动态矩阵控制的模型简化与预报误差校正

依据热工过程的特点,提出对其实施动态矩阵控制(DMC),用简化的模型和预报误差校正改进常规的控制算法.文中介绍了预测模型简化的原理,推导出简化后预测控制的即时控制律.为了补偿简化后带来的模型失配,用预报模型误差代替模型误差对系统的滚动优化控制进行校正.预报误差校正是基于时间序列分析、参数估计和最优预报原理形成的.文章最后,将上述算法在供热对象上进行了仿真,分析了仿真的结果,说明改进的DMC算法在保证实时性、改善跟踪和鲁棒性方面均有益处.     

基于RBF神经网络的时间序列交叉供热负荷预报研究

针对供热过程的特点及节能控制的需要,提出基于RBF神经网络的时间序列交叉供热负荷预报法.首先对现场实测的供热负荷数据进行预处理,取得建立预报模型所需的负荷样本阵列;随后,应用自相关法求取RBF神经网络的输入维数,并分别建立时间序列的横向及纵向预报模型;最后用最小二乘法求出横向与纵向负荷预报的交叉权系数,得到RBF神经网络的时间序列交叉预报模型.仿真结果表明,RBF神经网络交叉负荷预报的精度高于横向负荷预报及纵向负荷预报,其实时性要优于BP神经网络交叉负荷预报.     

局域支持向量回归与误差区间估计的概率预报方法及其应用研究

 针对工程应用对预报技术要求的不断提高以及现有预报方法存在的问题,本文提出了一种局域支持向量回归(Local Support Vector Regression,LSVR)和误差区间估计相结合的概率预报方法,该方法利用局域支持向量回归降低噪声对点预报的干扰,提高预报的可靠性和准确性,利用非参数核估计获取误差区间,避免误差概率分布特性假设,再将点预报和误差区间结合获得预报置信区间,并进一步给出了联合预报置信区间.最后,给出了电网负荷预报和供热负荷预报算例,验证了所提出方法的有效性和实用性.     

基于PM-RBF的电厂供热负荷预测研究

针对供热系统热负荷预测模型的非线性,给出了并行最小化RBF神经网络算法,同时修正了RPCL方法系统振荡问题。通过仿真得出,当训练样本合适时,综合误差在6%以内。具有良好的预测精度和速度。     

协同综合预报IGDSS通用模型的研究与应用

分析了自然灾害综合预报会商式决策系统的特点,研究并提出了一种将智能决策方法作为“特殊预报专家”的协同综合预报IGDSS的通用模型。以地震然灾害综合预报为应用背景,根据地震综合预报的工作原理应用协同综合预报IGDSS通用模型,重点介绍了地震综合预报IGDSS的智能决策方法—贝叶斯网络的设计与实现,建立地震综合预报贝叶斯网络分析器,采用改进后的桶消元算法作为网络推理。给出了应用协同综合预报IGDSS通用模型解决地震综合预报的较完整的解决方案,系统已应用于福建省地震局预报中心。     

基于汽车领域的神经网络动态矩阵的误差补偿控制研究

基于预测控制算法的动态矩阵控制理论,改进得到了算法模型的误差相关矩阵,给出约束多变量DMC模型以及神经网络误差补偿的动态矩阵控制验证,在误差控制仿真验证中,应用神经网络误差补偿的预测控制效果优势明显,这一研究对模糊预测技术的进一步推广应用有一定的促进作用。     

优化粒子群算法在组合供热负荷预测中的应用

分析了粒子群算法和组合预测的特点。将组合预测和粒子群算法结合,建立了一种组合形式的供热负荷预测模型。同时针对粒子群算法易于陷入局部最优解、进化后期收敛慢等缺点对粒子群算法进行改进,解决了组合预测中权重难以确定的问题。改善了预测模型的拟合能力,提高了预测精确度。最后选取大庆油田某一天供热数据作为测试数据,结果表明组合预测误差较小,精确度高于其他单项预测方法40%以上。     

供热系统中供热不平衡的调节控制策略

近些年来,集中供热系统在我国北方城市得到了飞跃式的发展,大幅提升了城市人民的生活水平。城市集中供热对有效利用资源,提高热力资源的使用效率具有重大意义。与其他城市基础设施相比,供热系统研究和应用起步较晚,技术水平相对较低。在供热系统在供热过程中,会存在着供热不平衡的现象,给人们的生活带来了极大不便。本文介绍了供热系统的基本工作原理,分析造成供热系统供热不平衡的原因,并在最后给出调节控制供热系统的平衡策略。为城市供热系统工程提供借鉴与参考。     

基于深度信念网络的短期电力负荷预测

对电力负荷预测的原理、步骤及方法做了简要分析,对深度信念网络做了细致描述,在此基础上,提出了用深度信念网络的方法预测短期电力负荷,并做了相应的实验,深度信念网络的预测值十分逼近实际值,预测误差的绝对值范围小,为0~0.08,且误差范围波动较小,预测稳定。表明基于深度信念网络的短期电力负荷预测模型预测精准,具有很高的预测精度和预测稳定性。     

基于预报误差法的神经网络辩识

针对传统系统辨识存在的缺点,提出了基于预报误差法的神经网络辨识方法,将神经网络的预报误差法应用于系统辨识中,通过调节神经网络连接权值可使网络输出逼近系统输出。神经网络作为实际系统的辨识模型,可以用于在线控制。仿真实例表明其收敛速度快于BP算法。     

变流量节能系统动态监测及实时调节系统设计分析

为提高变流量节能系统自适应性能和实时调节能力,提出基于双启发式动态规划的变流量节能系统动态监测及实时调节系统设计方法,构建变流量节能动态监测及实时调节控制对象模型,以供水流量及一次侧的供水温度、供水流量为初始约束参量设计系统的控制规则。在此基础上,结合双启发式动态规划模式并采用模糊优化分配控制策略,对系统进行调节和控制,通过线性规划学习方法,实现变流量节能动态监测及实时调节。仿真结果表明,采用该方法进行变流量节能系统动态监测及实时调节的自适应性能较好,各时段流量预测误差曲线波动介于-1.5至1.5之间,且热负荷值与预测热负荷之间相对误差最大值为2.5%,误差较低。     

A describing function for resonantly commutated H-bridge inverters

The paper presents the derivation of a describing-function to model the dynamic behavior of a metal oxide semiconductor field effect transistor-based, capacitively commutated H-bridge, including a comprehensive explanation of the various stages in the switching cycle. Expressions to model the resulting input current, are also given. The derived model allows the inverter to be accurately modeled within a control system simulation over a number of utility input voltage cycles, without resorting to computationally intensive switching-cycle level, time-domain SPICE simulations. Experimental measurements from a prototype H-bridge inverter employed in an induction heating application, are used to demonstrate a high degree of prediction accuracy over a large variation of load conditions is possible using the simplified model.     

热网全网平衡系统设计方案

热网全网平衡系统主要通过获得换热站内的二次网供水温度、二次网回水温度、一次网电动阀门的反馈值和设定值等参数,以及热源的温度、压力、流量等参数,采用神经网络、数据回归等算法,控制一次网电动阀门的动作,从而减小热网的水平水力失调度,达到均匀供热和节约能源、增加经济效益的目的。全网平衡系统采用均匀性控制策略,调节换热站的一次网电动阀门或分布式变频泵,使每个热力站的供热效果基本一致、冷热均匀,同时消除了一次网的时延产生的参数不同步所造成的二次网参数不同步的失调现象,有效地保证了热网的同步负荷调节,规避了系统的容量、惯性、时延特性,是合理、有效的系统调节和全网平衡策略。     

虚拟CPU负载预测算法性能评估

准确的虚拟CPU负载预测是提高虚拟机CPU调度性能的重要前提,然而,虚拟机操作系统环境下,虚拟CPU负载预测方法需要尽可能简单、有效.针对虚拟机CPU调度应用场景,以实际CPU负载为研究对象,选取五种简单的时间序列预测算法,详细评估其虚拟CPU负载预测性能,为虚拟机CPU调度的实现提供了研究基础.结果表明,平均移动法、...     

Predictive Current Control Strategy With Imposed Load Current Spectrum

This paper presents a modified predictive current control strategy which allows one to have control over the spectrum of the load current. The proposed method uses a model of the system to predict the behavior of the current for each possible voltage vector generated by the inverter. For that purpose, at each sampling interval, signal predictions are evaluated using a cost function that quantifies the desired system behavior. The cost function used in this work evaluates the filtered error of the load currents. The inclusion of a filter for the load error allows one to manipulate current spectra. Thus, by designing this filter appropriately, the load spectrum can be shaped. The performance of the proposed control strategy is verified by simulation and experimental results.     

Thermoelectricity in Self-Contained Heating and Boiler Systems

The use of thermoelectric generators (TEGs) in heating systems enables autonomous supply of power to automatic safety devices, creation of optimized gas mixtures, and automation and precise temperature control of exhaust gas and heat carriers. It is particularly important to make heating systems independent of the district electric grid. Results of research and development efforts on a TEG for supplying power to electric devices of self-contained heating and boiler systems are presented. A TEG physical model is proposed, and results of computer simulation and optimization of its basic power and design parameters are given. Two TEG design variants (single and double sided) are considered. Their advantages and shortcomings are discussed. On the basis of theoretical calculations, a prototype TEG for a 10.5-kW boiler is built. At water heating system temperatures from 35°C to 80°C, the TEG electric power is 50 W to 65 W, which is used to supply a circulation pump for forced liquid heat carrier delivery (30 W to 40 W) and a fan for removal of fuel combustion products from the boiler’s smoke chamber (5 W to 7 W).     

Optimal predictive control with constraints for the processing ofsemiconductor wafers on bake plates

An optimal control scheme is designed to improve repeatability by minimizing the loading effects induced by the common processing condition of placement of a semiconductor wafer at ambient temperature on a large thermal-mass bake plate at processing temperature. The optimal control strategy is a model-based method using linear programming to minimize the worst-case deviation from a nominal temperature set point during the load disturbance condition. This results in a predictive controller that performs a predetermined heating sequence prior to the arrival of the wafer as part of the resulting feedforward/feedback strategy to eliminate the load disturbance. This procedure is based on an empirical model generated from data obtained during closed-loop operation. It is easy to design and implement for conventional thermal processing equipment. Experimental results are performed for a commercial conventional bake plate and depict an order-of-magnitude improvement in the settling time and the integral-square temperature error between the optimal predictive controller and a feedback controller for a typical load disturbance     

Discrete-time current control of voltage-fed three-phase PWMinverters

The discrete-time current control of three-phase voltage-fed pulsewidth modulation (PWM) inverters is discussed, with emphasis on important practical aspects. The inverter feeds a balanced three-phase load of R-L impedances in series with back-EMFs. The inverter-load system is modeled by using space vectors. The practical difficulties of the predictive control method are pointed out. In particular, effects of the delayed application of the manipulated voltage caused by controller computing time are investigated. A modified control law that eliminates the undesirable effects of the computing delay is introduced. Also, prediction of current reference is utilized, which would enable operation with any reference function. An estimation and prediction procedure for the back EMF is also introduced. Steady-state errors in the load current due to these predictions are calculated for sinusoidal operation. Simulation results for the predictive control and the modified control laws are presented and discussed     

水质浊度红外光检测及聚类灰色融合预测模型

为了对水处理过程中水质浊度进行实时、准确检测,设计了基于红外光的散射浊度检测系统,并提出一种聚类灰色融合的预测模型对水质浊度的变化趋势进行有效预测。利用890 nm波长的红外发光二极管作为发光器件,光敏二极管作为接收器,检测装置响应时间短,零点误差小。采用灰色预测算法和聚类融合的方法对传感器所采集的数据进行处理,将聚类融合处理后的数据作为灰色预测控制的输入数据,灰色预测控制的输出数据与融合数据进行对比分析,确定预测浊度值。通过实际项目进行了数据跟踪和运算,聚类灰色融合算法的浊度预测输出值和实测值的平均误差值为0.008 7 NTU,聚类灰色融合算法预测性能优于单一的灰色预测算法,能够保证水质浊度参数的平稳,满足了水质的要求。