基于Reflected Sigmoid径向基函数重建温度场优化算法

微波加热在工业领域中广泛应用,利用超声波测温对其温度场重建可以实时监控设备内部温度,重建算法是实现温度场重建的关键。经典的重建算法为最小二乘法,其在重建过程中为保证结果唯一性,要求有效声波路径数大于区域划分数,重建结果的边界处会出现温度信息缺失。优化声波换能器安装位置,减少待测区域边界位置信息损失,提出基于Reflected Sigmoid径向基函数重建温度场算法,结合病态矩阵广义逆矩阵求解,可有效避免子区域划分条件限制,弥补了经典算法的缺陷。通过仿真实验及对比,所提算法可降低5%的误差。

Temperature field optimization algorithm based on Reflected Sigmoid radial basis function re-construction

Microwave heating is widely used in the industrial field. Using ultrasonic temperature measurement to monitor the internal temperature of the device in real time, the reconstruction algorithm is the key to temperature field recon-struction. The second most multiplication method is a classical algorithm. To ensure that the result is unique, the number of effective acoustic path needs to be larger than the number of regional divisions, and the information on the boundary of the reconstruction result will be missing. This paper optimizes the installation position of the acoustic transducer and reduces the information loss of the boundary position of the area to be tested. A temperature field algorithm based on Reflected Sigmoid radial basis function reconstruction is proposed, which is combined with the generalized inverse matrix of ill-conditioned matrix to avoid the sub-region division conditional constraints. Through experimental comparison, the algorithm reduces the error by 5%.