Research On Online Optimization PID Temperature Control Technology Of Large Brazing Furnace

In this paper,based on the first order delay model and two order delay model duringthe third stage of heating in vacuum brazing furnace as the research background,parameter identification and parameter optimization of PID controller for the twotemperature model is deeply studied.First,through the combined identification algorithmto identify the first-order time-delay model in vacuum brazing furnace and then toidentify the identification of Second-order plus dead-time model, three parameters ofthe PID controller optimized by using genetic algorithm after the model being obtained,we can obtain the optimal parameters of the controller to meet on-line control of timevarying temperature object. At the same time parameter optimization of parameteridentification and controller of Second-order plus dead-time model are studied. Thispaper will focuses on the first-order time delay objects online identification and thedesign of PID parameters use genetic algorithm optimization. The main contents andresearch results are as follows:1. As for the temperature of the time varying, we can use a first-order time-delaymodel to describe it. The traditional recursive least squares algorithm for discretizationcannot accurately identify the delay time. By combining zero frequency matchingmethod and recursive least squares and online identification algorithm with variableforgetting factor is improved, the parameters of online identification of a first-ordertime-delay cam be accurately identified. This algorithm is concise and fast, suitable forindustrial real-time control, and it is in the simulation and verification.2.In the process of first order optimization parameters for time-delay, due to thelag time-variability of first-order brazing furnace in the third stage, correspondencebetween the variation model and controller parameters on the lag time has been studied,and proposes a new way that the interpolation method is added to the PID parameters ofgenetic algorithm optimization, which saves the time of the same charges object tooptimize the parameters. This method is simple and effective, and has been studied itsvalidation.3.in the model of brazing furnace of third stage temperature object, as for the first-order time delay object whose time delay large, the study found that using geneticalgorithm to optimize the PID parameters cann’t achieve good results, and generatesserious overshoot phenomenon. In order to solve the problem of large overshoot causedby the lag time, based on exact model with Smith predictor to eliminate the influence oftime delay on the optimization of PID parameters, which solves the overshoot problemcaused by large delay time. Through building the model simulation which proves thatthe design is feasible.4.In order to ensure the stable and efficient operation of the controller, and todescribe the object more accurately, part of the temperature object using Second-orderplus dead-time model to replace the first-order time-delay temperature model, in theidentification and control of model also obtains very good effect, and built asimulation model to verify.Finally, the control system of the software is tested and experimented, whichpreliminary realized automatic control. According to the analysis and processing ofexperiment data, we can see the whole control system to the completion of a goodbrazing furnace heating.