Simulated Feedback Mechanism-Based Rotary Kiln Burning State Cognition Intelligence Method

Abstract

Burning state directly determines the clinker quality index in the rotary kiln sintering process. A simulated feedback mechanism-based rotary kiln burning state cognition intelligence method and a calculating model are explored for the purpose of imitating the human cognition process with repeated comparison and inference. The flame image feature space is optimized progressively using the evaluation of uncertain cognition results with different values of cognition demand to realize the simulated feedback cognition mode from global to local. First, the simulated feedback mechanism-based rotary kiln burning state cognition intelligence method with the coupling operation of the training layer and cognition decision layer is proposed, and the framework of the model is described. Second, the evaluation index system of uncertain cognition results based on the bag of words model, latent semantic analysis method, and entropy theory is constructed. Third, the simulated feedback mechanism based on the evaluation of uncertain cognitive results is established, and a concrete calculation model is given. Fourth, the simulated feedback mechanism-based rotary kiln burning state cognition intelligence system is designed, and the relevant cognition intelligence algorithm is provided. Finally, a simulation experiment is carried out with the collected burning zone flame image of a rotary kiln. An average recognition accuracy of 92.32% is achieved with a minor standard deviation in accuracy. The experimental results show that our method is effective and outperforms other open-loop recognition methods with the global configuration feature of flame image.