Optimization Of Boiler Combustion Systems Based On MRF And Adaptive Control

System of boiler combustion optimization target decomposition, it is concluded that the basic task of the boiler combustion control system is adapt to the heat generated by combustion of the fuel steam load demand. At the same time, so as to ensure the boiler to the safe and economic operation. From the point of view of control system optimization, the boiler combustion system as a three input three output system, the input respectively amount of fuel, air volume and air guide; the three output respectively the main steam pressure, furnace excess air coefficient and furnace negative pressure.Because the Markov random field and Gibbs random field with equivalent, MRF using and model parameters estimation are considered together, using the approximate method for calculating the joint probability, the a priori knowledge is integrated to the model. So the estimation of parameters and the recognition of the alternating iteration are the main methods of maximum likelihood estimation, maximum pseudo likelihood and dynamic Monte Carlo method.. Algorithm to improve the anti-jamming performance of the Gauss Markov random field model, starting from the boundary curve of density function and Markov chain and taking into account the disturbance compensation is proposed.Industrial production process of the controlled object with larger capacity lag, load changes larger or other disturbance is strong, this paper proposes a Markov with airport and model free adaptive control combination, in the estimation of parameters, using a dynamic Monte Carlo method, and the model of mining using Gauss Markov random field model and realize to compensate the disturbance, minus the vice loop burden.Using Markov random field and adaptive control are combined, in the estimation of parameters, using a dynamic Monte Carlo method and Gauss Markov random field model and realize the control of the boiler combustion system optimization based on. Simulation results show that the Markov random field and adaptive control combined control effect than PID control effect is much better, the improved adaptive control in the system of anti disturbance and delay adaptability has greatly improved, the ideal effect.