基于BP神经网络的水泥石电容-含水率关系建立

含水率测量对于水泥石材料来说,具有很大意义。该研究通过类比岩石与水泥石,借鉴地质中的测井技术,尝试通过测量电容,间接测量水泥石含水率。该文通过实验与研究,基于BP神经网络,通过抽样方式、训练方式、训练参数等优化,最终建立一个偏差率约11%的水泥石电容-含水率关系模型。展示出一个测量水泥石含水率的可行技术路线。

Water content measurement is of great significance for cement stone and concrete materials.This study attempts to indirectly measure the water content of cement stone by measuring capacitance by analogy with rock and cement stone and using the logging technology in geology for reference.Through experiments and research,the research team finally established a capacity-water content relationship model of cement paste with a deviation rate of about 11%by using BP neural network and optimizing the sampling method,training method,training parameters,etc.,which shows a feasible technical route for measuring the water content of cement stone.