基于改进粒子群算法的PID控制器参数自整定

针对标准粒子群算法寻优精度不高、易出现早熟收敛等缺陷,提出一种自适应混沌移民变异粒子群算法IPSO。该算法通过引入基因距离来反映粒子间合作与竞争的隐性知识,使粒子种群的多样性得到量化,采取自适应混沌移民变异策略对陷入聚集区域的粒子进行处理,使之获得继续搜索的能力,从而防止算法过早陷入局部最优。仿真结果表明,IPSO算法在PID控制器参数寻优问题上具有遗传算法和标准粒子群算法无法比拟的优势。     

基于改进粒子群优化算法的AGV全局路径规划

分析了用人工神经网络模型描述环境时,采用Sigmoid函数作为神经网络作用函数的不足之处,提出采用双曲正切函数作为神经网络的作用函数,使网络更有利于路径优化算法的寻优计算。粒子群优化（Particle Swarm Optimization,PSO）算法具有收敛速度快,需要调节的参数少等优点,但优化过程中容易发生“早熟”收敛,使优化陷入局部极小值。通过引入模拟退火算法、“交叉算子”和“变异算子”,提出了一种新的改进粒子群优化算法（Improved Particle Swarm Optimization,IPSO）来解决AGV全局路径规划问题。仿真结果表明,IPSO具有很强的全局寻优能力,并且收敛速度比PSO快,能够为AGV规划出理想的路径。     

基于改进的量子粒子群算法在QoS组播路由中的研究

针对QoS组播路由问题,提出了一种改进的量子粒子群优化算法。为了更好地求解该问题,算法采用预处理机制。首先将图形网络拓扑转换为树形网络拓扑,在此基础上进行粒子的编解码,从而杜绝了坏粒子及环路的产生,减少了重复粒子;并利用量子粒子群算法进行粒子群遍历寻优,同时在每次粒子位置移动后,均进行粒子群体的交叉和选择操作,以提高粒子群个体的多样性,增强算法的全局寻优能力,加快算法的收敛速度。最后,将该算法与传统的粒子群优化算法进行编程对比。实验仿真结果表明:改进后的量子粒子群优化算法能获得比传统粒子群优化算法更优的解,同时具有更快的收敛速度及全局寻优能力。     

改进粒子群算法在乙烯收率软测量中的应用

针对粒子群算法(PSO)容易陷入局部收敛的问题,提出一种引入反动因子并结合引力定律的方法来改进算法,增强其寻优能力,该改进算法命名为:GPSO算法.该算法利用引力定律快速确定粒子的寻优方向,寻优过程中当粒子陷入局部最优时利用反动因子的引入使粒子跳出局部最优.仿真实验证明该改进算法在收敛速度和寻优能力上都取得了显著效果.最后,用改进的算法优化BP神经网络的参数,获得了乙烯裂解转化率模型,实验结果表明,基于改进算法的神经网络模型能够较好地预测乙烯裂解转化率.     

基于混沌变异算子的小生境量子粒子群算法

针对粒子群算法早熟收敛和搜索精度低的问题,提出了基于混沌变异的小生境量子粒子群算法(NCQPSO).该算法结合小生境技术并加入了淘汰机制.使算法具有良好的全局寻优能力.变尺度混沌变异具有精细的局部遍历搜索性能.使算法具有较高的搜索精度,实验结果表明,NCQPSO算法可有效避免标准PSO(Particle Swarm Optimization)算法的早熟收敛,具有寻优能力强、搜索精度高、稳定性好等优点.也优于原始的量子粒子群算法QPSO(Quantum-behaved Particle Swarm Optimization).     

基于粒子群神经网络线控转向故障诊断的研究

改进的自适应粒子群优化算法根据群体早熟收敛程度和个体适应值来调整惯性权重和阈值系数,以及对粒子速度与位置进行更新,该算法兼顾全局寻优和局部寻优,有效地避免早熟收敛.使用改进的自适应粒子群优化算法训练神经网络,并根据汽车线控转向系统构建故障诊断模型.实验结果表明:与传统的粒子群优化算法、遗传算法训练神经网络相比,基于改进的自适应粒子群优化算法的神经网络能够有效改善神经网络的训练效率,加快了收敛速度,提高故障模式识别的准确率.     

基于改进粒子群优化算法的新型小波神经网研究

本文提出了改进的粒子群优化算法（Improved Particle Swarm Optimization,IPSO）的新型BP 小波 神经网络,并且对非线性辨识问题进行了仿真实验．实验结果表明,基于改进的粒子群优化算法的BP 小波网 络不仅具有小波分析良好的局部特性以及神经网络的学习、分类能力,而且具有粒子群优化算法全局快速寻 优的特点．与简单的粒子群优化算法相比,该方法在收敛性和稳定性方面都有了较明显的提高,验证了它的 合理性和有效性．     

基于IPSO-FNN的风电齿轮箱故障诊断方法

在对风电齿轮箱故障特征提取基础上,利用模糊神经网络(FNN)对齿轮箱故障诊断系统进行网络建模;为了提高网络学习算法效率,采用改进粒子群优化(IPSO)算法对网络参数进行学习。引入适应度方差表征粒子状态,对早熟粒子进行差分进化操作,改善粒子群的多样性。对惯性权重、学习因子进行改进,平衡算法的全局搜索和局部开发能力,提高算法性能。经仿真实验研究表明:所提出的故障诊断方法与FNN,PSO-FNN方法相比,具有较高的诊断精度和较快的收敛速度。     

一种改进的混沌量子粒子群优化算法

通过将量子粒子群优化算法和佳点集法相结合,提出一种改进的混沌量子粒子群优化算法,用于解决复杂函数问题。将佳点集融合到量子粒子群算法中,以提高解空间的遍历性,对函数实现全局寻优。用混沌序列改变惯性权重 w,调节粒子群优化算法的全局和局部寻优能力。采用线性递减速度比例收缩因子η提高搜索速度,避免早熟收敛。用量子Hadamard门对量子编码进行变异,增强种群的多样性,促使粒子跳出局部极值点。对典型复杂函数的仿真结果表明,该混合算法寻优效率高、收敛速度快,能有效避免早熟收敛。     

改进动量粒子群优化神经网络的语音端点检测

为了提高语音端点检测率,提出一种改进动量粒子群优化神经网络的语音端点检测算法(WA-IMPSO-BP)。利用小波分析提取语音信号的特征量,将特征向量作为BP神经网络输入进行学习,并采用粒子群算法优化BP神经网络参数,建立语音端检测模型,在Matlab环境下进行仿真实验。仿真结果表明,WA-IMPSO-BP提高了语音端点检测率,有效降低了虚检率和漏检率,表示WA-IMPSO-BP是一种检测率高,抗噪性能强的语音检测算法。     

Mixed constrained image filter design using particle swarm optimization

This article describes an evolutionary image filter design for noise reduction using particle swarm optimization (PSO), where mixed constraints on the circuit complexity, power, and signal delay are optimized. First, the evaluated values of correctness, complexity, power, and signal delay are introduced to the fitness function. Then PSO autonomously synthesizes a filter. To verify the validity of our method, an image filter for noise reduction was synthesized. The performance of the resultant filter by PSO was similar to that of a genetic algorithm (GA), but the running time of PSO is 10% shorter than that of GA.     

基于改进粒子群算法的自适应陷波器设计

在信号处理中,接收信号常伴随着干扰和噪声,这就需要最优滤波器来实现,其中工频干扰的消除则以自适应陷波器为最优。利用粒子群算法自适应地调节其权值,得到与干扰信号接近的期望信号,最终达到消除干扰得到有用信号的目的。同时,针对此算法存在局部收敛和收敛速度不高的问题,提出了改进方法。计算机仿真结果表明了该改进粒子群算法在自适应陷波器设计上的有效性,并取得了较高的效率。     

Hybridization of spectral filtering with particle swarm optimization for speech signal enhancement

Speech enhancement has received a significant amount of research attention over the past several decades. The enhancement of speech signal is needed so as to improve the degraded signal and the goal is to separate a single mixture into its underlying clean speech and interferer components. This is achieved by having prior knowledge through learning and generation of masks accordingly. Hybridization of the spectral filtering and optimization algorithm is employed for speech enhancement in this paper. The proposed technique uses MMSE (Minimum Mean Squared Error) and PSO (Particle Swarm Optimization) for effective enhancement. The proposed technique is three module technique consisting of pre-processing module, optimization module and spectral filtering module. Loizou’s database and Aurora dataset are used for evaluating the proposed technique using standard evaluation metrics consists of PESQ and SNR. Comparative analysis is also made by comparing with other existing techniques such as MMSE and BNMF. Highest PESQ for proposed technique is 2.75 and highest SNR came about 32.97. The technique gave average PESQ of 2.18 and average SNR of 20.53 which was higher than the average values for other techniques. Hence, we can observe that proposed technique yielded better evaluation metrics than the existing methods.     

Quadratic interpolation based orthogonal learning particle swarm optimization algorithm

Particle swarm optimization (PSO) is a population based algorithm for solving global optimization problems. Owing to its efficiency and simplicity, PSO has attracted many researchers’ attention and developed many variants. Orthogonal learning particle swarm optimization (OLPSO) is proposed as a new variant of PSO that relies on a new learning strategy called orthogonal learning strategy. The OLPSO differs in the utilization of the information of experience from the standard PSO, in which each particle utilizes its historical best experience and globally best experience through linear summation. In OLPSO, particles can fly in better directions by constructing an efficient exemplar through orthogonal experimental design. However, the global version based orthogonal learning PSO (OLPSO-G) still have some drawbacks in solving some complex multimodal function optimization. In this paper, we proposed a quadratic interpolation based OLPSO-G (QIOLPSO-G), in which, a quadratic interpolation based construction strategy for the personal historical best experience is applied. Meanwhile, opposition-based learning, and Gaussian mutation are also introduced into this paper to increase the diversity of the population and discourage the premature convergence. Experiments are conducted on 16 benchmark problems to validate the effectiveness of the QIOLPSO-G, and comparisons are made with four typical PSO algorithms. The results show that the introduction of the three strategies does enhance the effectiveness of the algorithm.     

基于可变学习率BP算法的空调控制系统的研究

通过分析控制器参数学习率和控制器性能之间的关系,设计一种基于可变学习速率反向传播算法VLRBP和模糊神经元网络的变频空调控制系统.该系统不仅可以通过反传误差信号训练控制器参数,而且可以根据网络的当前状态朝最优化方向调整控制器参数的学习率.实验结果表明,该控制系统不仅比传统的空调PID控制器和模糊控制器具有更好的控制性能,而且相比基于标准BP算法和动量BP算法的模糊神经网络控制系统,也具有更快的收敛速度和更好的控制精确度.     

基于RefineNet的端到端语音增强方法

为提高神经网络对语音信号时域波形的直接处理能力,提出了一种基于RefineNet的端到端语音增强方法.本文构建了一个时频分析神经网络,模拟语音信号处理中的短时傅里叶变换,利用RefineNet网络学习含噪语音到纯净语音的特征映射.在模型训练阶段,用多目标联合优化的训练策略将语音增强的评价指标短时客观可懂度(Short-time objective intelligibility,STOI)与信源失真比(Source to distortion ratio,SDR)融入到训练的损失函数.在与具有代表性的传统方法和端到端的深度学习方法的对比实验中,本文提出的算法在客观评价指标上均取得了最好的增强效果,并且在未知噪声和低信噪比条件下表现出更好的抗噪性.     

基于RefineNet的端到端语音增强方法

为提高神经网络对语音信号时域波形的直接处理能力,提出了一种基于RefineNet的端到端语音增强方法.本文构建了一个时频分析神经网络,模拟语音信号处理中的短时傅里叶变换,利用RefineNet网络学习含噪语音到纯净语音的特征映射.在模型训练阶段,用多目标联合优化的训练策略将语音增强的评价指标短时客观可懂度(Short-time objective intelligibility,STOI)与信源失真比(Source to distortion ratio,SDR)融入到训练的损失函数.在与具有代表性的传统方法和端到端的深度学习方法的对比实验中,本文提出的算法在客观评价指标上均取得了最好的增强效果,并且在未知噪声和低信噪比条件下表现出更好的抗噪性.     

A novel cuboid method with particle swarm optimization for real-life noise attenuation from heart sound signals

A novel cuboid method with particle swarm optimization (PSO) is proposed to attenuate real-life noise from heart sound (HS) signals. Firstly, the quasi-cyclic feature of HS is explored. It is found that for each cycle of HS, the fragmental signals at similar time section have similar frequency and energy. Based on this finding, short-time Fourier transform (STFT) is employed to decompose each HS cycle into time–frequency fragments which are called granules. Next, a cuboid is built for each granule to identify and see if it is a constituent of HS or noise. The dimensions of cuboid’s length, width, and height are optimized by PSO. An objective function of PSO based on the normalized autocorrelation coefficient is proposed. Then, granules representing HS are retained and merged into noise-quasi-free HS signal. The proposed de-noising method is assessed using mean square error (MSE) and compared with the recently proposed wavelet multi-threshold method (WMTM) and Tang’s method. The experimental results show that the proposed method not only filters HS signal effectively but also well retains its pathological information.     

基于粒子群优化算法的模拟滤波器设计

采用传统的网络综合法设计计波器存在带宽不精确及阻带衰减过小的问题,为此,提出一种基于粒子群优化算法的无源模拟滤波器优化设计方法.在网络综合法设计的滤波器电路基础上,利用粒子群优化算法对滤波器的整个参数空间进行高效并行搜索直到获得最优的参数值.实例表明,采用该方法设计的滤波器带宽更加准确,且具有更加陡峭的阻带衰减.