提出了一种非闭合电极电容层析成像(ECT)传感器结构参数的优化方法.采用均匀设计结合非线性偏最小二乘(NPLS)回归,提取传感器结构参数(电极极板的宽度l、绝缘外壳的壁厚δ1、屏蔽罩与绝缘外壳间距δ2及绝缘外壳材料的相对介电常数ε)与待优化指标(敏感场的均匀度及灵敏度指标p′、最大与最小电容的比值K)间的函数关系,建立相应的优化目标泛函,通过对优化目标泛函的求解,最终获得传感器结构参数的最优值。并以10电极非闭合电极ECT传感器为研究对象,进行了结构参数的优化设计,根据优化结果设计制作了非闭合电极ECT传感器,对其成像进行了仿真与实测。结果表明,参数优化后的传感器图像重建质量优于未优化的传感器。
A method to optimize the structural parameters of an unclosed-electrode sensor of electrical capacitance tomography was presented. The uniform design and non-linear partial least-square (NPLS) regression were used to extract the functional relationship between the structural parameters of the sensor (including the width of electrode l, the thickness of insulating shell δ1, the space between shield and insulating shell δ2, and the pemnttivity of insulating shell ε ), and the optimization targets ( including the index of uniformity and sensitivity p′, and the ratio between maximal and minimal capacitance K ). Based on that, an objective function was established for the optimization. The optimal parameters of the sensor were obtained by solving the objective function. Setting a 10-electrode unclosed-electrodes sensor as the investigation object, the optimization design was implemented for its geometric parameters. The unclosed-electrodes sensor was then fabricated according to the optimized parameters and the imaging simulations and testing were conducted.The simulation results indicated that the quality of the reconstructed images of the optimized sensor was better than that of conventional ones.