Research On Intelligent Predictive Control Method And Its Application For Coke Oven Gas Collector Pressure

Coke is the main raw material in the metallurgical industry for iron-smelting. In the coking process, large amount of by-product gas will be generated from coke ovens. Gas collecting process by effectively recycling shortage of gas is an important link in the iron and steel production, which not only saves energy but also reduces environmental pollution. The gas-collector pressure of coke ovens is an important parameter in the coking production, and its stability directly affects the quality of coke and gas, the life-time of ovens as well as the producing environment. According to a great quantity of literature and on-site study show that, the effect of gas collector pressure control in many coking plant is non-ideal.. Smoking and Burning phenomenon is serious. Gas collector pressure control is the problem of coking industry recognized. Therefore, it is significant for the iron and steel enterprise to study on the control of gas collecting process of coke ovens.Due to the effect of connection, there exists coupling effect between adjacent gas-collectors and the suction. In addition, the gas collector pressure is also affected by influence of the pulse interference of the intermittent coaling. But the traditional separate adjustment strategy is still the main method in current coking production site and the pressure of the gas collector and the suction of the main pipe are regulated separately, which is difficult to achieve ideal control effect. Thus, for the gas collector pressure control problem, the intelligent predictive control strategy as well as its corresponding dedicated configurable control software products are developed, which has provided a new approach in solving the gas-collector pressure optimization problem, and achieved good results in practical application. The main contributions of this article are as follows:In the light of the problems of multi-variable, closed-loop continuous operation and coal charging disturbance in coke oven gas collecting process, the paper presents an identification method for coke oven gas collecting system based on closed-loop subspace. Firstly, subspace identification does not have consistent estimation when the system variables have correlation with noise of multi-variable system. To solve the problem, closed-loop subspace matrix predicting equation binding PLS regression based on the closed-loop subspace identification method PARIM-E is presented. Aiming at the problem of computation complexity when the PARSIM-E method is used for solving the state space model, a closed-loop subspace identification method which is able to solve the state space model quickly combining with open-loop subspace identification method N4SID is presented. At last, through system analysis, the disturbance signal of coal charging is simulated with pulse signal with certain width. Identification and prediction of output pressure of the dataset of coke oven gas collecting process is done through closed-loop subspace identification method.In the premise of consideration of measurable disturbance input of the gas collector pressure system and soft constraints of process index, subspace predictive control for gas collector and calculation method of steady-state target are designed. Firstly, on the basis of direct structure subspace predictive control method, the incremental subspace predictive control method including feed-forward for gas collector pressure is designed aiming at measurable disturbance and constraints of gas collecting system. Then in the light of the timeliness of online optimization solution of predictive control for multiple coke oven gas collectors, laguerre function is introduced to realize the parameterization of input trajectory of predictive control for gas collector pressure. Thus complexity of optimization problem is reduced to save the computing time. At last, in order to obtain the best input and output target value in the process interval of input and output of coke oven gas collector system, the calculation method of steady-state target of gas collector pressure based on subspace matrix prediction model is designed. When the system is under disturbance, through adjusting the control target value, the steady-state input and output of the system can meet the soft constrains.To solve the problems of the complexity of coke oven gas collector pressure and the fact that single control method can hardly get the ideal control effect, the intelligent predictive control algorithm for coke oven gas collector pressure is presented binding the suction optimization of the main pipe cooling drum workshop. Firstly, fuzzy predictive control method is introduced for the problems of model mismatch because of unmodeling dynamic, time-varying parameter, bounded disturbance and so on. The fuzzy control method is used for the correction of the output of subspace predictive control. For the problem that the suction of main pipe is regulated alone without considering the control effect for gas collector pressure, the paper presents multi-level fuzzy judgment method of multiple collector integrated suction state and regulation expert rules of speed of induced draft fan. At last, the data of suction system is identified through ABC-BP algorithm to build the hybrid model of gas collector pressure. Verification of the effectiveness of the method presented in the paper is done based on the model.The control algorithm for gas collector pressure is now designed on different platform, in order to solve the problems of bad portability.long development period and spread. The intelligent predictive control software for the pressure of coke oven gas collector is designed and developed, the application of intelligent predictive control strategy for gas collector pressure is realized in production of coke oven. The software provides general OPC communication interface and secondary development function in the form of special configurable software to keep the applicability for different coke types. The software has already been applied in two coke ovens of different types, and it has reached the pross requirement.