An improved krill herd algorithm with global exploration capability for solving numerical function optimization problems and its application to data clustering

Krill herd algorithm is a stochastic nature-inspired algorithm for solving optimization problems. The performance of krill herd algorithm is degraded by poor exploitation capability. In this study, we propose an improved krill herd algorithm (IKH) by making the krill the global search capability. The enhancement comprises of adding global search operator for exploration around the defined search region and thus the krill individuals move towards the global best solution. The elitism strategy is also applied to maintain the best krill during the krill update steps. The proposed method is tested on a set of twenty six well-known benchmark functions and is compared with thirteen popular optimization algorithms, including original KH algorithm. The experimental results show that the proposed method produced very accurate results than KH and other compared algorithms and is more robust. In addition, the proposed method has high convergence rate. The high performance of the proposed algorithm is then employed for data clustering problems and is tested using six real datasets available from UCI machine learning laboratory. The experimental results thus show that the proposed algorithm is well suited for solving even data clustering problems.     

自适应磷虾群优化Elman神经网络的目标威胁评估

提出一种自适应磷虾群算法,在基本磷虾群算法中引入遗传繁殖机制,并加入进化算子和优化算子构成自适应环节,提高了算法的全局搜索能力和预测精度;通过自适应磷虾群算法对Elman神经网络的初始权值和阈值进行寻优,并在此基础上建立目标威胁评估模型。仿真实验表明,自适应磷虾群优化Elman神经网络既保证了一定的收敛速度,又能够使寻优精度得到明显提升,其对测试集的预测结果优于传统Elman神经网络和基本磷虾群优化Elman神经网络,从而验证了算法模型在目标威胁评估中的可行性、有效性。     

遗传算法在跨超声速风洞总压控制中的应用

总压作为风洞控制中的重要流场参数,其调节性能是风洞控制系统能否满足试验要求的重要指标,为提高跨超声速风洞的总压控制水平,需对总压控制策略进行设计。针对某跨超声速风洞对总压控制系统提出的快速性和精确性要求,提出串级控制、智能PID控制和总压分段控制等方法,并利用MATLAB系统辨识工具箱对流场调节阶段的总压系统模型进行了辨识。提出将遗传算法应用于风洞流场调节阶段的PID控制器参数整定中,重点对基于遗传算法的PID控制原理和参数整定步骤进行介绍,并针对遗传算法的遗传算子进行了设计。系统仿真和风洞实际运行情况表明：该方法较常规PID参数整定与优化方法,具有更好的控制性能指标,满足总压控制系统精确性、快速性、鲁棒性等要求,为后续风洞建设和设备改造提供了新方法。     

基于改进磷虾群算法的SVDD参数优化

支持向量数据描述（SVDD）是构造单类数据描述的分类算法,惩罚参数[C]和核参数[σ]作为影响SVDD分类效果的关键,其合理选取一直是个难点。针对这一问题,提出了一种基于改进磷虾群算法的SVDD参数优化算法（IKH-SVDD）。依据仿真实验,分析参数[C]和[σ]对描述边界的影响;引入磷虾群算法并分析其优劣,通过在随机扩散行为中定义扰动因子,增强算法的全局搜索能力;将一种新的精英选择和保留策略引入迭代过程,提高算法的收敛精度;将改进的磷虾群算法引入SVDD参数优化过程,构建了IKH-SVDD参数优化模型。基于UCI标准数据库进行实验并与其他几种参数优化算法进行比较,结果表明了IKH-SVDD算法具有更高的分类准确性。     

Supersonic, variable-throat, blow-down wind tunnel control using genetic algorithms, neural networks, and gain scheduled PID

In this paper the design and application of a control algorithm is discussed to control the test conditions within plenum chamber and the test section of a supersonic blow-down, variable throat wind tunnel at the University of Alabama. The artificially intelligent controller algorithm was designed using a gain scheduled Proportional-Integral-Differential (PID) control approach. The PID controller was augmented to work with time variant properties of the control problem by determining a functional form of the integral term of the controller from the governing equations of the tunnel. The controller was optimized using genetic algorithms (GA) on a neural network (NN) model of the tunnel and was compared to a conventional PID controller using the same NN model. The process was repeated for different throat settings to find the control gains for each setting. The controller algorithm was next applied to the actual wind tunnel at different throat settings and the results were compared. The optimized controller is proven to work very well at every throat setting.     

奇异摄动系统的H∞控制: 基于奇异系统的方法

实际系统的奇异系统模型往往通过忽略系统的奇异摄动系统模型微分项系数矩阵中的小时间参数得到.然而, 传统的奇异系统控制器设计很少考虑微分项系数矩阵的摄动. 本文拓展了现有奇异系统输出动态反馈H_∞控制器的设计, 使得当实际闭环系统中存在小时间参数时, 仍然能保持稳定并且满足一定的H_∞性能指标. 仿真算例说明了本文提出方法的有效性.     

一类大时滞非线性网络控制系统的H∞保成本控制

本文研究一类大时滞不确定非线性网络控制系统的H∞保成本控制问题. 首先将具有随机大时滞的网络控制系统模型化为具有不确定系数的离散时间系统模型. 然后利用Lyapunov稳定性理论和线性矩阵不等式方法, 给出H∞保成本控制律的存在条件和H∞保成本控制器的设计方法. 进而通过求解凸优化问题得到最优鲁棒H∞保成本控制器. 最后仿真结果验证了该控制算法的有效性.     

Delay-dependent H-infinity control for continuous time-delay systems via state feedback

The delay-dependent H-infinity analysis and H-infinity control problems for continuous time-delay systems are studied. By introducing an equality with some free weighting matrices, an improved criterion of delay-dependent stability with H-infinity performance for such systems is presented, and a criterion of existence and some design methods of delay-dependent H-infinity controller for such systems are proposed in term of a set of matrix inequalities, which is solved efficiently by an iterative algorithm. Further, the corresponding results for the delay-dependent robust H-infinity analysis and robust H-infinity control problems for continuous time-delay uncertain systems are given. Finally, two numerical examples are given to illustrate the efficiency of the proposed method by comparing with the other existing results.     

一种基于改进KH与KHM聚类的混合数据聚类算法

为解决K-means聚类对初始聚类中心敏感和易陷入局部最优的问题,提出一种基于改进磷虾群算法与K-harmonic means的混合数据聚类算法.提出一种具有莱维飞行和交叉算子的磷虾群算法以改进磷虾群算法易陷入局部极值和搜索效率低的不足,即在每次标准磷虾群位置更新后加入新的位置更新方法进一步搜索以提高种群的搜索能力,同时交替使用莱维飞行与交叉算子对当前群体位置进行贪婪搜索以增强算法的全局搜索能力. 20个标准测试函数的实验结果表明,改进算法不易陷入局部最优解,可在较少的迭代次数下有效地搜索到全局最优解的同时保证算法的稳定性.将改进的磷虾群算法与K调和均值聚类融合,即在每次迭代后用最优个体或经过K调和均值迭代一次后的新个体替换最差个体. 5个UCI真实数据集的测试结果表明:融合后的聚类算法能够克服K-means对初始聚类中心敏感的不足且具有较强的全局收敛性.     

线性不确定系统的H∞状态反馈鲁棒重复控制

现有的重复控制设计不能同时优化低通滤波器的参数和重复控制器的参数.我们在设计重复控制系统以控制线性不确定对象时,解决了这个问题.首先,引入状态反馈以保证闭系统的鲁棒稳定性,把重复控制器设计问题转化为H∞状态反馈增益的设计问题.为获得低通滤波器最大转折频率,进一步将设计问题转化为基于线性矩阵不等式约束的凸优化问题.提出了一种迭代算法,用以计算低通滤波器的最大转折频率和H∞状态反馈增益.在保证系统鲁棒稳定性的同时,获得最高控制精度的重复控制器和低通滤波器的参数组合.该方法与已有方法比较,它的结果容易验证和求解,因而更适合于实际应用.最后,通过数值实例验证了本文所提方法的有效性.     

Fault-tolerant decentralized H-infinity control for multi-channel systems using homotopy method

This paper considers a fault-tolerant decentralized H-infinity control problem for multi-channel linear time-invariant systems. The purpose is to design a decentralized H-infinity output feedback controller to stabilize the given system and achieve a certain H-infinity performance requirement both in the normal situation and in the situation where any one of the local controllers fails. The designed problem is reduced to a feasibility problem of a set of bilinear matrix inequalities (BMIs). An algorithm is proposed to solve the BMIs. First, the normal situation is considered where all the local controllers are functioning. The local controllers are obtained from a standard centralized H-infinity controller by using a homotopy method imposing a structural constraint progressively. Secondly, the above case is extended to the one where any one of the local controllers fails. We again use a homotopy method where the coefficient matrices of the failed controller are decreased rogressively to zero. The efficiency of the proposed algorithm is demonstrated by an example.     

Fault-tolerant decentralized H-infinity control for multi-channel systems using homotopy method

This paper considers a fault-tolerant decentralized H-infinity control problem for multi-channel linear time-invariant systems. The purpose is to design a decentralized H-infinity output feedback controller to.stabilize the given system and achieve a certain H-infinity performance requirement both in the normal situation and in the situation where any one of the local controllers fails. The designed problem is reduced to a feasibility problem of a set of bilinear matrix inequalities （BMIs）. An algorithm is proposed to solve the BMIs. First, the normal situation is considered where all the local controllers are functioning. The local controllers are obtained from a standard centralized H-infinity controller by using a homotopy method imposing a structural constraint progressively. Secondly, the above case is extended to the one where any one of the local controllers fails. We again use a homotopy method where the coefficient matrices of the failed controller are decreased progressively to zero. The efficiency of the proposed algorithm is demonstrated by an example.     

Design of mixed H-two/H-infinity optimal control systems using multiobjective differential evolution algorithm

In this paper, the mixed H-two/H-infinity control synthesis problem is stated as a multiobjective opti-mization problem, with objectives of minimizing the H-two and H-infinity norms simultaneously. Instead of building a LMIs-based synthesis algorithm, a self-adaptive control parameter multiobjective differential evolution algorithm is developed directly in the controller parameters space. In the case of systems with polytopic uncertainties, the worst case norm computation is formulated as an implicit optimization problem, and the proposed self-adaptive differential evolution is employed to calculate the worst case H-two and H-infinity norms. The numerical examples illustrate the power and validity of the proposed approach for the mixed H-two/H-infinity control multiobjective optimal design.     

喷水推进式水面机器人H∞鲁棒航向跟踪控制

针对水面机器人(unmanned surface vehicle, USV)航向跟踪容易受到风、浪与水流干扰影响的问题,提出了一种基于线性变参(linear parameter varying, LPV)模型的H_∞鲁棒航向跟踪控制器.首先从水动力学机理出发,提出了基于速度变参的LPV模型.然后基于提出的速度变参LPV模型,利用线性矩阵不等式设计了USV的H_∞鲁棒航向控制器,用以抵抗风、浪与水流对机器人的影响.最后,在自主研发的3自由度欠驱动喷水推进式USV平台上进行了实验.实验结果表明,控制器可以实现鲁棒的航向跟踪控制.     

船舶航向非线性H∞逆优化输出反馈控制

为抑制船舶航向非线性优化控制中模型参数摄动和由状态观测器引入的不确定观测误差,提出了一种非线性H∞逆优化控制算法.首先,基于无源理论设计观测器以实现海浪滤波,该观测器无需海浪扰动的方差信息从而减少了观测器参数数量.然后,考虑模型参数摄动对观测误差的影响,给出了描述系统局部(全局)性态的局部(全局)H∞优化性能指标.在以广义黎卡提方程(GARE)对局部优化问题的求解的基础上,应用逆优化方法将全局H∞优化问题转化为构造闭环系统的Lyapunov函数问题,得到同时满足两种指标的优化控制器,并证明了稳定性.仿真结果证明了该算法的有效性.     

磁悬浮直线电动机H∞鲁棒控制器及其蚁群算法优化设计

磁悬浮永磁直线电动机在结构上取消了机械传动的中间环节,具有磁悬浮和直接驱动的特点.针对磁悬浮永磁直线电动机由于悬浮高度不同,在磁悬浮和直接驱动运行过程中存在自身参数摄动和外界干扰突出的问题,设计基于蚁群算法的H_∞鲁棒控制器,以保证系统对这些不确定性具有良好的鲁棒性.建立包含磁悬浮永磁直线电动机参数摄动和外界干扰的状态空间模型.推导出无须满足正则条件约束的Riccati不等式,给出H_∞鲁棒控制器的解析表达式.针对H_∞控制器中加权矩阵选择的困难,采用蚁群算法对加权矩阵进行寻优.最后在MATLAB环境下对控制系统进行仿真研究.仿真实验表明基于蚁群算法的磁悬浮永磁直线电动机控制系统的性能比优化前有较明显改善,说明该方法的可行性和有效性.     

导弹的LPV建模与混合H2/Hinf自动驾驶仪设计

导弹在飞行过程中,高度、速度、攻角、动压等变量变化剧烈,如何在其大范围变化情况下设计满足要求的自动驾驶仪是研究的重点所在.论文利用线性参数时变方法对导弹动态系统完成建模过程,并采用线性分式变换将控制器设计问题转换为设计名义主控制器与根据飞行状态实时调整的参数控制器的过程,同时以混合H2/Hinf为性能指标,利用线性矩阵不等式对上述问题完成求解.最后以导弹自动驾驶仪设计为例对文中算法进行仿真验证,经验证表明此算法可以满足跟踪指令输入并具有较好的鲁棒稳定性.     

基于动态压力控制算子的磷虾群算法

针对基础磷虾群（KH）算法在求解复杂函数优化问题时局部搜索能力差、求解精度低、收敛速度慢、容易陷入局部最优等问题,提出一种基于动态压力控制算子的磷虾群算法（DPCKH）。该算法将一种新的动态压力控制算子加入了标准磷虾群算法,使其处理复杂函数优化问题更有效。动态压力控制算子通过欧氏距离量化了多个不同优秀个体对目标个体的诱导效应,进而在优秀个体附近加速产生新磷虾个体,提高了磷虾个体的局部探索能力。通过比较蚁群算法（ACO）、差分进化算法（DE）、磷虾群算法（KH）、改进的磷虾群算法（KHLD）和粒子群算法（PSO）,DPCKH算法在7个测试函数上的结果表明,DPCKH算法与ACO算法、DE算法、KH算法、KHLD算法和PSO算法相比有着更强的局部勘测能力,其开采能力更强。     

分散混杂自适应模糊H∞控制器及其在自动化公路系统中的应用

本文针对一类SISO不确定非线性大系统,提出了一种混杂间接和直接自适应分散模糊H∞控制器.通过组合模糊系统和H∞跟踪技术开发的分散自适应模糊控制算法避免了控制设计中含有的符号函数.两种自适应模糊控制器的组合消除了它们各自均不能够同时融合被控对象知识与控制知识的局限.闭环大系统被证明是稳定的,且具有H∞跟踪性能.该算法应用于自动化公路系统中车辆的纵向跟随控制,仿真结果表明混杂自适应模糊H∞控制系统的跟踪性能更好而相应的控制幅值却更小.     

Hybrid clustering analysis using improved krill herd algorithm

In this paper, a novel text clustering method, improved krill herd algorithm with a hybrid function, called MMKHA, is proposed as an efficient clustering way to obtain promising and precise results in this domain. Krill herd is a new swarm-based optimization algorithm that imitates the behavior of a group of live krill. The potential of this algorithm is high because it performs better than other optimization methods; it balances the process of exploration and exploitation by complementing the strength of local nearby searching and global wide-range searching. Text clustering is the process of grouping significant amounts of text documents into coherent clusters in which documents in the same cluster are relevant. For the purpose of the experiments, six versions are thoroughly investigated to determine the best version for solving the text clustering. Eight benchmark text datasets are used for the evaluation process available at the Laboratory of Computational Intelligence (LABIC). Seven evaluation measures are utilized to validate the proposed algorithms, namely, ASDC, accuracy, precision, recall, F-measure, purity, and entropy. The proposed algorithms are compared with the other successful algorithms published in the literature. The results proved that the proposed improved krill herd algorithm with hybrid function achieved almost all the best results for all datasets in comparison with the other comparative algorithms.