Multivariable Setting Software System For Municipal Solid Waste Incineration Process

The municipal solid waste incineration power generation has been widely used in the world in terms of volume reduction,harmlessness and resource utilization.The feeder speed,the second combustion stage speed,the drying stage speed and the secondary air volume are key parameters to achieve complete combustion and meet the emission standards.Due to technical reasons,the control level of domestic incinerators still stay at the logic control of process equipment and the loop control of some process parameters.There is also short of methods to judge and decide on operating conditions in the optimization of incineration process to replace the operators' work.In addition,it is difficult to set key parameters mentioned above because the characteristics of non-linear,time-varying,strong coupling and large lag which the incineration process has.Therefore,this thesis developed a multivariable setting software system using case-based reasoning technology to solve the setting problems existing in the current municipal solid waste incineration.The main work as followed:(1)Building the multivariable setting model for incineration process.A case-based reasoning setting model using learning pseudo metric retrieval was proposed for that it is difficult to set variables in real time in the incineration process.The model including four parts: case library construction,case retrieval in which BP neural network was trained,case reuse and case storage.(2)Developing the multivariable setting software system.The system including a database module,setting module,output module and alarm module.According to the optimization model code,the value of key variables in the incineration process can be calculated in real time.A historical database of set variables and influencing factors for case retrieval was also designed according to the process flow.In order to reduce the workload of staff,alarm was set up to monitor the operation of the incineration process in real time.(3)Testing the setting model and software system.A comparison and robustness tests for the setting model showed that the fitting error of this model is lower than other typical methods,which indicated that the introduction of pseudo metric can improve the regression performance,meet the requirement of setting key parameters and prove the reliability and practicability of the model.In addition,the multivariable setting software system can operate normally from the function module performance experiment one by one.It realized functions of setting the key parameters automatically and giving alarm prompts promptly and successfully replaced the existing artificial experience method to provide technical support for the full and stable burning of municipal solid waste.