LPV Robust Gain Control For Power Station Boiler-Steam Turbine Coordinated System

The boiler-turbine coordinated control system in a thermal power generating unit can follow the load instruction to quickly coordinate the energy balance between the power plant boiler and the turbine.Due to the strong coupling between the main steam pressure control circuit and the power output control circuit of the boiler turbine,with large time delay and uncertainty,the automatic coordinated control of the two circuits is a very challenging problem.In this paper,load is used as a scheduling variable,and the LPV model of the machine-boiler coordination system is improved by a particle swarm optimization algorithm.Based on th e LPV model,the H_∞/H₂ robust variable gain control method is improved.The system is verified for several operating conditions that are frequently studied Control effect.The main research results obtained are summarized as follows:（1）Using local modeling method based on state space model,using Jacobian linearization method to obtain the local model of the inverted pendulum,using particle swarm optimization algorithm to identify the LPV model of the inverted pendulum,and obtaining the multi-cell LPV based on singular value decomposition model.In the original robust variable gain control,a symmetric matrix X is required to satisfy multiple LMIs,which greatly increases the conservativeness of the system.This paper introduces relaxation variables in the original robust H_∞variable gain control to reduce The conservativeness of the controller design is compared.In the case of model mismatch,simulation comparison with robust control shows that the improved robust variable gain controller has a good control effect.（2）For the Bell-?str?m machine-furnace coordination system,the particle swarm algorithm was used to identify the global LPV model of the machine-furnace coordination control system,and the singular value decomposition method was used to obtain the multicellular LPV model.In the original robust H_∞/H₂ variable gain control,a symmetric matrix X_∞/X₂ is required,and multiple LMIs are also required,which greatly increases the conservativeness of the system.In this paper,the relaxation variable is introduced into the original robust H_∞/H₂ variable gain control,which reduces the conservatism of the controller des ign.In the case of model mismatch,a simulation comparison with robust control shows that the improved robust variable gain controller has a good control effect.