Research On Intelligent Decoupling Control Of Coke-oven Collector Pressure

Energy has important influence in the economic development of a country. Coal is the main energy in our country. In China, its consumption accounts for about75%of the nation’s energy, and it is the basic state policy of Chinese energy development to vigorously develop coal enterprise for meeting the need of the development of economic construction. But coal applied in industrial production directly are not common, most of them are used in industry after reprocessing. At present, the most widely using and the most reasonable way is coking chemical industry. It is an important coal reprocessing industry. In coking production process, coke oven gas collector pressure is important influence in coking production, the stability of which directly affects coke oven life and coke production, so this paper considers the collector pressure as the research object and use the intelligent control algorithm to ensure the pressure stabilize in a certain range. It can also ensure the productive safety and bring enterprises the most profitable.This paper gives the research situation of the collector pressure control at home and abroad, and analyzes different factors influencing in the collector pressure control. Based on it, take a method with mechanism model to build a nonlinear model of the collector pressure system of coke oven. With the simulation, it shows that coke oven gas collector pressure control system is complex nonlinear system whit coupled multi-variable system.Then, this paper brings a method of inverse system based on dynamic fuzzy neural network. Considered the industrial environment, the model expression of inverse system of coke oven gas collector pressure is not accurately known, therefore, take fuzzy neural network approximates its inverse system expression to enhance its adaptability and anti-interference for the original system. It can break through the bottleneck of traditional inverse system analytical method to realize decoupling. Also, take the single neuron adaptive PID controller to control the decoupling system, it can change control parameters in real-time to improve system performance.Finally, in the environment of MATLAB, build simulation models of coke oven gas collector pressure system and its decoupling control system. With simulation curve, it shows that the coupling between the collectors pressure is basically eliminated and the effect of the control is good.