基于鸽群算法的传感器优化布置方法研究

发布时间：2018-05-03 15:29

本文选题：结构健康监测 + 传感器优化布置　；参考：《大连理工大学》2016年硕士论文

【摘要】：传感器优化布置是结构健康监测中的首要环节,如何在给定传感器数量的情况下,优化传感器的位置有重要的现实意义。群智能优化算法具有较强的全局寻优能力,可为优化准则寻找全局最优解。目前大多数群智能优化算法存在早熟收敛,算法循环次数较多,在高维情况下收敛速度较慢,易陷入局部最优等问题。为解决全局数值最优解问题,本文提出了一种新型群智能优化算法鸽群算法(Pigeon Colony Algorithm,简称PCA)并实现了PCA的离散化,成功用于传感器优化布置。本文主要研究工作如下：(1)原创地提出了一种新型群智能优化算法：PCA。介绍了PCA的构思,详细给出了PCA算法的三大过程：起飞、飞行和归巢；总结了算法的具体步骤和公式；以及PCA的流程图和伪代码。(2)对PCA进行了参数分析。将PCA参数进行了分组对比,给出了参数选取的合理范围；通过算例对PCA的迭代过程进行了演示,证实了参数范围的适用性。(3)采用PCA对低维函数、高维函数和非线性方程组进行优化测试,并与标准遗传算法以及粒子群算法进行了对比分析,结果表明PCA的特点：1)算法对目标函数的性质要求不高,可以是函数表达式,也可以是非函数形式的表示形式；2)对低维函数算法具有全局收敛性较强、算法循环次数少、收敛速度快的特点；3)算法对高维、多峰值、复杂问题具有较强的全局收敛性、较少的循环次数以及较高的稳定性。(4)将PCA算法引入传感器优化布置。以三维模态置信准则作为优化目标函数,采用美国中佛罗里达大学建立的桥梁基准模型为例,分别对考虑冗余和不考虑冗余性的两种三维模态置信准则进行了对比分析,结果表明,PCA算法可以很好地应用于传感器优化布置领域,考虑冗余性的三维模态置信准则可以有效地拟合模态振型,传感器的布设位置更为合理。(5)以润扬长江大桥作为工程算例,对大桥南汊桥的主梁建模,对主梁93个节点进行传感器优化布置,分别选取三维模态置信准则的非对角元最大值最小化和非对角元平均值最小化作为目标函数,用PCA对两种目标函数进行了优化,并对比分析了考虑冗余性的优化布置方案,证实了考虑冗余性的优化布置方案的传感器布设位置更为合理,证实了PCA在传感器优化布置应用中的可行性。
[Abstract]:The optimal arrangement of sensors is the most important link in structural health monitoring. How to optimize the position of sensors is of great practical significance when the number of sensors is given. The swarm intelligence optimization algorithm has a strong global optimization ability and can find the global optimal solution for the optimization criteria. At present, most swarm intelligence optimization algorithms have premature convergence, many circulations, slow convergence speed in high dimensional cases, and easy to fall into local optimization problems. In order to solve the problem of global numerical optimal solution, a new pigeon Colony algorithm (pigeon Colony algorithm) is proposed in this paper, and the PCA is discretized and successfully used in the optimal placement of sensors. The main research work of this paper is as follows: (1) A new swarm intelligence optimization algorithm: PCA-1 is proposed in this paper. This paper introduces the conception of PCA, and gives the three major processes of PCA algorithm: take-off, flight and homing, summarizes the concrete steps and formulas of the algorithm, and analyzes the parameters of PCA by flow chart and pseudo-code of PCA. The PCA parameters are grouped and compared, the reasonable range of parameter selection is given, the iteration process of PCA is demonstrated by an example, and the applicability of the parameter range is proved. The optimization test of high-dimensional function and nonlinear equations is carried out, and compared with the standard genetic algorithm and particle swarm optimization algorithm. The results show that the characteristic of PCA is that the property of the objective function is not high, and it can be a function expression. It can also be expressed in the form of non-function, which has strong global convergence to low dimensional function algorithm, less cycle times and faster convergence speed.) the algorithm has strong global convergence for high-dimensional, multi-peak and complex problems. The PCA algorithm is introduced to the sensor layout optimization with less cycle times and higher stability. Taking the three-dimensional modal confidence criterion as the optimization objective function and the bridge reference model established by the University of Central Florida in the United States as an example, two kinds of three dimensional modal confidence criteria with or without redundancy are compared and analyzed. The results show that the PCA algorithm can be well applied in the field of sensor optimal layout, and the 3D modal confidence criterion considering redundancy can fit the modal mode effectively. Taking Runyang Yangtze River Bridge as an engineering example, the main girder of the bridge is modeled and the 93 nodes of the main beam are optimized. The maximum value minimization of non-diagonal elements and the minimization of the mean value of non-diagonal elements are selected as objective functions respectively. The two objective functions are optimized by PCA, and the optimal layout scheme considering redundancy is compared and analyzed. It is proved that the optimal sensor placement with redundancy is more reasonable, and the feasibility of PCA in the application of sensor optimal layout is confirmed.
【学位授予单位】：大连理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TU317;TP212

【相似文献】