The Application Of Particle Swarm Optimization In1000mMW Thermal Power Model Identification

Model Identification of thermal power plants is the foundation of all thermalpower plant control problems. For the actual thermal control problems, research timeto determine model parameters typically occupy a control process design,commissioning and put into operation a higher proportion. Only after determining oridentifying the control system model, the model parameters can be correspondinglyoptimized before applied to practical engineering, and this paper uses PID controltechnology as the control strategy, mainly because of its wide range of applications,robustness, simple algorithm and so on. This paper takes model identificationtechnology and PID parameter optimization techniques as the background to studytypical thermal power plant process based on actual operating data. Particle swarmalgorithm is used to identify the model and optimize the parameter, so that we cananalyze the main factors that affect the results of system identification and pay moreattention to these problems. This provides a exploring way of thermal systemmodeling.PSO is easy to fall into local optimal solution exists, prematurity, and the searchaccuracy is not high, on the basis of particle swarm algorithm, the algorithm isimproved and enhanced. First, particle swarm evolution speed factor and aggregationfactor are introduced in the algorithm, and the inertia factor is expressed as particleswarm evolution speed factor and aggregation factor function, so that the algorithm isdynamic adaptability. Secondly, the algorithm combines genetic algorithm realmutation operator, and updates the particle individual extreme point and globaloptimum through mutation operator, so that the search capability is enhanced. Theimproved algorithm is validated by the validation function, and the results show thatthe algorithm convergence, stability, and global search capability has beensignificantly improved. And improved particle swarm algorithm combines withmature knowledge of thermal model structure, which is used for model identificationand parameter optimization of1000MW thermal power plant, and achieves thedesired results.