From the point of view of function approximation, a new nonlinear time series prediction model based on wavelet network and its learning algorithm by conjugate gradient method are developed. This model is applied to predict faults of propulsion system for spacecraft.
研究了小波网络模型，从函数逼近角度出发，提出基于小波网络模型的非线性时间序列预报模型，给出了基于共轭梯度的学习算法，并把该模型用于空间推进系统的早期故障预报。
Experimental study has revealed that the artificial nerve network method is one of the most effective methods to solve such nonlinear problem as the evaluation of liquefaction of sand bed.
实例研究表明，人工神经网络法是解决砂土液化评判等非线性问题最有效的新方法之一。
Part of the analysis data is used as training sample to establish three layes BP Neural Network, and in stead of the dynamics model with trained BP Neural Network, the nonlinear mapping relationship between the design variable and the mode parameter is acquired.
将部分分析数据作为训练样本，建立了三层BP神经网络，以该神经网络模型取代结构动力学模型，来得到设计变量与模态参数之间的非线性映射关系。
Moreover the authors also analyze the improvement of burst-mode's BER and network transmission capacity when the nonlinear pre-amplifier is used in the burst-mode receiver.
进一步指出了改进非线性突发模式光接收机的BER性能以及网络传输能力的方法。
Artificial neural network is a kind of nonlinear dynamic system that consists of large numbers of simple processing units.
人工神经网络是由大的简单处理单元所构成的非线性动力学系统。
automatic control of channel network is a key solution for modern water-saving irrigation, and it is also a key technique problem that the middle route of the south-to-north water diversion project, the design of controller demands a liner mathematical model, but the channel networks system is a high nonlinear system. a mimo model of multi-channel system built is used to analyze the uncertainty introduced in the procedure of linearization; this uncertainty will be a precondition for the design of robust controller. take the liner model as the nominal one, the uncertainty is expressed by the largest singular value of the distance matrix of the models. at last a simulation example of six channels is given.
多渠段联合的自动化运行是南水北调中线运行控制的关键技术问题，其控制器的设计求解依据于线性化的系统数学模型，而多渠段系统本身是高度非线性耦合系统。针对所建立的多渠段多输入多输出模型，分析数学模型在线性化过程中引入的不确定性，作为系统鲁棒控制器设计的基础。将线性化的模型作为标称模型，不确定性采用模型之间差异矩阵的最大奇异值表示。最后针对一个六渠段串联的算例进行求解，得出其模型不确定性的定量描述。
The runoff that uses existing relevant site measures historical data and data of quantity is defeated by sand, fall, be being handled into travel normalization and advocate on the foundation that composition analyses, use model of network of three-layer BP artificial nerve, stand to prosperous of upper reaches of drainage area of river of billows dark blue between diameter discharge and each correlation factor complex nonlinear map relation undertakes imitate, use draft Newtonian algorithm to undertake training to the model, imitate result achieves expectation precision to ask, use 1982 ~1985 year actual measurement data undertakes model test and verify.
利用现有的相关站点的径流量历史数据和输沙量、降水量数据，在进行规格化处理和主成分分析的基础上，利用三层BP人工神经网络模型，对澜沧江流域上游昌都站径流量和各关联因子之间复杂的非线性映射关系进行模拟，采用拟牛顿算法对模型进行练习，模拟结果达到期望精度要求，并利用1982年~1985年实测数据进行模型验证。
The results validate more validity of nonlinear error correction model on the wavelet neural network than linear vector autoregressive model, and forecast validly the nonlinear economy system.
研究证明，小波神经网络所建立的非线性误差校正模型有较好的预测效果，能够有效地预测非线性经济系统。
It is based on nonlinear homotopic BP neural network with nonlinear normalization method to identify the oil pumping model, and the genetic algorithm to optimize the downtime.
利用非线性规范化方法的非线性同伦BP神经网络对采油模型进行辨识，采用遗传算法优化停机时间。
Artificial neural network inverse method is helpful to system stability and robustness when it is applied to nonlinear TCP flow process.
神经网络逆系统方法应用于非线性的流量控制过程中有助于系统稳定性和鲁棒性。
This paper is a systematic study on continuous nonlinear predictive control theory, control theories combining fuzzy logical and neural network, their application in electric power system. The study is based on nonlinear models of electric power systems with measured interaction, and it makes use of electric power system's own symmetrical and similar structure. It is the first of such studies to propose a design of holographic structure control laws with measured interaction, which suits the characteristics of electric power system. With this design, holographic and decentralized control can be achieved in the simulation of electric power system. Theoretically, this design can simultaneously improve power angle stability and synchronous machine terminal voltage characteristic, and the difficult problems of electric power system control can be solved efficiently.
本文利用电力系统自身的对称性和相似性的结构属性，以关联测量非线性模型为依据，对非线性预测控制理论及其与模糊逻辑、模糊神经网络控制相结合在电力系统中的应用进行了系统的研究，提出了适应电力系统特点的且有关联测量的结构全息预测控制规律的设计方法，使电力系统实现了结构全息分散控制，这种控制能从理论上同时改善电力系统的功角稳定性和发电机端电压的动态特性，从而有效地解决了电力系统控制中所存在的难题，为电力系统的分层控制、局部稳定和全局稳定之间的协调控制提供了一个很好的控制方法。
We use wavelet neural network to construct the nonlinear cointegration function, and give the DFP algorithm for network training.
提出了利用小波神经网络逼近非线性协整函数的方法，并给出了训练小波神经网络的变尺度算法。
Advanced research product both here and abroad in this field is absorbed. The virtue and shortage of traditional compensation method in dry friction is summarized. Then a new damper of magneto-rheological fluid is designed. The new damper is very suitable for control. But thinking of its nonlinear model and disturbance in the dry friction model, the paper adopt intelligent control, by the feedback of speed, a RBF neural network is trained to control input current of damper of magneto-rheological. Ultimately the damper can compensate the negative stiffness of dry friction.
吸收了国内外在该领域的先进研究成果，总结了传统摩擦补偿方法的优点及不足，设计了一种新型磁流变阻尼器作为汽轮发电机电液伺服系统补偿阻尼，考虑到其模型的非线性，以及干摩擦模型存在扰动，采用智能控制的办法，设计一个RBF神经网络来控制磁流变阻尼器的输入电流，最终达到补偿系统干摩擦负刚度的目的。
The artificial neural network approach is used to describe the nonlinear relationship between the properties of materials and the...
该系统可用于结构抗疲劳设计和疲劳可靠性分析、有损结构损伤评估及维修决策。
Based on Spall, the Neural Network that concealed level can automatically increase is adopted to control the Neural Network structure and it can realize optimization control of unfounded nonlinear random system.
因为系统模型未知，本文在Spall［1］的基础上采用了隐层能自动增加的神经网络，通过实时的对神经网络的结构进行控制，可以实现对未建模型的非线性随机系统的最优控制。
The article uses nonlinear modeling of modern statistical to simulate and analyze the effects of landscape pattern by Natural disturbance and human disturbance.1. It chooses 10 indices as a natural disturbance factors such as size, volume, age, soil thickness, shrub cover, herbaceous cover, average altitude, slope, aspect and slope position to construct of artificial neural network model, multi-model adaptive regression splines, Support vector machine model and project pursuit regression model to analog analyze of fractal index separately.
也说明了研究地的景观斑块大小、景观斑块面积和分维数大部分服从正态分布，部分服从Gama分布。3、运用现代统计学中的非线性建模方法，分析自然干扰和人为干扰对景观格局的影响。1选择面积、蓄积、年龄、土层厚度、灌丛盖度、草本盖度、平均海拔、坡度、坡向和坡位10指标作为自然干扰因子，构建人工神经网络模型、多元自适应回归样条模型、支持向量机模型和投影寻踪回归模型分别模拟分析分维数。
This paper utilizes the nonlinear mapped competence and consilient approach to faculative function of Artificial Neural Networks and discusses the application of Back Propagation Networks model in integrated bid evaluation in order to set up a scientific bid evaluation method based on impersonality and experts knowledge. With the unceasing development of computer network technology, this method has a bright perspective of being conducted on the computer network.
本文利用人工神经网络的非线性映射能力以及对任意函数的一致逼近性能，对BP (Back Propagation，多层前向网络)网络模型在综合评标中的应用进行了探讨，目的是建立一种科学的基于客观和专家知识的评标方法；并且随着计算机网络信息技术的不断发展，可将该方法拓展为网上在线评标。
The applications of wavelet neural network in approaching nonlinear function or signal, sorting signal, system identification, dynamic model building and predicting nonstationary time series were research...
对飞控系统的故障诊断是飞行过程中的容错控制和控制律重构的基础，及时准确的对飞控系统中出现的故障进行辨识、定位就成了飞行安全的重要保证。
CHEA is applied to the daily hydroelectric maximum production in the whole Minjiang valley in Fujian Electrical network, and the simulated calculation results are accurate and reasonable, which demonstrates that CHEA provides an efficient way to solve the large-scale dynamic nonlinear programming problems with complicated constraints.
并应用于福建省电网整个闽江流域水电站群的短期日发电量最大问题中，通过采用CHEA算法优化求解，所得结果正确合理，表明CHEA算法为此类具有复杂约束条件的大型、动态非线性规划问题的求解开辟了一种新的行之有效的途径。
The output equation and the current expression of the motor have been analytically derived. The way to extend the operation speed range has been explored and a split-winding theme has been proposed to extend the speed range of the DSPM motor. Based on the finite element analysis, a nonlinear varying-network magnetic circuit model has been proposed. Moreover, a new method of measuring the inductance has ben proposed, which can reflect the strebthening and weakening actions of the armature flux to the PM flux. The reasons of torque ripple in the motor has been analyzed and measures to minimize the torque ripple have been developed.
导出了电机输出方程和绕组电流解析表达式，探讨了扩展电机转速范围的途径，提出的分裂绕组方案能有效扩展电机运行区；在有限元分析基础上进一步提出非线性变网络等效磁路模型；探讨了绕组电感测量方法，提出了峰值电流法以正确测取电枢磁场为增磁和去磁作用时所对应的电感；分析了转矩脉动的原因并提出了抑制转矩脉动的措施；研究了电机设计方法，设计制作了三种规格样机4台。