

提出了一种基于Ｈｏｐｆｉｅｌｄ神经网络优化的图像重建算法。将图像重建问题转化为ＨＮＮ优化问题,取重建图像熵函数最大以及原始投影与再投影之间的误差平方和最小作为图像重建的优化目标,作为能量函数构造连续型ＨＮＮ模型,由ＨＮＮ能量函数极小化可得到重建问题的优化解。     

BP神经网络在入侵检测系统中的应用及优化

BP神经网络运用于入侵检测系统有很多优点，但是也存在一些缺点，如执行速度比较慢的问题等．常用的LMBP算法，虽然收敛速度很快，但是应用于入侵检测系统执行速度仍然不能满足要求．结合KDD99数据集，选取适当的数据，通过加入一些限制条件，对LMBP算法进行了优化．通过实际计算，比较算法优化前后的计算结果，验证了优化算法是有效的．优化后的算法比较明显的提高了BP神经网络应用于入侵检测系统时的执行速度，具有一定的实用价值．     

人工神经网络中隐含层节点与训练次数的优化

目前构建定量构效关系人工神经网络模型中隐含层节点数和网络训练次数大多是依靠试验方法来确定,针对该方法运算工作量较大、模型质量和预测精度没有保证等问题,通过编写程序获得有关网络的预测精度和百分误差与网络隐含层节点数和训练次数之间关系的大量数据,采用Matlab语言分别绘制预测精度和百分误差与网络隐含层节点数和训练次数之间的三维关系图,从图中可以很容易判断出达到最佳预测精度和最小百分误差的隐含层节点数和训练次数。该方法和技术从根本上提高了选择人工神经网络隐含层节点数和训练次数方法的效率。     

Using particle swarm optimization algorithm in an artificial neural network to forecast the strength of paste filling material

In order to forecast the strength of filling material exactly, the main factors affecting the strength of filling material are analyzed. The model of predicting the strength of filling material was established by applying the theory of artificial neural networks. Based on cases related to our test data of filling material, the predicted results of the model and measured values are compared and analyzed. The results show that the model is feasible and scientifically justified to predict the strength of filling material,which provides a new method for forecasting the strength of filling material for paste filling in coal mines.     

基于神经网络优化的迭代学习控制方法的研究及应用

聚合反应的动态特性具有时变性、非线性等特点,应用传统的控制方法已不能满足实际的控制要求,且达不到需要的控制精度,急需提出一种先进的控制方法.本文提出了一种新的基于神经网络优化的迭代学习控制方法,介绍了由迭代学习控制理论设计迭代学习控制器,提出用神经网络对控制器参数进行优化计算,找出最优的学习增益;并将该方法应用于ABS树脂聚合反应过程的温度控制中,仿真结果表明了该方法的有效性,且能在较少的迭代次数下,以最快的收敛速度、较高的跟踪精度逼近期望轨迹.     

Optimizing neural network forecast by immune algorithm

1 INTRODUCTIONThe forecast of complicated nonlinear relation-ships among multi-factor ti me series can be ex-pressed as a multivariate ti me series analysis ,which si multaneously observes the dynamic processof a group of (multi-di mension) correlative randomvariables as a whole , while traditional linear fore-casting methods such as multiple regression mod-el[1], GM(1 ,N) model[2]seemto have many bar-riers on solvingthis complex nonlinearissue amongthese multi-factor ti me series .Artifi…     

基于神经网络和遗传算法的锅炉燃烧优化方法

针对锅炉燃烧控制系统送风调节系统存在的弊端,遵照火电厂锅炉燃烧既要提高效率又要降低污染物排放的要求,对神经网络和遗传算法在火电厂锅炉燃烧优化中的应用进行了研究。首先借助燃烧特性试验数据,建立了火电厂锅炉燃烧特性的神经网络模型,然后应用遗传算法寻找送风调节系统最佳氧量设定值,进而调节送风量,实现锅炉燃烧的整体优化。仿真结果表明:应用该方法指导锅炉燃烧,不仅能使锅炉节能,还能降低排放的烟气中氮氧化物的含量,减少对环境的污染。     

Beneficiation of the gold bearing ore by gravity and flotation

Gold concentration usually consists of gravity separation, flotation, cyanidation, or the combination of these processes. The choice among these processes depends on the mineralogical characterization and gold content of the ore. Recently, the recovery of gold using gravity methods has gained attention because of low cost and environmentally friendly operations. In this study, gold pre-concentrates were produced by the stepwise gravity separation and flotation techniques. The Knelson concentrator and conventional flotation were employed for the recovery of gold. Gold bearing ore samples were taken from Gümüshane Region, northern east part of Turkey. As a result of stepwise Knelson concentration experiments, a gold concentrate assaying around 620 g/t is produced with 41.4wt% recovery. On the other hand, a gold concentrate about 82 g/t is obtained with 89.9wt% recovery from a gold ore assaying 6 g/t Au by direct flotation.     

改进BP神经网络算法在机器人轨迹优化中的研究

针对未知的全局环境,将整体任务分解为环境信息已知的一系列中间任务,即将整个机器人的运动分解成一系列直线运动,降低神经网络函数复杂度,并且将障碍物封装为简单图形,以减少整个运动过程中的中间任务的个数,进一步降低神经网络函数复杂度．利用BP神经网络高速并行计算的优点,建立神经网络函数,提出一种实时性较高的求取函数的负梯度方向的方法,控制机器人快速高效地完成中间任务,从而驱使机器人到达目标点并进行仿真．     

A chaos genetic algorithm for optimizing an artificial neural network of predicting silicon content in hot metal

A genetic algorithm based on the nested intervals chaos search (NICGA) has been given. Because the nested intervals chaos search is introduced into the NICGA to initialize the population and to lead the evolution of the population, the NICGA has the advantages of decreasing the population size, enhancing the local search ability, and improving the computational efficiency and optimization precision. In a multi-layer feed forward neural network model for predicting the silicon content in hot metal, the NICGA was used to optimize the connection weights and threshold values of the neural network to improve the prediction precision. The application results show that the precision of predicting the silicon content has been increased.     

三维重建的神经网络算法

本文根据Ｆｕｃｈｓ等人的结果,将三维重建问题转化为一有向不图的最小费用路径搜索问题,然后构造了一反馈型神经网络来求解这一组合优化问题,并对如何使网络跳出某些局部极小点,收敛到可行解给了了一解决方法,软件模拟结果证实了这一技术的可行性。     

A hybrid particle swarm optimization approach with neural network and set pair analysis for transmission network planning

Transmission network planning (TNP) is a large-scale, complex, with more non-linear discrete variables and the multi-objective constrained optimization problem. In the optimization process, the line investment, network reliability and the network loss are the main objective of transmission network planning. Combined with set pair analysis (SPA), particle swarm optimization (PSO), neural network (NN), a hybrid particle swarm optimization model was established with neural network and set pair analysis for transmission network planning (HPNS). Firstly, the contact degree of set pair analysis was introduced, the traditional goal set was converted into the collection of the three indicators including the identity degree, difference agree and contrary degree. On this bases, using shi(H), the three objective optimization problem was converted into single objective optimization problem. Secondly, using the fast and efficient search capabilities of PSO, the transmission network planning model based on set pair analysis was optimized. In the process of optimization, by improving the BP neural network constantly training so that the value of the fitness function of PSO becomes smaller in order to obtain the optimization program fitting the three objectives better. Finally, compared HPNS with PSO algorithm and the classic genetic algorithm, HPNS increased about 23% efficiency than THA, raised about 3.7% than PSO and improved about 2.96% than GA.     

多氯酚QSAR建模的人工神经网络信息流分析

针对目前采用人工网络（ANN）建模中存在的问题,以多氯酚的定量构效关系（QSAR）人工神经网络模型研究为基础,开展了ANN的信息流分析,对该网络的连接权值和阈值进行了分析、比较,找出了影响网络输出的稳含层的主要节点;并由这些隐含层节点进一步确定出主要输入层节点;经对网络内部信息传播与分配过程的分析,得到ANN模型信息流流径分布图。