Research On Modeling And Optimization Of Fired Brick Tunnel Kiln Firing Process

The key to sustainable development in today’s society is energy and environment, but with the development of industry, energy shortage and environmental pollution problems are becoming more and more serious. However, in Guangxi, most brick enterprises in the tunnel kiln fired brick roasting production process generally may have some issues such as low degree of automation, low production efficiency, product quality is unstable and management is not standardized and so on. Therefore, modeling and optimization of fired brick tunnel kiln firing process, in order to improve the efficiency of resource use and reduce waste emissions, are of great significance.In this paper, on the basis of research and analysis of theory and control key of fired brick tunnel kiln firing process. Firstly, the data variables of the fired brick tunnel kiln firing process are determined; then in (?)he training process of incremental extreme learning machine, due to the random determination of the input weights and the threshold of hidden layer neurons, there may be part of the output weights of the hidden layer neurons is too small to make small contribution to the network output and these neurons are invalidity, it not only makes the network more complicated, but also reduces the stability of the network, so we propose an improved method:incremental extreme machine learning random hidden layer output matrix plus an offset (named by Ⅱ-ELM), and effectiveness of the method is analyzed. Finally, the validity of Ⅱ-ELM is verified by comparing to Ⅰ-ELM in classification and regression problems by using various data in UCI database, and applies the method to model fired brick tunnel kiln firing process.In this paper, a case based reasoning method is selected to optimize the fired brick tunnel kiln firing process. Firstly, in order to improve the efficiency of case retrieval, this paper describes a case classification method based on clustering analysis; Secondly, the process of case based reasoning also uses the the weight extraction based on particle swarm optimization algorithm and case modification based on genetic algorithm; Finally, according to finding similar history case, case reuse and case modified, it gives the case optimized solution in the current operating conditions.Based on the above mentioned research on modeling and optimization of the fired brick tunnel kiln firing process, finally it verifies the feasibility of optimal control of the fired brick tunnel kiln firing process at the semi-physical simulation platform.