Study On The Application Of Intelligent Control And Evolutionary Algorithm To Gas Fired Shuttle Kiln

In this thesis, combustion technology, control system, and design feature of shuttle kiln are analyzed. And also discuss the influence of processing technology on characteristics of inorganic materials such as microstructure, body properties, quality of product. Some kinds of research methods, mathematical models, and control algorithm applied for shuttle kiln are summarized.Problems in application of kiln such as temperature zone coupling, automatic control difficulty, temperature-field disturbance and low heat consumption rate are deeply researched. It is found that pulling the air in the shuttle kiln to adjust the temperature is reliable to achieve kiln atmosphere control with low cost and also verified to be an appropriate means from both technology and control viewpoints. Defects of burning with fixed air-gas ratio in the whole burning period are found, according to analysis of the heat balance and shortage of turbulent convection caused by top exhaust port. Data with intrinsic are extracted from vast scale of data by cluster analysis. The thesis presents the method to train the ANN with the intrinsic data to approach the real kiln for the control of shuttle kiln atmosphere. To realize the betteradvanced control of shuttle kiln atmosphere, soft measure technique, neural network model and evolutionary algorithm are utilized in the research .On the basis of the soft measure model of the content of oxygen in fume and the forecast model of the gas, the air-gas ratio can be hunting in large scale with gene algorithm by judgment of kiln state and setting temperature. With the air-gas ratio, the rule of the fuzzy control can be acquired by Adaptive Network-based Fuzzy Inference System(ANFIS). This ANFIS is effective for shuttle kiln, while the common fuzzy control or PID control is unsuitable.