A novel chemical composition estimation model for cement raw material blending process

Raw material blending process is an essential part of the cement production process. The main purpose of the process is to guarantee a certain oxide composition for the raw meal at the outlet of the mill by regulating the four raw materials. But the chemical compositions of raw materials vary from time to time, resulting in difficulties to control the oxide compositions to a predefined value. Therefore, a novel algorithm to estimate the chemical compositions of the raw materials is developed. The paper mainly consists of two parts. In model construction part, a novel constrained least square model is proposed to overcome the deviation introduced by long-term drift of the material components, and the model parameters are estimated with an online strategy. And in validation part, the approach is implemented to two examples including datasets from simulation model and the actual industrial process. The final results show the effectiveness of the proposed method.

A Local Quadratic Embedding Learning Algorithm and Applications for Soft Sensing

Inspired by the tremendous achievements of meta-learning in various fields,this paper proposes the local quadratic embedding learning(LQEL)algorithm for regression problems based on metric learning and neural networks(NNs).First,Mahalanobis metric learning is improved by optimizing the global consistency of the metrics between instances in the input and output space.Then,we further prove that the improved metric learning problem is equivalent to a convex programming problem by relaxing the constraints.Based on the hypothesis of local quadratic interpolation,the algorithm introduces two lightweight NNs;one is used to learn the coefficient matrix in the local quadratic model,and the other is implemented for weight assignment for the prediction results obtained from different local neighbors.Finally,the two sub-mod els are embedded in a unified regression framework,and the parameters are learned by means of a stochastic gradient descent(SGD)algorithm.The proposed algorithm can make full use of the information implied in target labels to find more reliable reference instances.Moreover,it prevents the model degradation caused by sensor drift and unmeasurable variables by modeling variable differences with the LQEL algorithm.Simulation results on multiple benchmark datasets and two practical industrial applications show that the proposed method outperforms several popular regression methods.