基于改进LS-SVM的制粉出力软测量建模

针对火电厂双进双出钢球磨煤机制粉出力较难直接测量的问题,拟建立其最小二乘支持向量机(LS-SVM)模型实现软测量.由于LS-SVM算法缺乏“稀疏性”,采用经Remez算法优化后的FIR数字滤波器与LS-SVM相结合,形成一种新的软测量建模方法——FIR-LSS-VM.仿真结果表明,经Remez优化后的FIR滤波效果更好.将该软测量方法应用于火电厂双进双出钢球磨煤机直吹式制粉系统,通过现场实验对比可知,改进的软测量模型学习速度较快且误差较小,更加适用于在线学习.     

矿渣粉磨智能控制系统的研究及应用

立磨矿渣粉磨系统经常出现立磨振动大,磨机负荷波动频繁等故障,同时立磨系统又是一个多变量、强耦合、非线性系统.针对以上问题,采用专家系统和模糊控制设计了立磨优化控制方案,利用现场实际数据辩识出立磨系统近似数学模型,并以此通过仿真实验验证了立磨优化控制算法.通过加入阈值和限幅控制将优化控制算法应用于现场实际控制中,结果表明立磨优化控制比人工手动控制的料层厚度波动范围和磨机振动值小,立磨没有出现异常振动停机.     

数字化钢球磨机制粉系统

本文简要介绍了在火电厂有着广泛应用的钢球磨煤机制粉系统的自动控制以及在线决策、在线自寻优问题,通过对原有系统的自动化改造．达到系统稳定安全运行、节能降耗的目的。     

球磨机制粉系统的解耦PID控制

中储式钢球磨煤机制粉系统具有多变量、强耦合、大惯性等特点,使用常规的PID控制回路进行控制无法取得理想效果。针对这种情况,本文提出了一种解耦PID控制策略。该方案在PID回路中加入解耦补偿器,以实现对多变量的解耦控制。仿真试验表明,该策略有效地解决了多变量的耦合问题,取得了较好的控制效果。     

球磨机制粉系统参数自整定PID解耦控制器

针对球团厂钢球磨煤制粉系统具有多变量强耦合、时滞、非线性以及生产工况变化大的复杂性,已有的控制方法难以满足其控制要求的现状,以磨煤机的数学模型为依据,提出了一种新型的多变量PID解耦控制策略.该控制系统由比例因子模糊自整定单元、PID控制器和基于对角矩阵法的解耦补偿器组成;PID控制器参数由粒子群算法进行优化.仿真结果表明,所建模型和所提控制方法的有效性.该方法可在大范围内解决系统的非线性和强耦合问题,具有较强的鲁棒性和运行工况适应性.     

基于人工神经网络的中速磨存煤量软测量方法

中速磨直吹式制粉系统是火力发电厂锅炉运行稳定的一个重要环节,其磨机内存煤量能否获得可靠监测对发电厂的自动控制、经济运行和安全生产均有十分重要的意义。通过基于工艺机理知识,分析了中速磨直吹式制粉系统筒内存煤量的各相关因素,提出了利用人工神经网络进行存煤量软测量的方法,给出了基于BP改进算法的回归神经网络与延时神经网络综合模型,为中速磨存煤量的检测提供了新方法。离线训练与计算机仿真结果证实了软测量方法的可行性。     

闭路球磨机的非线性动态辨识及最佳循环负荷量求取

1 引言磨机是水泥厂的主机之一,其能耗大,效率低。据有关文献介绍,水泥生产中约有70%的能耗用于粉磨作业,而用于粉磨功(有用功)的能耗仅占磨机耗电量的2%～10%,其余为损耗。因此,合理使用磨机与控制磨机负荷并使之在最佳工作状态下运转,是提高磨机粉磨效率和降低能耗的重要途径。粉磨系统具有非线性、大滞后和慢时变参数等特性,其最佳负荷即最大生产能力随着生产过程中原料的粒度分布,易磨性及研磨体研     

基于粗糙集理论的煤磨故障诊断模型

针对水泥煤磨系统的常见故障,将基于粗糙集理论的数据挖掘技术应用于水泥煤磨故障诊断。结合具体的数据,采用粗糙集的方式,建立了故障诊断系统,通过该系统对识别数据进行故障诊断实验,表明该系统能有效地对煤磨堵仓和钢球小故障进行诊断。诊断结果的精度可以满足现场应用的要求。     

基于模糊前馈解耦的变风量空调控制方法研究

变风量(VAV)空调系统具有多变量、多回路耦合、大滞后、强非线性特点;采用数学建模和仿真实验法,针对送风温度、送风管道静压、室内温度和室内二氧化碳浓度4个相互耦合的主要被控量,通过建立热量平衡和物质传输机理模型和传递函数矩阵,设计基于神经网络的前馈模糊解耦方法;所提出的方法既利用了数学模型可清晰描述被控量之间耦合关系的优点,又避免了传统解耦算法控制效果过于依赖建模精度等问题,仿真和实验结果表明系统获得了良好的控制性能。     

模糊控制在磨机出入口温度上的应用

矿渣微粉一般采用集烘干、粉磨、选粉于一体的立磨工艺,该工艺对磨机出入口温度的控制要求较高。入口温度过高容易使磨辊内润滑油变质,过低则满足不了烘干效果;特别是出口温度,过高容易烧坏价格昂贵的收粉布袋,过低容易导致产品的水分过高,且产品细度很难控制,从而导致产品不合格。     

燃料电池温度模糊自适应控制

如何控制燃料电池温度性能是燃料电池的一个重要问题。首先基于模糊辨识建模方法建立质子交换膜燃料电池温度性能的T-S模型。模型结构简单,精度高,方便地应用于质子交换膜燃料电池系统控制中。其次针对该模型设计电堆温度的模糊自适应控制器。最后在Matlab平台进行仿真,模糊自适应控制器在较大幅度变化的系统参数下都得到较好的控制性能,证明模糊自适应控制系统具有很好的鲁棒性和良好的控制品质,能够满足质子交换膜燃料电池温度控制系统的要求。     

Intelligent control of braking process

Intelligent modeling, prediction and control of the braking process are not an easy task if using classical modeling techniques, regarding its complexity. In this paper, the new approach has been proposed for easy and effective monitoring, modeling, prediction, and control of the braking process i.e. the brake performance during a braking cycle. The context based control of the disc brake actuation pressure was used for improving the dynamic control of braking process versus influence of the previous and current values of the disc brake actuation pressure, the vehicle speed, and the brake interface temperature. For these purposes, two different dynamic neural models have been developed and integrated into the microcontroller. Microcontrollers are resource intensive and cost effective platforms that offer possibilities to associate with commonly used artificial intelligence techniques. The neural models, based on recurrent dynamic neural networks, are implemented in 8-bit CMOS microcontroller for control of the disc brake actuation pressure during a braking cycle. The first neural model was used for modeling and prediction of the braking process output (braking torque). Based on such acquired knowledge about the real brake operation, the inverse neural model has been developed which was able to predict the brake actuation pressure needed for achieving previously selected (desired) braking torque value in accordance with the previous and current influence of the pressure, speed, and the brake interface temperature. Both neural models have had inherent abilities for on-line learning and prediction during each braking cycle and an intelligent adaptation to the change of influences of pressure, speed, and temperature on the braking process.     

重油输送系统基于流量-压力的建模与仿真

应用于阳极焙烧的重油输送的精确控制是一个典型交叉耦合的复杂控制问题，它对炭阳极焙烧质量起关键作用，输送过程中涉及到重油流量、压力和温度三个参量的变化。为了简化所建模型，假定重油输送过程中油温恒定，这样在建模过程中就剩下重油流量与压力之间的耦合关系。为了实现阳极焙烧重油输送的精确控制，通过对阳极焙烧重油输送工作机理的分析。以现场实际采集的数据为基础，采用管道流体传输的物理机理建立对象的模型的方法，提出了一种阳极焙烧萤油输送的基于流量-压力精确控制的数学模型，经仿真验证了该模型的有效性。     

BP神经网络在煮糖结晶过程建模的应用

煮糖结晶过程是一个非常复杂的物理、化学过程,其内部机理复杂,各变量间相互耦合,很难建立机理模型.利用神经网络具有逼近任意复杂非线性函数的能力,将神经网络应用到煮糖结晶过程建模中.模型以煮糖结晶过程中糖液的过饱和度为输出,以影响过饱和度的温度,压力和各阀门的开度等等作为网络的输入,采用改进的BP算法,建立了煮糖结晶过程的动态响应模型.给出了模型的结构、网络学习流程图和仿真结果.仿真结果表明,此BP网络模型能较好地解决煮糖结晶过程的建模问题.     

硅压阻式压力传感器的温度补偿

分析了硅压阻式压力传感器的温度误差来源，给出了通过长时间实验的温度与传感器输出电阻的数据，提出了对此类传感器温度误差的测量和修正方法，从而保证此类传感器在大温度范围内得到高精度使用。     

基于改进的BP算法的过热汽温建模

采用机理建模方法获得热工被控对象的数学模型非常困难,相比之下,利用电厂采集到的大量历史数据,不用人为地加入特殊扰动信号,来建立火电厂热工过程的动态数学模型,是很有发展前途的方向。本文利用电厂采集到的现场数据,应用加入二次动量项和改进激活函数的BP算法,对电站锅炉的过热汽温建模进行了研究。仿真结果表明,改进后的BP算法在误差精度、收敛速度、泛化能力等方面明显优于标准BP算法,对过热汽温的建模结果令人满意。     

直接甲醇燃料电池的混合模型及其仿真分析

讨论了一种直接甲醇燃料电池基于机理模型和神经网络模型的混合模型构建方法,利用人工神经网络的非线性逼近能力,对机理模型的不精确性进行有效补偿。混合模型中的神经网络模型根据输入变量和补偿量训练后与机理模型相结合,对电池电压提供了良好的近似预测。     

Positioning error guarantee method with two-stage compensation strategy for aircraft flexible assembly tooling

Flexible assembly tooling is an important manufacturing equipment for aircraft products, the positioning error of its end locating effectors determines the final assembly precision directly. To realize its accurate positioning, a precise and effective error compensation method based on two-stage strategy, is proposed. Based on the structure and working conditions of locating unit, an assembly measuring field is set up, for acquiring the actual geometric error status and the various time-varying assembly parameters in practical assembly site. Considering the transfer and accumulation of actual geometric error and deformation error, the positioning error mechanism model at the pre-compensation stage is proposed. Then with the actual assembly conditions and the measured training/testing data set of the positioning error, the SLFN motion prediction model for the locating unit is proposed, aiming at clarifying the nonlinear and strong coupling relationship between the input motion value and the output positioning error. And the SLFN data model has a good fitting and prediction ability. Combining the advantages of error mechanism modeling and measurement data modeling, a semi-mechanism model is proposed at the accurate compensation stage. Where the positioning error is predicted with SLFN model for a given input motion value, acting as an error generator. Experiment results showed that the final measured average error is 0.036 mm, having an enhancement ratio of 25.77 % contrasting with the error mechanism model. This accurate modification and compensation solution makes the relationship between causal factors and output positioning error more comprehensive, and the overall effect is more consistent with the practical situation, i.e. a higher precision value and an accurate input/output relationship can be gained.     

Optimal control of a nonlinear induction heating system using a proper orthogonal decomposition based reduced order model

This paper considers the problem of obtaining control inputs for induction heating processes involving nonlinearities and nontrivial geometries. The considered system falls within the much broader class of multiphysics or coupled, distributed parameter systems that are well known for the challenges they pose towards modeling, simulation and control. The approach is based on the versatile finite element method (FEM) widely popular in industry for modeling of such systems. The large dynamic model obtained through FEM is reduced using proper orthogonal decomposition (POD) technique, applicable to nonlinear systems, while preserving the input–output characteristics almost exactly. In the present work, using the reduced model, an optimal control input current for the induction heating system is obtained, for transferring the temperature profile over an axi-symmetric work piece from any initial temperature profile to any other desired temperature profile. The reduced model enables application and numerical implementation of established optimal control techniques based on Hamilton–Jacobi–Bellman theory for the system. Further, although the controller is developed using the reduced model, it is shown to perform perfectly well when used with the original unreduced model as well. Results of the numerical implementation are presented to evaluate the overall proposed approach.     

基于二元多项式拟合法的压力传感器输入信号重构

为了同时解决压力传感器输出特性的非线性和温度漂移问题,提出利用二元多项式根据传感器输出信号及其内部温度来重构被测压力.给出了被测压力重构模型的具体形式,以传感器静态特性标定试验数据为依据,利用最小二乘法求解模型系数矩阵,利用单点最大拟合误差、误差平方和模型阶次三项指标进行模型择优.基于重构模型给出了测量电路检测精度估算方法.本方法在传感器静态特性标定数据范围内得到了便于工程计算的一致表达式,同时解决了传感器的非线性和温度漂移两个主要问题,计算所用数据量大大小于插值法.某型硅谐振压力传感器的应用验证了本方法的合理性和有效性.