PID,Predictive and fuzzy temperature control for nuclear magnetic resonance spectroscopy experiments

Temperature control inside the probe of an NMR spectrometer is dealt with. First a simple model of the system is derived by means of step response experiments. Then four control algorithms are implemented and compared. Specifically, a PID regulator is first synthesized and applied to the real plant. Then, in order to cope with the non-linearity of the plant, two adaptive predictive control algorithms are implemented, namely the generalized predictive control and the constrained receding horizon predictive control algorithms. They both show superior performance with respect to PID for control of a simulated plant model and of the real plant. Finally a feasibility study is carried out for a fuzzy controller; its unsatisfactory behaviour in simulation prevents one from applying it to the plant.     

预测控制算法的研究现状及展望

本文简要回顾了预测控制的发展过程，对当前引起控制界广泛重视的几种预测控制算法进行了比较，着重对该领域的热点课题：鲁棒性问题、非线性系统控制、多变量系统控制、实用化问题和综合功能预测控制器的研究现状及发展趋势进行了讨论。     

基于神经网络的非线性预测控制

本文简单地介绍了基于神经网络预测控制的基本原理;并且根据神经网络所采用网络模型的不同对预测模型进行了分类以及按照求解最优控制作用方式的不同介绍了几种常用的优化算法。     

一种隐式广义预测自校正控制算法及其仿真研究

广义预测控制是一种新型的远程预测控制方法,它集多种算法的优点为一体,具有较强的鲁棒性。采用不辨识对象模型参数的隐式算法,大大减少了计算工作量,节省了时间。介绍了一种隐式广义预测自校正控制算法且对其进行仿真研究,分析了参数变化对整个系统性能的影响。通过仿真验证了该算法的优越性和可行性。     

单元机组负荷多变量模型预测控制

提出了一种具有输入约束的多变量模型预测控制算法，并将其应用于单元机组负荷控制进行仿真研究，结果表明，该多变量约束预测控制算法是有效的，并具有算法简单，在线计算量小的优点。     

基于CARMA模型的自校正预测PID调节器的应用

本文分析了用于离散控制系统常规ＰＩＤ调节器的两种典型算法，提出了基于ＣＡＲＭＡ模型自校正极点配置预测ＰＩＤ调节器新结构。并在电阻加热炉温度控制系统中获得了成功应用。取得了令人满意的控制效果。     

Typical predictive control algorithms

It is shown that the changeover from the typical algorithms commonly used to control the process parameters of power facilities to typical predictive algorithms makes it possible to achieve essentially better quality of control and at the same time to use the possibilities offered by microprocessor engineering to a fuller extent.     

非线性预测控制在电厂热工过程中的应用研究

非线性预测控制（NMPC）研究在电厂热工过程控制中有很大的应用潜力。目前NMPC算法的研究主要集中在线性化控制方法、基于特殊模型的NMPC方法、先进控制策略与预测控制相结合的控制方法。分析了NMPC方法在电厂各热工过程控制系统中的应用研究现状,并对其前景进行了展望。分析表明,系统辩识与模型预测控制（MPC）的协同作用、约束NMPC算法、非线性系统的建模与参数估计、多变量有约束系统的稳定性与鲁棒性定量分析、预测控制的智能化将是电厂热工过程预测控制问题未来的研究方向。     

The intravenous route to blood glucose control

This article discusses closed-loop blood glucose regulation algorithms that use the intravenous route for insulin delivery to insulin-dependent diabetic patients. Classical control methods and advanced algorithms using implicit knowledge or explicit models (empirical, fundamental, or “gray-box”) of the diabetic patient are examined. Current research on characterizing patient variability is presented, in the context of a model predictive controller able to adjust to changes in patient glucose and insulin sensitivity     

Partial state model reference adaptive control of a rapid thermal processor

The temperature control problem in a rapid thermal processor is addressed using first a conventional PID controller and then a partial state model reference adaptive controller derived via the long-range predictive control and linear quadratic control approaches. These two algorithms are shown to provide such fundamental design features as offset-free disturbance rejection and stability robustness with respect to unmodelled and time-varying dynamics.     

Linear and non-linear adaptive control of alcoholic fermentation process: Experimental results

This paper deals with the experimental implementation of several control algorithms on a continuous flow fermentation process. The regulation and tracking problems of the substrate concentration are considered. Our objective is to show the advantages and drawbacks of these algorithms in tracking and regulation behaviour, overshoot, knowledge contribution, number of tuning parameters, etc. The different controllers described in this paper are the PI controller, the non-linear adaptive L/A algorithm, the linear adaptive controller (predictive control with partial state reference model) and the adaptive controller based on the non-linear structure of the process model (pole placement control). These controllers are applied to a pilot-scale fermentation system with satisfactory results.     

用户侧微电网的能量管理方法综述

用户侧微电网是协调分布式发电与用户负荷的有效载体,而能量管理系统是提高其运行效用的重要途径。从用户侧微电网的能量管理需求与应用特点出发,系统论述了在线凸规划、在线模型预测控制、李雅普诺夫优化和在线线性规划等四类在线算法的基本原理。综述了其在微电网或类似系统能量管理中的应用情况,分析了不同在线算法的特点、性能、适用范围及评价指标。最后,简要总结了目前存在的难点问题与发展前景。     

容错型混合励磁磁通切换电机的模型预测控制

为了提高电机驱动系统的带故障运行性能,实现电机的最小铜耗容错运行,提出一种基于模型预测控制算法的容错型混合励磁磁通切换(FTHEFS)电机容错控制方法。以1台6/13极FTHEFS电机为控制对象,针对电机单相绕组断路故障,基于三相四桥臂逆变器拓扑,分别对模型预测转矩控制和无差拍模型预测磁链控制算法下的最小铜耗容错控制方法进行研究分析,并对所提容错控制方法的可行性和有效性进行验证。研究结果表明:两种控制方法均能使故障前后转矩、转速和定子磁链保持不变,同时降低故障后的电机铜耗,保证系统稳定运行。与模型预测转矩控制相比,无差拍模型预测磁链容错控制能够在降低开关频率的同时减小故障前后的磁链脉动。     

APPLICATION OF ADAPTIVE PREDICTIVE CONTROL TO A SHAKER CONVEYOR FOR TRANSPORTATION OF PLASTIC BEADS

After having briefly portrayed the basic principles of predictive control, three adaptive predictive control algorithms–the GPC algorithm and two new modifications of it proposed by the authors, called the MGPC (modified GPC) algorithm and the DGPC (direct GPC) algorithm—will be discussed. The main disadvantage of the GPC algorithm, namely the noise sensitivity, will be highlighted, when the real-time application to a plant with variable dead-time, a shaker conveyor system for transportation of plastic beads, will be studied. Both new algorithms proposed here show good behaviour in reducing the computational burden and achieving a better disturbance rejection than the direct application of the GPC algorithm. © 1997 by John Wiley & Sons, Ltd.     

模糊控制在三相PWM整流器无差拍控制中的应用

无差拍电流控制为了补偿控制延时,需要对参考电流进行超前预测,由于参考变化的不确定性,单一的预测算法在不同参考变化情况下可能产生较大的电流超调或者相位滞后。本文通过对三种常用预测算法进行比较分析,得出预测参数与电流超调和相位滞后之间的关系,以此提出一种基于模糊控制的预测算法。该算法针对不同的参考电流变化选择合适的预测参数,折衷考虑电流超调和相位滞后的影响,限制电流超调最大值,优化相位滞后。最后,结合TMS320F2812 数字信号处理器(digital signal processor,DSP)控制平台在三相电压型脉宽调制(pulse width modulation,PWM)整流器上进行实验,验证了理论分析的正确性。     

改进差拍控制算法在有源滤波器中的应用

为克服有源滤波器的无差拍控制法因算法复杂导致计算延时而影响补偿效果的缺点,优化了有源滤波器的预测控制目标函数,提出了一种改进差拍控制算法并仿真验证了其可行性。用该算法控制三相有源滤波器时,每个控制周期内只需12次加法和5次乘法运算,计算复杂度明显降低。仿真结果表明,改进算法的控制精度与传统算法十分接近;与定时比较控制法相比,改进算法可减少部分硬件环节,提高了数字化程度。将主要结论与谐波检测预报相结合可实现对无差拍控制法的简化。     

Fast parallel LQ regulator design for adaptive control

A new fast algorithm for linear quadratic (LQ) control optimization is derived using a partial state formulation and is exploited for the synthesis of long-range predictive controllers. The new algorithm requires O(Nn) computations for an nth-order plant and N-step prediction horizon and is about twice as cheap as existing fast algorithms. Systolic implementation on arrays of O(n) processors is also considered to get an O(N) processing time. Both fast algorithms and systolic implementation can be used to considerably speed up the control design task in adaptive predictive control schemes and thus increase the adaptation bandwidth. © 1997 John Wiley & Sons, Ltd.     

Comparison between various control methods for thermal power plant main control system

In thermal power plants, it is important to improve the control accuracy of main steam pressure and temperature and so forth during load up/down. This paper focuses on temperature control, the most difficult aspect of control due to the nonlinearity and long dead time of power plants. We applied control methods such as MRAC, neural network, and long‐range predictive control to the power plant main control system. Each method was evaluated by a simulator using detailed physical models that represent accurately the power plant dynamics. We confirmed that each method can provide proper control, but long‐range predictive control is better than the two other methods. In addition, thermal power plants are so complex that further analysis (e.g., persistently exciting condition, learning method of neural networks) is necessary for the application of theoretical algorithms. © 1999 Scripta Technica, Electr Eng Jpn, 128(3): 72–81, 1999     

Algorithms of adaptive tracking of unknown multisinusoidal signals in linear systems with arbitrary input delay

The problem of adaptive tracking control is addressed for the class of linear time‐invariant plants with known parameters and arbitrary known input delay. The reference signal is a priori unknown and is represented by a sum of biased harmonics with unknown amplitudes, frequencies, and phases. Asymptotic tracking is provided by predictive adjustable control with parameters generated by one of three designed adaptation algorithms. The first algorithm is based on a gradient scheme and ensures zero steady‐state tracking error with all signals bounded. The other two algorithms additionally involve the scheme with fast parametric convergence improving the closed‐loop system performance. In all the algorithms, the problem of delay compensation is resolved by special augmentation of tracking error. The adjustable control law proposed do not require identification of the reference signal parameters.