Enhancing Efficiency and Sustainability: A Predictive Modeling Approach Using Extra Tree Regressor for Cement Kiln Cooler Process

Abstract

Abstract The clinker cooler is an important part of the cement manufacturing process, as it plays a crucial role in controlling the temperature of the clinker leaving the kiln. To optimize the clinker cooler in terms of energy and throughput, this paper presents a methodology based on artificial intelligence and mainly on machine learning, which could revolutionize the cement industry by improving efficiency, reducing costs, and increasing production. The sensor data are used after preprocessing to implement the different models of the supervised regression problem. The models have four output variables which include pressure, temperature, and speed. This aims to identify the right model which is Extra Tree regressor after comparison according to two metrics that are the mean absolute error, and score R2 and to optimize the process parameters in real time.