采用双相机多尺度方法的机制砂级配测量及空隙率预测

Gradation Measurement and Void Content Prediction of Manufactured Sand Using Dual Camera Multi-Scale Method

为了解决图像法难以准确测量粒径为0.150 mm以下的机制砂颗粒的问题,设计双相机多尺度测量装置,并提出一种使用形态参数预测空隙率的方法.使用基础相机和精密相机搭建测量装置,基于该装置测量机制砂的粒形参数和粒径参数,通过实验数据构建空隙率预测模型.结果表明:双相机多尺度方法的最大级配测量误差为-2.57%,构建的随机森林模型最大空隙率预测误差为0.62%,测量精度满足工程要求.

In order to solve the problem that it is difficult to accurately measure manufactured sand particles with particle size less than 0.150 mm by image method,a dual camera multi-scale measuring device was designed,and a method of predicting void content using morphological parameters was proposed.The basic camera and precision camera were used to build the measuring device,the particle shape parameters and particle size parameters of manufactured sand were measured based on the device,and the void content prediction model was constructed through the experimental data.The results show that the maximum gradation measurement error of the dual camera multi-scale method is-2.57%,and the maximum void content prediction error of the constructed random forest model is 0.62%,the measurement accuracy of which meets the engineering requirements.