基于振动信号分析法的催化剂积炭量测量模型

基于催化剂固体颗粒运动碰撞反应器壁面产生振动信号的机理，借助平滑消噪及偏最小二乘（PLS）回归等多元数据处理手段，结合振动信号的功率谱分析，建立了在一定表观气速下测量结焦催化剂积炭量的PLS回归模型。实验室冷模研究表明，在表观气速为0.06～0.12 m·s^-1的范围内，模型所得催化剂积炭量预测值与真实值的相关系数均在0.893以上，而预测均方根误差(RMSEP)均在0.51以下，具有较高的精度。由此，获得了一种便捷灵敏、安全环保的非侵入式结焦催化剂积炭量测量技术，能够实现反应器内结焦催化剂积炭量的实时在线检测与监控。

A model for determination of amount of coke deposit on catalyst based on vibration signal analysis

Based on the interaction between particles and the internal wall of the reactor and the method of partial least squares (PLS) regression, a PLS model is proposed to predict the amount of coke deposit on catalyst by analyzing the vibration signal power spectrum after denoising at specific superficial gas velocity.Cold model experiments showed that the PLS model was precise for measuring the amount of coke deposit on catalyst at different superficial gas velocities between 0.06—0.12 m·s^-1.The correlation coefficients of the model were all above 0.893, and the values of the root mean square error of prediction (RMSEP) were all below 0.51.Therefore, a sensitive, safe and environmental-friendly method to determine the amount of coke deposit on catalyst was acquired, which is applicable to the online measurement and monitoring of the amount of coke deposit on coked catalyst in reactors.