Performance Optimization And Diagnostic Analysis Of Hydrogen Engines Based On Fusion Algorithms

In the 21 st century,with the serious shortage of petroleum resources and the worsening of environmental pollution,finding "clean alternative fuels" has become the common goal of governments and people from all walks of life.Hydrogen energy is regarded as the most promising alternative fuel for its clean,efficient and renewable properties,more importantly,the car engine we are using can burn hydrogen with a little modification.However,due to the characteristics of low density,fast combustion speed and wide ignition limit,hydrogen-fueled engines are more prone to abnormal combustion phenomena such as pre-ignition,backfire and knocking,which will affect the economy and power performance of the engine.Therefore,the performance optimization and abnormal combustion analysis of hydrogen engines are a hot topic.This paper makes full use of the existing experimental conditions to conduct a more comprehensive analysis and research on hydrogen engines.In this paper,the new information fusion method is applied to the research field of hydrogen engine.Genetic algorithm,neural network fusion algorithm,artificial bee colony algorithm and FOA-GRNN fusion algorithm are used to study the optimal control and fault diagnosis of hydrogen fuel engine.Firstly,based on each algorithm,the fusion optimization model is constructed on MATLAB platform,then the experimental data are analyzed,and some reasonable suggestions are put forward to improve the optimization results.To improve the economic and dynamic performance of the hydrogen engine,the genetic-neural network fusion algorithm is selected to optimize,the genetic algorithm is used to improve the deficiency of neural network which is prone to fall into local extremum.By optimizing the operation parameters of excess air coefficient and ignition advance angle,the economic performance and dynamic performance of hydrogen engine can be effectively improved;Aiming at the pollution of exhaust gas of hydrogen engine,this paper firstly analyzes the harmful components of exhaust gas,mainly including HC,CO,NOx,then,artificial bee colony algorithm is used to predict the emission trend of various harmful components,and the reasonable measures and methods to reduce the emission of harmful substances are found out;For the problem that hydrogen engine is prone to abnormal combustion such as pre-ignition and backfire,the variation of cylinder pressure and intake pipe pressure of hydrogen engine is studied in this paper,and the FOA-GRNN fusionalgorithm is used to analyze and predict the abnormal combustion phenomenon.The experimental results show that this method can make a relatively reliable diagnosis for abnormal combustion of the hydrogen engine.