乙醇-水溶液电化学浸渍体系的神经网络优化

乙醇-水溶液体系电化学浸渍具有很高的实用价值。采用人工神经网络对实验数据进行建模。采用基于学习参数模糊自适应调整的误差反向传播算法提高网络训练速度。适当拓宽实验各因素的水平范围,经过不同因素、不同水平间的组合模拟,利用建立的神经网络模型预测出二元电化学浸渍体系的适当工艺条件。     

Principles and methodology of fuzzy sets

The principles of fuzzy sets and their role in processing uncertain information will be discussed. The question of knowledge representation that is of significant importance in problems of system modelling will be formulated and considered at the level of fuzzy sets. Modelling and simulation realized with the aid of fuzzy sets are studied in a unified methodological framework. First a notion of the cognitive perspective is applied to articulate the problem in terms of specialized linguistic labels. Fuzzy models are constructed to capture logical relationships between the elements (linguistic labels) of the cognitive perspective. Several different classes of the models distinguished with regard to their structural dependencies will be analysed in depth. Finally a linguistic-numerical transformation constituting a type of model-environment interface will be studied.     

基于软计算思想解决含干扰输入多值异或问题的研究

介绍一种解决含干扰输入多值异或问题的方法。该方法利用遗传算法寻找神经网络结构,使用模糊逻辑控制神经网络的学习率和冲量,并用传播算法训练得到神经网络。结果显示,这种方法能针对问题得到泛化性能、时间复杂度上表现都较为优秀的网络。     

基于模糊穴─穴映射的多变量模糊系统分析

在简要介绍多变量模糊系统的模糊穴－穴映射（ＦＣＭ－Ｉ）模型的基础上，提出了一种对多变量模糊系统进行分解的方法。此法的核心在于认为任何复杂系统的动态都是由其隐含的稳定子动态和不稳定于动态聚合而成，而系统特性则主要取决于决定性子动态的性质。并介绍了确定系统决定性子动态和基于其上的多变量模糊系统渐近分析方法。     

非匹配不确定系统的模糊滑模控制

研究多输入不确定系统的模糊滑模控制问题，利用动态补偿器和不等式技巧设计了滑动模态，解决了非匹配不确定性的抑制问题，并给出了保守性较小的滑模稳定性条件．为克服滑模的抖振，提出了用模糊数模型和双二次函数插值构造模糊滑模控制律的方法．仿真结果表明，该方法对于参数摄动和外部扰动具有良好的鲁棒性．     

Finite‐time attitude‐tracking control for rigid spacecraft with actuator failures and saturation constraints

In this article, the problem of finite‐time attitude‐tracking control for rigid spacecraft is addressed. Uncertainties caused by external disturbances, unknown inertial matrix, actuator failures, and saturation constraints are tackled simultaneously. First, a smooth function that is more qualified to approximate the standard saturation characteristics is presented to deal with the actuator saturation constraints. Second, a fast nonsingular terminal sliding mode (FNTSM) manifold is constructed as a foundation of controllers design. By incorporating the fuzzy logic system into FNTSM technique, a less demanding solution of coping with model uncertainties is provided because the requirement of a prior knowledge of unknown inertial parameters and external disturbances in many existing achievements is removed. To reduce the number of parameters to be estimated, the norm approximation approach is exploited. Subsequently, an antichattering attitude controller is presented such that all the tracking errors converge into arbitrary small domains around the origin in finite time. The result is further extended to obtain a fault‐tolerant controller against completely failed actuators. Finally, numerical simulation is conducted to verify the effectiveness of the proposed control scheme and comparison with relevant literature demonstrates its high performance. Furthermore, an experiment for the large satellite Hubble Space Telescope is carried out to validate the practical feasibility.     

基于模糊神经网络的母线充电保护研究

针对当前母线充电试验时继电保护中存在的一些问题,提出了一种智能型的母线充电保护新方法。在分析母线正常充电和故障元件充电的基础上,找到了正确快速地识别两种情况的特征量,结合模糊神经网络这一新型的人工智能技术,综合利用母线充电时的多种电气量分别提取形成网络的特征输入量,并采用Simpson模糊极小-极大神经网络来形成区分母线正常充电和故障元件充电的模糊模式识别器。利用EMTP程序仿真来获取不同系统参数和各种不同故障情况下模糊神经网络训练和测试所需要的大量样本,通过对模糊神经网络的训练,使网络具备了很强的故障识别能力和较强的泛化能力,结果表明,训练后的网络能快速准确地区分母线在各种运行工况下的正常充电和故障元件充电,从而验证了该方法的有效性。     

基于模糊层次分析法的自动化机房供电风险评估

针对电网调度自动化主站供电系统运行状态和潜在发生的风险,构建自动化机房供电评估指标体系,提出了一种基于模糊层次分析法的自动化机房供电风险评估方法,通过权重选择和分级评估,同时考虑各类影响因素间的相互作用,实现了对自动化机房供电的科学评估与判断,结合实例分析证明具有较强的科学性和实用性。     

3-RRRT并联机器人鲁棒自适应模糊滑模控制

针对3-RRRT型搬运并联机器人提出了一种直接自适应模糊滑模控制方法.根据机器人的系统动力学模型特点,基于Lyapunov函数的综合设计方法和滑模控制理论,利用模糊逻辑系统与自适应技术相结合,提出了一种控制律,然后进行了系统控制仿真实验.结果表明,采用这种直接自适应模糊滑模控制方法的3-RRRT型并联机器人在周期干扰状况下达到了较高的轨迹控制精度,其闭环系统具有较强的自适应性和鲁棒稳定性.     

选矿浓缩过程矿浆浓度的模糊控制

针对治金行业选矿浓缩过程的工艺特点,介绍了基于变频技术的矿浆浓度模糊控制,较好地控制了浓度指标,同时也稳定了矿浆流量并大幅度节能降耗。该设计已在宝钢、武钢等选矿厂中得到成功应用,效益显著。