Research On The Combustion Optimization Of Coal-fired Boiler Based On LSSVM And Improved PSO Algorithm

with the requirement of energy conservation, consumption reduction and emission reduction deepening, the boiler industry could’t reach the requirements of high efficiency and low emission pollution in most cases because of the defects or deficiencies that the most boiler combustion systems have, and the coal-fired boiler is the largest source point in it. Research of the coal-fired boiler combustion optimization adjustment in this paper has a certain practical significance.First of all, build the boiler combustion characteristic models and implement parameters optimization. In view of multi-variable, strong coupling, multi-interference, large lag and other complex characteristics in boiler combustion process, this paper takes the least squares support vector machine(LSSVM) theory into the coal-fired boiler combustion optimization modeling. In order to optimize the internal structure of the LSSVM model to improve its forecast accuracy and according to LSSVM forecasting theory as well as the uncertainly of LSSVM inter parameters selection, we use an improved particle swarm optimization algorithm(PSO) to optimize the parameters of the model, which is to set the inertia weight and learning factor in PSO algorithm to improve its optimization performance. Comparing with the other two parameter optimization algorithm for study simultaneously, the results show that: LSSVM is an effective method to build model which has high fitting degree; a combination of an improved PSO and LSSVM can improve prediction accuracy and generalization ability, and it is superior to the other two methods in the NOX emissions concentration forecast.Secondly,implement the boiler combustion multi-objective optimization. For the lack of being easy to lose the diversity of particles in the late flight and the problem of premature convergence in local optimum value of the PSO algorithm,PSO was corrected. The optimization results show that: the combination of PSO and LSSVM was useful in solving the multiple and conflicting objectives and performing a search to find the best settings of the air conditioning system for ensuring the boilerefficiency while reducing the NOX emission. It gives a feasible optimization adjustment program of each throttle opening and other operation amounts, finally realize the operation of coal-fired boiler combustion optimization.Consequently, design the remote monitoring system of the boiler and the optimization platform. The paper designs a monitoring system based on GSM/GPRS communication network firstly, and then designs a web platform based on the contents of the boiler remote monitoring and the combustion optimization, to realize to monitor the operation condition of boiler and obtain adjustable parameters through online optimizing, which can supply to the boiler operating personnel for referencing and executing. The working efficiency will be improved.