Effect of geometrical factors on stability characteristics of portal frames

This paper is concerned with the effect of slanting the columns and changing the ratio of the member lengths on the buckling and post-buckling behaviour of portal frames. Two types of structures are considered: hinged- and rigid-base frames, each type subjected to two loading patterns. In the first pattern (perfect loading), equal vertical loads are applied to the nodes. The second pattern (imperfect loading) consists of the same vertical loads and, additionally, of a horizontal nodal force. Numerical results are presented that show the beneficial effect of slanting the columns by negative angles (i.e. clockwise for the left-hand column and anticlockwise for the right-hand one), and the degrading effect of slanting the columns by positive angles. The effect has been observed both in perfect frames and in the frames subjected to imperfect loading. The paper considers also the effect of the frame geometry on the structure's sensitivity to imperfections. The imperfection sensitivity analysis of the frames is carried out according to a definition of sensitivity which differs somewhat from that commonly used.     

Optimal Placement of Phasor Measurement Units with Linear and Non-linear Models

Abstract—This article presents a non-linear programming-based model for the optimal placement of phasor measurement units. The optimal phasor measurement units placement is formulated to minimize the number of phasor measurement units required for full system observability and to maximize the measurement redundancy at all buses in a power system. A sequential quadratic programming algorithm is used for the solution of the proposed model. The existence of power flow and injection measurements, the limited phasor measurement units channel capacity, the lack of communication facilities in substations, and the single phasor measurement units loss are also incorporated into the initial proposed formulation. The non-linear programming model is applied to IEEE 14- and 118-bus test systems in MATLAB. The accuracy and the effectiveness of the proposed method is verified by comparing the simulation results to those obtained by a binary integer programming model also implemented in MATLAB. The comparative study shows that the proposed non-linear programming model yields the same number of phasor measurement units as the binary integer programming model. A remarkable advantage of the non-linear programming against binary integer linear programming is its capability to give more than one optimal solution, each one having the same minimum number of phasor measurement units (same minimum objective value), but at different locations.     

An Evaluation of Different Controllers Based on the Network Frequency Maintenance Strategy for a Large and Multi-control-area Interconnected Power System

Abstract—Maintaining the network frequency at its nominal value due to the load changes is one of the most important issues in the control and operation schedule of a large and multi-area interconnected power system. For the effective control solution, two types of tie-line bias control strategy-based controllers have been applied, i.e., classical and improved controllers. The classical controllers using conventional regulators, including integral, proportional–integral and proportional-integral-derivative, have achieved initial control results to bring the steady state back to the network. However, due to the very poor control features (e.g., large overshoots and long settling times), they need to be replaced with improved controllers, such as fuzzy logic and artificial neural network. To obtain an entire evaluation of the application of different load-frequency controllers, a five-control-area interconnected power system was built as a typical case study. Three improved controllers using proportional-plus-integral-based fuzzy logic and artificial neural network-based non-linear autoregressive moving average (NARMA-L2) architectures as well as their hybrid combination will also be investigated in this study. Simulation results obtained reveal that the improved controllers have obtained smaller overshoots (from 13.95 to 84.18%) and shorter settling times (from 19.91 to 65.71%) compared with the classical controllers.     

Practical Implementation of Novel Robust Position Control Scheme for Synchronous Reluctance Motor

A variable structure robust position controller is presented for a three-phase synchronous reluctance motor. Linear control and variable structure control and pulse-width modulation generation are combined. These provide robust, fast, and accurate position control without the penalty of high chattering. Schemes, including conventional sliding-mode control and the proposed scheme, are tested under parameter variations, and sudden perturbations are applied to the system at a specific time, then compared. This scheme uses both advantages of traditional variable structure control methods and linear methods. The disadvantages of each method, such as chattering and parameter sensitivity, are minimized. Results demonstrate that the proposed technique preserves the classical linear position control merits, while both the steady-state and transient behavior are significantly improved in terms of robustness, accuracy, and low ripple. The results also prove that the position reference command is perfectly tracked in spite of motor parameter uncertainties and load torque disturbance in control of the system that uses the new schemes.     

基于稀疏分解和混沌理论的微弱信号检测

针对数字锁定放大器中稀疏分解算法检测精度受限的缺点,结合非线性方法进行补偿,采用混沌检测联合稀疏分解的算法设计了数字锁定放大器。数字锁定放大器借助稀疏分解算法估计信号参数,根据估计得到的参数设计对应混沌系统;利用混沌系统检测信号的过程中产生间歇性混沌状态,并通过测量间歇性混沌状态的周期,精确检测信号。最终,在信号的信噪比低至-20 dB时可以成功检测信号,频率精度达到0.05 Hz,提供了一种非线性微弱信号检测方法。     

Simplified probabilistic non-linear assessment of existing railway bridges

An objective of the European Commission's Sixth Framework Research Project, Sustainable Bridges, is to advance understanding of the behaviour of existing railway bridges and develop tools to assess their ability to safely handle future traffic demands and extend their service lives. This paper presents the findings of a study that reviews structural safety models applicable to the assessment of existing bridges. The study proposes the use of simplified probabilistic non-linear structural analysis methods to provide more accurate assessments of the load capacity of bridge systems than traditional methods. The simplified methods use the results of a limited number of deterministic non-linear structural analyses and apply these results into a reliability framework. The application of the proposed methods is illustrated by assessing the safety of an existing bridge. The accuracy and efficiency of the simplified methods are verified by comparing the results of the simplified methods to those obtained from full probabilistic non-linear analysis procedures.     

Ti-6Al-4V钛合金网篮组织与拉伸性能 定量关系研究

建立钛合金显微组织-拉伸性能定量关系对于组织性能的控制及优化具有重要意义。本文基于定量金相学和体视学原理,利用多元非线性回归方法建立了Ti-6Al-4V钛合金网篮组织-拉伸性能定量关系模型。模型自变量包括片状α相厚度、片状α相长宽比,因变量为合金的室温拉伸性能。经验证表明,所建立的多元回归模型对合金的拉伸性能具有较高的预测精度,说明模型的构建方法是可行的。     

应用RB无迹卡尔曼滤波组合导航提高GPS重获信号后的导航精度

针对微机电-船舶惯性导航/全球定位(MEMS-SINS/GPS)组合导航系统在GPS信号中断时造成的强非线性误差及重获信号后精度变差的问题,设计了基于Rao-Blackwellised无迹卡尔曼滤波(RB-UKF)的组合导航算法。首先,基于捷联平台欧拉失准角定义了姿态误差,建立了捷联惯导系统的非线性误差传播方程。然后,针对组合导航的状态方程为非线性而量测方程呈线性的特点,设计了RB-UKF算法,在保证精度的同时降低了计算量。最后,设计了滤波算法总体结构,分别给出了GPS信号正常时和中断时组合导航滤波计算的流程。将提出的算法用于跑车实验,结果表明:在GPS失锁20s和40s再重获信号之后,使用RB-UKF算法的组合导航系统位置精度分别优于6m和7.5m,比扩展卡尔曼滤波(EKF)算法精度提高了1.5倍以上,误差收敛速度提高了1.88~16.5倍,计算量比UKF量测更新的计算量减小了41.7%。实验显示该方法显著提升了组合导航系统GPS信号中断再恢复后的滤波精度,且易于工程实现。     

煤炭需求量预测非线性动力学模型

突破传统分析方法，引入浑沌分析，对煤炭需求量演规律进行探索，以煤炭需求量时间序列数据为基础，从非线性动力学观点出发，研究其所蕴含的动力学特征。求解描述该系统的相空间特征及其吸了，建立煤炭需求量预测非线性动力学模型，采用Ｋｏｌｍｏｇｏｒｏｖ熵分析给出煤炭需求量确定性预测的时间尺度。     

非线性函数协方差传播公式的两种方法及其等价性

本文将非线性观测值函数泰勒级数展开，并取至二次项，得到线性一二次项形式。一方面在线性空间[L，M]］n中推导出含有二次项的非线性函数的广义协方差传播公式，该公式在形式上类似于经典欧氏空间中的协方差传播公式；另一方面又利用矩阵理论等推导出含有二次项的非线性函数的协方差传播公式的矩阵形式。分析并证明了二者的等价性及与经典传播公式间的关系。