Tuning PID Parameters Based on a Combination of the Expert System and the Improved Genetic Algorithms

a new strategy combining an expert system and improved genetic algorithms is presented for tuning proportional-integral-derivative （PID） parameters for petrochemical processes. This retains the advantages of genetic algorithms, namely rapid convergence and attainment of the global optimum. Utilization of an orthogonal experiment method solves the determination of the genetic factors. Combination with an expert system can make best use of the actual experience of the plant operators. Simulation results of typical process systems examples show a good control performance and robustness.

Application of Time Scale to Parameters Tuning of Active Disturbance Rejection Controller for Induction Motor

Active disturbance rejection controller (ADRC) has good performance in induction motor (IM) control system, but controller parameter is difficult to tune. A method of tuning ADRC parameter by time scale is analyzed. The IM time scale is obtained by theoretical analysis. Combining the relations between scale time and ADRC parameters, ADRC parameter tuning in IM vector control based stator flux oriented is obtained. This parameter tuning method is validated by simulations and it provides a new technique for tuning of ADRC parameters of IM.

Automated Arabic Text Classification Using Hyperparameter Tuned Hybrid Deep Learning Model

The text classification process has been extensively investigated in various languages,especially English.Text classification models are vital in several Natural Language Processing(NLP)applications.The Arabic language has a lot of significance.For instance,it is the fourth mostly-used language on the internet and the sixth official language of theUnitedNations.However,there are few studies on the text classification process in Arabic.A few text classification studies have been published earlier in the Arabic language.In general,researchers face two challenges in the Arabic text classification process:low accuracy and high dimensionality of the features.In this study,an Automated Arabic Text Classification using Hyperparameter Tuned Hybrid Deep Learning(AATC-HTHDL)model is proposed.The major goal of the proposed AATC-HTHDL method is to identify different class labels for the Arabic text.The first step in the proposed model is to pre-process the input data to transform it into a useful format.The Term Frequency-Inverse Document Frequency(TF-IDF)model is applied to extract the feature vectors.Next,the Convolutional Neural Network with Recurrent Neural Network(CRNN)model is utilized to classify the Arabic text.In the final stage,the Crow Search Algorithm(CSA)is applied to fine-tune the CRNN model’s hyperparameters,showing the work’s novelty.The proposed AATCHTHDL model was experimentally validated under different parameters and the outcomes established the supremacy of the proposed AATC-HTHDL model over other approaches.

A Two-stage Tuning Method of Servo Parameters for Feed Drives in Machine Tools

Based on the evaluation of dynamic performance for feed drives in machine tools, this paper presents a two-stage tuning method of servo parameters. In the first stage, the evaluation of dynamic performance, parameter tuning and optimization on a mechatronic integrated system simulation platform of feed drives are performed. As a result, a servo parameter combination is acquired. In the second stage, the servo parameter combination from the first stage is set and tuned further in a real machine tool whose dynamic performance is measured and evaluated using the cross grid encoder developed by Heidenhain GmbH. A case study shows that this method simplifies the test process effectively and results in a good dynamic performance in a real machine tool.

A method of tuning PID parameters for P-GMAW based on physical experiments

To improve welding quality, a method of proportional-integral-differential （PlD） parameters tuning based on pulsed gas metal arc welding （P-GMAW） control was put forward. Aiming at the request of dynamic responsiveness of PGMA W constant current control, a self-developed welding waveform wavelet analyzer was employed. By tuning the proportional parameter, integration time and differential time in sequence, the optimal PID parameters could be achieved. The results showed that, due to the PID parameters tuned by this method, the welding process was stable and the weld bead appearance was nice. The requirement of dynamic responsiveness of P-GMAW constant current control was fully met.

Fault Diagnosis Based on Fuzzy Support Vector Machine with Parameter Tuning and Feature Selection

This study describes a classification methodology based on support vector machines(SVMs),which offer superior classification performance for fault diagnosis in chemical process engineering.The method incorporates an efficient parameter tuning procedure(based on minimization of radius/margin bound for SVM's leave-one-out errors)into a multi-class classification strategy using a fuzzy decision factor,which is named fuzzy support vector machine(FSVM).The datasets generated from the Tennessee Eastman process(TEP)simulator were used to evaluate the clas-sification performance.To decrease the negative influence of the auto-correlated and irrelevant variables,a key vari-able identification procedure using recursive feature elimination,based on the SVM is implemented,with time lags incorporated,before every classifier is trained,and the number of relatively important variables to every classifier is basically determined by 10-fold cross-validation.Performance comparisons are implemented among several kinds of multi-class decision machines,by which the effectiveness of the proposed approach is proved.

Controller Parameter Tuning of Delta Robot Based on Servo Identification

High-speed pick-and-place parallel robot is a system where the inertia imposed on the motor shafts is real-time changing with the system configurations.High quality of computer control with proper controller parameters is conducive to overcoming this problem and has a significant effect on reducing the robot＇s tracking error.By taking Delta robot as an example,a method for parameter tuning of the fixed gain motion controller is presented.Having identifying the parameters of the servo system in the frequency domain by the sinusoidal excitation,the PD＋feedforward control strategy is proposed to adapt to the varying inertia loads,allowing the controller parameters to be tuned by minimizing the mean square tracking error along a typical trajectory.A set of optimum parameters is obtained through computer simulations and the effectiveness of the proposed approach is validated by experiments on a real prototype machine.Let the traveling plate undergoes a specific trajectory and the results show that the tracking error can be reduced by at least 50%in comparison with the conventional auto-tuning and Z-N methods.The proposed approach is a whole workspace optimization and can be applied to the parameter tuning of fixed gain motion controllers.

Parameter Tuning Method for Dither Compensation of a Pneumatic Proportional Valve with Friction

In the practical application of pneumatic control devices, the nonlinearity of a pneumatic control valve become the main factor affecting the control effect, which comes mainly from the dynamic friction force. The dynamic friction inside the valve may cause hysteresis and a dead zone. In this paper, a dither compensation mechanism is proposed to reduce negative effects on the basis of analyzing the mechanism of friction force. The specific dither signal（using a sinusoidal signal） was superimposed on the control signal of the valve. Based on the relationship between the parameters of the dither signal and the inherent characteristics of the proportional servo valve, a parameter tuning method was proposed, which uses a displacement sensor to measure the maximum static friction inside the valve. According to the experimental results, the proper amplitude ranges are determined for different pressures. In order to get the optimal parameters of the dither signal, some dither compensation experiments have been carried out on different signal amplitude and gas pressure conditions. Optimal parameters are determined under two kinds of pressure conditions. Using tuning parameters the valve spool displacement experiment has been taken. From the experiment results, hysteresis of the proportional servo valve is significantly reduced. And through simulation and experiments, the cut-off frequency of the proportional valve has also been widened. Therefore after adding the dither signal, the static and dynamic characteristics of the proportional valve are both improved to a certain degree. This research proposes a parameter tuning method of dither signal, and the validity of the method is verified experimentally.

Control parameter optimal tuning method based on annealing-genetic algorithm for complex electromechanical system

A new searching algorithm named the annealing-genetic algorithm(AGA) was proposed by skillfully merging GA with SAA. It draws on merits of both GA and SAA ,and offsets their shortcomings.The difference from GA is that AGA takes objective function as adaptability function directly,so it cuts down some unnecessary time expense because of float-point calculation of function conversion.The difference from SAA is that AGA need not execute a very long Markov chain iteration at each point of temperature, so it speeds up the convergence of solution and makes no assumption on the search space,so it is simple and easy to be implemented.It can be applied to a wide class of problems.The optimizing principle and the implementing steps of AGA were expounded. The example of the parameter optimization of a typical complex electromechanical system named temper mill shows that AGA is effective and superior to the conventional GA and SAA.The control system of temper mill optimized by AGA has the optimal performance in the adjustable ranges of its parameters.

A Statistical Parameter Analysis and SVM Based Fault Diagnosis Strategy for Dynamically Tuned Gyroscopes

Gyro's fault diagnosis plays a critical role in inertia navigation systems for higher reliability and precision. A new fault diagnosis strategy based on the statistical parameter analysis (SPA) and support vector machine (SVM) classification model was proposed for dynamically tuned gyroscopes (DTG). The SPA, a kind of time domain analysis approach, was introduced to compute a set of statistical parameters of vibration signal as the state features of DTG, with which the SVM model, a novel learning machine based on statistical learning theory (SLT), was applied and constructed to train and identify the working state of DTG. The experimental results verify that the proposed diagnostic strategy can simply and effectively extract the state features of DTG, and it outperforms the radial-basis function (RBF) neural network based diagnostic method and can more reliably and accurately diagnose the working state of DTG.

Proportion integral-type active disturbance rejection generalized predictive control for distillation process based on grey wolf optimization parameter tuning

The high-purity distillation column system is strongly nonlinear and coupled,which makes it difficult to control.Active disturbance rejection control(ADRC)has been widely used in distillation systems,but it has limitations in controlling distillation systems with large time delays since ADRC employs ESO and feedback control law to estimate the total disturbance of the system without considering the large time delays.This paper designs a proportion integral-type active disturbance rejection generalized predictive control(PI-ADRGPC)algorithm to control the distillation column system with large time delay.It replaces the PD controller in ADRC with a proportion integral-type generalized predictive control(PI-GPC),thereby improving the performance of control systems with large time delays.Since the proposed controller has many parameters and is difficult to tune,this paper proposes to use the grey wolf optimization(GWO)to tune these parameters,whose structure can also be used by other intelligent optimization algorithms.The performance of GWO tuned PI-ADRGPC is compared with the control performance of GWO tuned ADRC method,multi-verse optimizer(MVO)tuned PI-ADRGPC and MVO tuned ADRC.The simulation results show that the proposed strategy can track reference well and has a good disturbance rejection performance.

Adaptive Parallel Particle Swarm Optimization Algorithm Based on Dynamic Exchange of Control Parameters

Updating the velocity in particle swarm optimization (PSO) consists of three terms: the inertia term, the cognitive term and the social term. The balance of these terms determines the balance of the global and local search abilities, and therefore the performance of PSO. In this work, an adaptive parallel PSO algorithm, which is based on the dynamic exchange of control parameters between adjacent swarms, has been developed. The proposed PSO algorithm enables us to adaptively optimize inertia factors, learning factors and swarm activity. By performing simulations of a search for the global minimum of a benchmark multimodal function, we have found that the proposed PSO successfully provides appropriate control parameter values, and thus good global optimization performance.

The Analysis of Peculiar Control Parameters of Artificial Bee Colony Algorithm on the Numerical Optimization Problems

Artificial bee colony (ABC) algorithm is one of the popular swarm intelligence algorithms. ABC has been developed by being inspired foraging and waggle dance behaviors of real bee colonies in 2005. Since its invention in 2005, many ABC models have been proposed in order to solve different optimization problems. In all the models proposed, there are only one scout bee and a constant limit value used as control parameters for the bee population. In this study, the performance of ABC algorithm on the numeric optimization problems was analyzed by using different number of scout bees and limit values. Experimental results show that the results obtained by using more than one scout bee and different limit values, are better than the results of basic ABC. Therefore, the control parameters of the basic ABC should be tuned according to given class of optimization problems. In this paper, we propose reasonable value ranges of control parameters for the basic ABC in order to obtain better results on the numeric optimization problems.

基于PI参数的二阶线性自抗扰控制参数整定

线性自抗扰控制是一种不依赖于被控过程模型的控制方法,具有很好的工程应用潜力。目前,工业控制过程中大多采用PI控制器,为了确保由调试好的PI控制器平稳切换到线性自抗扰控制器,并使系统仍保持稳定状态,需要合理设置线性自抗扰控制器的参数。为此,在分析二阶线性自抗扰控制器的二自由度等效结构的基础上,推导出其反馈控制器与PID控制器的对应关系,给出一种基于现有PI控制参数直接获取二阶线性自抗扰控制初始参数的方法。最后,在MATLAB/Simulink平台上采用若干典型传递函数和双容水箱液位控制系统进行仿真研究,仿真结果验证了所提方法的有效性。

基于Q学习的高超声速飞行器自抗扰控制研究

为实现高超声速飞行器姿态自抗扰控制的参数整定,提出一种模糊Q学习算法。首先,采用强化学习中的Q学习算法来实现姿态自抗扰控制参数的离线闭环快速自适应整定;然后,根据模糊控制的思路,将控制参数划分为不同区域,通过设定奖励,不断更新Q表;最后,将训练好的Q表用于飞行器的控制。仿真结果表明,相对于传统的线性自抗扰控制(linear active disturbance rejection control,LADRC)和滑模控制,基于Q学习的LADRC省去了人工调试参数的繁琐过程,且仍具有良好的跟踪效果。蒙特卡罗仿真测试结果验证了基于Q学习的LADRC的鲁棒性。

海上风电柔直送出系统改进频率协同控制策略

针对远海风场经柔直系统接入岸上电网后无法提供频率主动支撑的问题,该文提出一种远海风场与柔直系统改进频率协同控制方案。协同控制方案分为柔直系统层面和远海风场层面。针对柔直系统,将岸上电网的频率变化通过下垂控制转化为直流电压的变化,分析下垂控制系数与柔直系统虚拟惯量系数之间的关系,提出一种计及电网频率偏差的柔直系统自适应控制方法。针对远海风场,在频率故障前以最大化风电场转子动能为目标提出一种改进的功率备用方法,能有效减小风电场的风能损失并将该部分能量用于后续频率调节中,在频率故障后考虑风速差异根据转子转速实时调整风电机组频率控制系数,能在保证机组运行安全性的前提下提升风电场整体调频能力。仿真结果表明,所提控制方法在不同风速以及风电渗透率下均具有良好的调频性能。

基于m×2正则化交叉验证的神经网络超参数调优方法

超参数调优是神经网络建模的关键问题。针对传统的超参数调优方法存在的问题,该文提出了一种基于m×2正则化交叉验证的超参数调优方法。目的是给出一种适用于复杂模型、大数据集背景下的计算开销较小且稳健的超参数调优方法。该方法的思想是从完整的数据集上选取少部分数据进行调优,避免模型在数据集较大时非常耗时的超参数调优难题;在m×2交叉验证的基础上设置正则化条件均衡训练集与验证集之间的分布差异,从而减少分布不一致带来的性能波动;使用信噪比作为调优的优化目标,从而可以综合考虑模型性能评价指标的均值和方差;并采用正交设计选择相关性较低的超参数组合以提高调优效率。以命名实体任务为例进行实验,在CoNLL 2003数据集上的实验结果显示,提出的调优方法能够选到和网格搜索性能上没有显著差异的超参数组合,且调优时间可显著降低约66%。

基于合作关系的多智能体数据库参数调优模型

高维数据库参数空间中的参数调优是提高数据库性能的难点,现有方法更多关注于如何识别重要参数,在如何有效提高可调参数数量的问题上仍存在不足。针对上述问题,基于低维度映射技术和多智能体(Multi-Agent)强化学习技术,提出基于合作关系的Multi-Agent数据库参数调优(Cooperative Multi-Agent Database Parameter Tuning,CMADPT)模型,CMADPT将数据库参数进行分类调优,极大增加了可调参数的数量;提出基于低维度映射的降维模型(Low Dimensional Mapping Model,LDMM),通过低维的合成参数调优高维的数据库参数。实验结果表明,CMADPT模型有效地扩大了可调参数的数量,比主流方法平均提升1.117%的数据库性能。此外,CMADPT每300次迭代训练平均节省1.32 h,极大地提升了算法的时间性能。

基于非线性动态重心粒子群优化的分数阶PI^(λ)D^(μ)控制器设计

针对现有Oustaloup滤波器拟合精度不佳、结构复杂的缺点,提出了最优精简Oustaloup滤波器。针对粒子群优化算法整定分数阶PI^(λ)D^(μ)控制器参数时学习能力不充分、迭代收敛乏力的问题,提出了一种改进的粒子群优化算法。该算法设计了双异步非线性动态学习因子,以提高粒子的思考能力与信息共享能力,并增加了粒子群质量重心项,用以加速收敛过程。将改进的算法结合最优精简Oustaloup滤波器应用于分数阶PI^(λ)D^(μ)控制器的设计过程,选取了2个分数阶系统模型进行仿真验证。结果表明,改进的算法收敛速度更快且不易陷入局部最优,所设计的控制系统超调量更小、调节时间更短、稳态误差更小,提高了系统的抗干扰能力。

基于LMI的二自由度飞行器多变量控制

针对二自由度飞行器存在非线性特性、强耦合性以及易受外界干扰等问题,提出了基于LMI的二自由度飞行器多变量控制方法。首先,根据二自由度飞行器系统的输入和输出变量,结合空间动力学方程建立了系统的状态空间模型;其次,设计满足静态输出反馈稳定的多变量控制律;然后,通过李亚普托夫稳定性定理与极点配置法,将控制系统稳定性问题转化为优化问题,利用线性矩阵不等式(LMI)求解满足最优性能指标的多变量控制器参数;最后,通过对比仿真分析与单变量PID,表明该方法能有效提升二自由度飞行器的轨迹跟踪精度。