极限学习机在电容层析成像中的应用

针对复合材料的无损检测技术,介绍了一种基于极限学习机(ELM)的电容层析成像算法(ECT)。为保证样本的代表性,随机生成各类训练样本模型。对训练样本及测试样本模型的仿真电容值及其灰度值进行归一化处理,采用极限学习机进行训练及测试。给出了仿真结果及分析,探讨了文中方法的优缺点,该方法具有优越的泛化能力和学习速度,且成像精度得到较大提高,具有一定的参考价值。

Application of Extreme Learning Machine in Electrical Capacitance Tomography

The Extreme Learning Machine (ELM) is a simple hidden layer feedforward neural network algorithm that is simple in structure, easy to use, and effective. In this paper, ELM-based capacitance tomography (ECT) image reconstruction methods are studied. Firstly, in order to ensure the representativeness of samples, various training sample models are randomly generated, including: single object, two objects, three objects, ring and layer. Secondly, the simulated capacitance value and its gray value of the training sample and the test sample model are normalized, and the ultimate learning machine is used for training and testing. Finally, the simulation results and analysis are given, and the ELM algorithm is discussed. The advantages and disadvantages show that compared with the LBP algorithm and the Landweber iterative algorithm, the ELM based ECT image reconstruction algorithm has superior generalization ability and learning speed, and the imaging accuracy is greatly improved.