Variable Parameter Optimization Control Strategy Of Control System Based On Information Fusion Technology

With the proposal of the energy strategic goal of "carbon peak and carbon neutrality",the scale of my country’s renewable energy consumption will be further expanded,and higher requirements will be put forward for the flexible operation of thermal power units.Thermal power units need to be more frequent and greater Participating in deep peak shaving,its safe and stable operation is facing huge challenges.As a result,researchers need to further optimize the thermal power unit coordinated control system and its multiple subsystems to meet the needs of maintaining the safe and stable operation of the unit while participating in deep peak shaving.In response to the above problems,this article has done the following work:(1)By participating in the on-site commissioning of a subcritical unit,collecting the operating data of its high load,medium load,low load and load-up and load-down conditions,and analyze the data fluctuations to obtain the operating characteristics of the sub-critical unit,and summarize the furnace The harm caused by the fluctuation of fuel quantity on the side and the fluctuation of the steam regulating valve on the engine side.(2)In view of the fact that the steam-water subsystem of the unit is prone to "false water level" when the unit participates in deep peak shaving,the actual measured water level of the unit is regarded as the superposition of the weight water level and the false water level by analyzing the factors that cause the drum water level change,and the weight water level can be accurately quantified.Describe,establish the drum water level model to calculate the weight water level,remove the weight water level from the actual measured water level,and get the false water level.Using wavelet multi-scale correlation analysis,selecting a suitable wavelet base,and decomposing the obtained false water level signal,feedwater flow signal,steam flow signal,fuel quantity signal and drum pressure signal,etc.into different frequency bands,and observe them by calculating their correlation coefficients.Volatility,determine its relevance.Experimentally verified,this method effectively eliminates the influence of feed water flow and steam flow on the false water level,and enhances the correlation between fuel volume and drum pressure and false water level signals.There is a strong correlation between the bag pressure and the false water level signal,which can assist in judging the generation of false water level and its fluctuation range.(3)Aiming at the problem that the control object of the front pressure of the closed-loop unit on the unit side is a non-self-balancing object,which is difficult to control,based on the simplified nonlinear dynamic model of a typical subcritical unit,the transfer function of the object is derived and built in SIMULINK.The subcritical unit model used starts with a simple first-order non-self-balancing object control system to analyze its control characteristics.In the built model,the furnace side controller is simulated using PD control and PID control to analyze its applicable operating conditions.Drawing lessons from the idea of online learning and automatic adjustment,using the self-learning ability of the fuzzy neural network,an optimized controller is designed to replace the original controller on the furnace side.Experiments have verified that this method can solve the problem of artificially determining when the unit without self-balanced object is subjected to load disturbance to perform integral switching,and the simulation results meet the requirements of unit operation.