基于改进相关向量机的锂电池剩余使用寿命预测

精确预测锂离子电池剩余使用寿命对于保障设备安全运行十分重要。但电池寿命预测中存在诸如数据噪声和容量再生等不确定性来源,这将导致预测精度大幅下降。为解决这一问题,使用变分模态分解方法对从充电和容量数据中提取的健康因子进行滤波分解,并利用贝叶斯优化方法对相关参数进行优化,提出一种基于多核相关向量机的锂离子电池剩余使用寿命预测模型。利用美国国家航空航天局(NASA)和Oxford电池数据集对所提出的模型进行验证,研究结果表明：所提出的基于变分模态分解和贝叶斯优化的多核相关向量机(VMD-BAYES-HRVM)方法的预测性能不受预测起始点和截止电压的影响,预测结果准确性更高,95%置信区间的跨度更小,证明了所提出方法的有效性。     

基于贝叶斯理论的新能源锂离子电池剩余寿命预测方法

针对锂离子电池健康因子衰退指标预测不佳,影响电池有效更换的问题,设计基于贝叶斯理论的新能源锂离子电池剩余寿命预测方法。提取新能源锂离子电池的衰退特征,并分析电池衰退变化;通过贝叶斯理论确定电池剩余寿命先验分布,提高电池剩余寿命预测的置信度;根据先验分布结果,构建锂离子电池剩余寿命预测模型,对电池寿命期望函数进行分析,进而实现新能源锂离子电池的有效利用。采用对比实验的形式,验证了该预测方法新能源锂离子电池剩余寿命预测效果更佳,可以应用于实际生活中。     

基于凸优化-寿命参数退化机理模型的锂离子电池剩余使用寿命预测

针对锂离子电池寿命预测中模型普适性差、预测精度不足等问题,提出一种基于凸优化-寿命参数退化机理模型的锂离子电池剩余使用寿命RUL预测方法。首先构造锂离子电池实际容量与其循环周期的退化机理模型。对锂离子电池寿命试验数据进行凸优化降噪处理;基于预处理得到的可靠性较高的数据,采用最小二乘法对所建机理模型的参数进行辨识,从而得到精确的模型表达式,实现锂离子电池RUL的预测。基于NASA锂离子电池数据集预测并评估锂离子电池的RUL,预测结果验证了模型良好的通用性,误差范围为4%左右。     

锂离子电池剩余寿命间接预测方法研究

针对锂离子电池的容量在线测量困难的问题,提出了一种基于优化的融合型间接健康因子和改进的最小二乘支持向量机的锂离子电池剩余寿命间接预测方法.首先采用自适应健康因子提取方法提取了等电压升充电时间序列和等电流降充电时间序列,通过健康因子的线性组合和Box-Cox变换构建了优化的融合型健康因子.然后针对最小二乘支持向量机的超参数调整困难的问题,提出了基于粒子群优化的改进算法.在此基础上,构建了基于优化的融合型健康因子的锂离子电池剩余寿命间接预测方法.实验结果表明,提出的间接预测方法能够较好地拟合容量的退化过程,同时预测精度比基于单一健康因子的方法更高.     

基于CNN-BiLSTM的锂电池剩余使用寿命概率密度预测

通过预测锂离子电池的剩余使用寿命(RUL)，可以对电池实现管理和维护，提升电池的耐用性和安全性。由于锂离子电池在使用过程中，不同的工况条件会增加锂离子电池RUL预测的不确定性，传统的点预测不能对电池的不确定性进行表达，因此提出了一种基于卷积神经网络(CNN)和双向长短时记忆循环神经网络(BiLSTM)的混合神经网络分位数回归的概率密度预测。该方法通过预测不同分位数条件下的电池容量，不仅可以利用中位数和众数对剩余寿命进行点估计，还可以利用核密度估计得到每个循环周期下电池的容量和剩余使用寿命的概率密度分布，为使用者提供更多有效的决策信息。     

基于BiGRU网络的锂电池寿命预测

锂离子电池寿命预测是电池健康管理的一项重要内容,针对锂离子电池使用寿命预测难和单向LSTM神经网络不能充分利用数据信息等问题,提出了一种基于BiGRU的神经网络模型并将电池容量数据作为关键因子的锂离子电池循环寿命预测方法.BiGRU的神经网络预测精度更高,且训练参数较少,验证了BiGRU模型的有效性.     

基于能量加权高斯过程回归的锂离子电池健康状态预测

针对容量再生现象影响锂离子电池健康状态预测(SOH)建模精度的问题,提出一种经验模态分解(EMD)的能量加权高斯过程回归(EWGPR)方法。该方法将容量再生现象看作是锂离子电池容量衰减过程的能量凸现,利用EMD分解获得样本的能量分布情况,根据能量情况计算每个样本的权重,进而建立基于能量加权高斯过程回归的锂离子电池SOH预测模型。基于NASA锂电池数据集的仿真实验结果表明,EWGPR方法比基本GPR算法具有更高的精度和适应性,单步预测和多步预测的均方根误差(RMSE)分别减少了3%和10%。     

基于GWO-SVR的锂电池剩余使用寿命预测

锂离子电池剩余使用寿命预测在电池管理系统中发挥着重要作用,准确预测其剩余使用寿命能够保障电池的安全稳定运行。由于支持向量回归SVR（support vector regression）参数内核选择较为困难,为此提出灰狼优化—支持向量回归GWO-SVR（gray wolf optimization-SVR）方法,使用灰狼算法优化其内核参数,根据NASA预测中心提供的电池数据集对该方法进行了验证。通过与SVR方法进行对比发现,所提GWO-SVR方法的预测精度得到显著提高;在此基础上与ALO-SVR方法进行对比,证明所提方法平均相对误差降低了7.16%,预测精度更高,有效地提高了锂离子电池剩余寿命预测的精确性。     

一种多时间尺度下的锂离子电池剩余寿命混合预测方法

针对锂离子电池储能系统寿命预测不准确问题,提出了一种多时间尺度下的电池剩余寿命混合预测方法。首先,通过带自适应噪声的完全集成经验模态分解(CEEMDAN),将原始数据分解为多个低频和高频分量。接着,利用GS-LM模型对其低频分量进行拟合预测,再将分解后的高频分量作为LSM-STW模型(滑动时间窗口和长短期记忆神经网络)的输入,得到高频分量预测结果。最后,将所有预测结果相加,得到最终的锂离子电池剩余寿命预测。实验结果表明,所提出的预测方法可以准确地预测锂离子电池容量退化趋势,不同预测起点对预测结果影响较小,并能够捕捉电池容量的再生现象。     

基于IGWO-MKELM的锂离子电池剩余使用寿命预测

随着锂离子电池在储能系统中比例迅速增大,为避免因电池性能退化导致的事故,如何准确预测锂离子电池剩余使用寿命就成为保障储能系统可靠运行的关键。针对锂离子电池剩余使用寿命预测的问题,提出一种改进灰狼优化多核极限学习机（IGWO-MKELM）预测方法。首先从电池充放电过程中提取能够表征电池寿命退化的间接健康因子作为输入量,然后采用改进灰狼算法对多核极限学习机参数进行寻优,建立改进灰狼优化多核极限学习机预测方法,最后使用NASA电池数据集进行仿真实验。结果表明,IGWO-MKELM方法可以更加精确地预测锂离子电池剩余寿命。     

An Application of the Recombination and Generation Theory by Shockley, Read and Hall to Biological Ion Channels

We have applied the Shockley-Read-Hall (SRH) model for the generation and recombination of charged carriers to biological ion channels. We show how to include this important effect in the traditional PNP model. The idea is to use the software of computational electronics that has been developed to solve Shockley’s equations. In particular we have used the simulator PROPHET to simulate biological ion channels and to include particle like properties and dynamics such as the capture and release of ions. The considerable reduction of effective diffusion coefficients can be well simulated. The saturation effect observed in current-concentration curves, which is not predicted by the conventional PNP model, has been successfully reproduced in our simulation. We also show that PROPHET can be used to perform both steady state and time dependent simulations for ion channels. The timescale can be microseconds, far beyond the range of molecular dynamics simulations. Our results demonstrate the useful role of PROPHET simulations in a multi-scale simulation approach.     

分布式电源的配电网重构和电容器优化投切

综合考虑最优潮流、配电网重构和电容器优化投切,建立了计及分布式电源的配电网经济运行模型。鉴于模型的复杂性,采用结合最优潮流和改进蚁群算法的混合优化方法进行求解。该方法将最优潮流嵌入改进蚁群算法中,利用最优潮流求解考虑网络安全约束的分布式电源优化调度问题,利用改进蚁群算法优化配电网结构和电容器档位。为了提高蚁群算法的优化效率,建立含有局部搜索蚂蚁的混合蚁群,平衡蚁群算法的全局和局部搜索能力。通过对16节点和33节点的测试系统仿真,表明提出的模型和算法正确有效。     

基于电磁感应和超宽带雷达的新型探测系统

超宽带探地雷达是近年来迅速发展起来的一种高新无损检测技术,可对地下目标进行高分辨率成像,但很难直接判断目标是金属或非金属的,而且无法得到准确的探测深度。针对地下目标材质识别和获取准确探测深度的需求,研制了一种将电磁感应与超宽带雷达相结合的新的探测系统。该系统可以直接分辨金属和非金属目标,并能反演出介质的介电常数,从而得到目标的准确深度。将该探测系统在测试模型上进行了实验,取得了很好的效果。     

Optimal location and sizing determination of Distributed Generation and DSTATCOM using Particle Swarm Optimization algorithm

A Particle Swarm Optimization algorithm for finding the optimal location and sizing of Distributed Generation and Distribution STATicCOMpensator (DSTATCOM) with the aim of reducing the total power loss along with voltage profile improvement of Radial Distribution System is proposed in this paper. The new-fangled formulation projected is inspired by the idea that the optimum placement of the DG and DSTATCOM can facilitate in minimization of the line loss and voltage dips in Radial Distribution Systems. A complete performance analysis is carried out on 12, 34 and 69 bus radial distribution test systems and each test system has five different cases. The results analyzed using Loss Sensitivity Factor shows the optimal placement and sizing of DG and DSTATCOM in Radial Distribution System effectively improves the voltage profile and reduces the total power losses of the system.     

MMC子模块控制器设计与测试

子模块控制器（ sub-module controller , SMC ）的控制保护功能对模块化多电平换流器直流输电（ modular multilevel converter based high voltage direct current ,MMC-HVDC）的运行有着极大的影响。首先,介绍了子模块与SMC硬件架构,开发调试了基于复杂可编程逻辑器件（ complex programmable logic device , CPLD）的SMC控制程序,研究了SMC与其上层控制阀基控制器（ valve based controller , VBC ）之间的协调控制以及子模块保护策略。最后对子模块进行了稳态的空载、带载实验以及MMC-HVDC物理模拟系统的系统实验,实验结果证明了子模块的性能良好,SMC可以与VBC进行有效通讯并可以对子模块进行有效控制保护以及稳态测试平台的有效性。     

SIP协议与ISUP协议互通研究

会话初始协议(SIP)是IETF定义的基于IP的应用层控制协议，是NGN网络控制的核心协议；ISDN用户部分(ISUP)协议是7号信令系统的一种主要协议。文章介绍并深入分析了SIP和ISUP协议，从互通的层面对二者进行了分析比较，给出了2种协议之间具体的互通结构、映射方法以及呼叫流程。基于网络融合的思想做了有益的探索和实践，为设计MGC／软交换核心设备提供了依据，为实现PSTN网络向软交换网络的平滑过渡提供了解决方案。     

An Approach to Speech Emotion Classification Using k-NN and SVMs

The interaction between humans and machines has become an issue of concern in recent years. Besides facial ex-pressions or gestures, speech has been evidenced as one of the foremost promising modalities for automatic emotion recognition. Effective computing means to support HCI (Human-Computer Interaction) at a psychological level, al-lowing PCs to adjust their reactions as per human requirements. Therefore, the recognition of emotion is pivotal in High-level interactions. Each Emotion has distinctive properties that form us to recognize them. The acoustic signal produced for identical expression or sentence changes is essentially a direct result of biophysical changes, (for example, the stress instigated narrowing of the larynx) set off by emotions. This connection between acoustic cues and emotions made Speech Emotion Recognition one of the moving subjects of the emotive computing area. The most motivation behind a Speech Emotion Recognition algorithm is to observe the emotional condition of a speaker from recorded Speech signals. The results from the application of k-NN and OVA-SVM for MFCC features without and with a feature selection approach are presented in this research. The MFCC features from the audio signal were initially extracted to characterize the properties of emotional speech. Secondly, nine basic statistical measures were calculated from MFCC and 117-dimensional features were consequently obtained to train the classifiers for seven different classes (Anger, Happiness, Disgust, Fear, Sadness, Disgust, Boredom and Neutral) of emotions. Next, Classification was done in four steps. First, all the 117-features are classified using both classifiers. Second, the best classifier was found and then features were scaled to [-1, 1] and classified. In the third step, the with or without feature scaling which gives better performance was derived from the results of the second step and the classification was done for each of the basic sta-tistical measures separately. Finally, in the fourth step, the combination of statistical measures which gives better per-formance was derived using the forward feature selection method Experiments were carried out using k-NN with different k values and a linear OVA-based SVM classifier with different optimal values. Berlin emotional speech da-tabase for the German language was utilized for testing the planned methodology and recognition rates as high as 60% accomplished for the recognition of emotion from voice signal for the set of statistical measures (median, maximum, mean, Inter-quartile range, skewness). OVA-SVM performs better than k-NN and the use of the feature selection technique gives a high rate.     

Robust fuzzy logic power system stabilizer based on evolution and learning

A robust fuzzy logic power system stabilizer (FLPSS) based on evolution and learning is proposed in this paper. A hybrid algorithm that combines learning and evolution is developed whereby each one complements other’s strength. Parameters of FLPSS are encoded in chromosome (individual) of genetic algorithm (GA) population. Population of FLPSS in GA learns to stabilize electromechanical oscillations in power system at an operating point, as the best fitness becomes large steady value during successive generations. Operating region of FLPSS is enlarged by learning more operating points over the operating domain. Best FLPSS drawn from last generation is saved as designed FLPSS. Effectiveness of the proposed method is validated on a single machine infinite bus (SMIB) power system. Promising optimal stabilizing performance with designed FLPSS for considered power system is obtained at wide range of operating points.     

数字式电力载波技术现状及其应用

介绍了挪威Ｎｅｒａ公司Ａ．Ｃ．Ｅ．３２ＤＰＬＣ系统的组成及特点,着重谈了语音编码和正交幅度调制技术,并简要介绍了其在电力系统通信组网中的现状和前景。     

基于GATS混合算法的PSS与SVC控制器参数设计

随着电力网络规模的扩大，电力系统优化问题日益复杂，故提出了一种采用遗传禁忌GATS混合优化策略对电力系统稳定器PSS和静止无功补偿器SVC附加线性稳定控制器进行参数协调优化的设计方法。该方法结合遗传算法GA和禁忌搜索算法TS各自的优点，将禁忌搜索引入到遗传算法的变异操作，改进了遗传算法的变异算子，具有比常规遗传算法更强的局部搜索能力。在10机新英格兰电力系统上对该优化方法进行了测试。特征值分析表明，该设计方法能有效地将多种不同运行方式下系统的特征根移到复平面目标函数限定的区域内，保证了小扰动稳定性控制的鲁棒。同时还对不同优化方法的收敛性及计算时间进行了比对，结果表明遗传禁忌混合策略的性能优于常规遗传算法以及遗传模拟退火混合优化策略。