摘要
分析电容层析成像系统的组成结构和电容传感器的数学模型,采用有限元法对传感器的敏感场进行仿真,并根据敏感场的分布特点,提出一种将三角形剖分与四边形剖分相结合的方法,完成了敏感场的剖分以及数值计算。在综合考虑敏感场分布的均匀性、传感器的灵敏程度以及测量电路的量程范围要求等性能指标的条件下,确定了传感器的优化设计函数。采用正交设计法对传感器的结构参数进行优化试验,应用RBF神经网络对正交设计的试验结果进行回归分析,并基于改进后的粒子群算法进行寻优,采用优化后的结构参数完成了对电容传感器的优化设计。由实验结果的对比分析可知,正交试验设计的传感器重建图像的精度高于基准传感器的重建精度,而采用RBF神经网络与混沌模拟退火粒子群算法优化的传感器成像效果优于正交试验设计的结果,为获得灵敏度高且可靠性强的电容传感器提供了一种新的优化途径。
The composition of the capacitance tomography system and the mathematical model of the capacitive sensor are analyzed.The finite element method is used to simulate the sensitive field of the sensor.According to the distribution characteristics of the sensitive field,a combination of triangular and quadrilateral splitting is proposed.The method completes the segmentation of the sensitive field and the numerical calculation.The optimal design function of the sensor is determined under the condition that the uniformity of the sensitive field distribution,the sensitivity of the sensor and the range requirement of the measuring circuit are comprehensively considered.The orthogonal design method is used to optimize the structural parameters of the sensor,and the RBF neural network is used to perform regression analysis on the orthogonal design test results.After the particle swarm optimization algorithm is used for optimization,the optimized design of the capacitive sensor is completed by using the optimized structural parameters.From the comparative analysis of the experimental results,the accuracy of the reconstructed image of the sensor designed by orthogonal experiment is higher than that of the reference sensor,while the imaging effect of the sensor optimized by RBF neural network and chaotic simulated annealing particle swarm optimization algorithm is better than the orthogonal experimental design.The results of the experimental design provide a new optimization approach for obtaining capacitive sensors with high sensitivity and reliability.
作者
张晋荣
王莉莉
杨博韬
刘笑
ZHANG Jin-rong;WANG Li-li;YANG Bo-Tao;LIU Xiao(School of Computer Science and Technology, Harbin University of Science and Technology, Harbin 150080,China)
出处
《哈尔滨理工大学学报》
CAS
北大核心
2021年第2期59-67,共9页
Journal of Harbin University of Science and Technology
基金
国家自然科学基金(60572153,60972127)
黑龙江省教育厅计划项目(11541040,12511097)
黑龙江省青年科学基金(QC2012C059)
黑龙江省博士后资助项目(LBH-Z11109).
关键词
电容层析成像
传感器
有限元分析
RBF神经网络
混沌模拟退火粒子群算法
electrical capacitance tomography
sensor
finite element analysis
RBF neural network
chaotic simulated annealing particle swarm algorithm
