Research And Application Of Power Electronic Learning Guidance Artificial Intelligence System

Switching power supplies have important application values in various fields such as scientific research,power construction,and national defense facilities.Therefore,it is particularly important to design and manufacture high-performance,low-volume switching power supplies.Magnetic components are the core components of switching power supplies.The saturation characteristics of magnetic core components are common to magnetic components.In practical applications,the main factor for components to heat up or even burn the circuit is the oversaturation of magnetic components.Magnetic components are mainly composed of windings and magnetic cores,which have important functions such as energy storage,energy conversion and isolation of components.Therefore,in the design process of switching power supply,the magnetic components are required to achieve high efficiency,low loss,low noise and other effects,and while satisfying these conditions,the volume of the magnetic components is as small as possible.However,due to the limited cognitive ability of switching power supply designers,magnetic component design lacks knowledge of their own capabilities and development.At the same time,it is difficult to form a knowledge network for the overall design of switching power supplies,and it is difficult to establish a knowledge network of specific circuit faults and their influencing factors.A network of relationships that initiate circuit failures and related factors affecting failures.Therefore,this article designs a power electronics artificial intelligence guidance system,aimed at specific complex engineering problems in power supply design,with the goal of improving the recognition of cognitive diagnosis results,indepth study of the unique physical characteristics of the switching power supply magnetic components,and carrying out consultations for switching power supply designers Intelligent evaluation,combined with knowledge map technology for personalized guidance,thereby improving the design quality and efficiency of the switching power supply.First,based on the principle of BUCK and BOOST circuit,analyze the performance indicators and calculation methods in the circuit,introduce the type of inductance core material,use WEBENCH software to analyze the inductance saturation characteristics of different core materials,and the influence of different inductance saturation characteristics on the current limiting function,And serve as the research basis for the construction of cognitive diagnosis models and the push of personalized experimental projects.Secondly,according to the specific cognitive skills of different inductance core material types and core saturation characteristics,design the experimental project task model that reflects the characteristics of cognitive skills,and build a cognitive diagnosis model based on Bayesian networks.We introduce item response theory(IRT)to analyze the relationship between the variables of the proficiency model and the evidence model in the cognitive diagnosis model,and calculate the conditional probability between the variables.According to the designer’s test results,and using the Markov chain Monte Carlo(MCMC)estimation method,the WINBUGS software calculates the posterior probability distribution of the variables in the cognitive diagnosis model,and diagnoses the designer’s design of switching power supply magnetic components.Cognitive ability,distinguish the different proficiency of designers in the design ability of magnetic components,and then provide personalized guidance.And at the same time,the expectations,standard deviations,MC errors and 95% HPDI of the parameters in the variables are obtained.Analyzing the convergence of parameters based on simulation results to improve the correctness,effectiveness and reliability of the cognitive diagnosis model.Finally,based on the knowledge graph technology and Neo4 j software simulation,the switching power supply knowledge graph is constructed to help designers quickly and accurately find the required experimental tasks,as well as the knowledge points contained in the experimental tasks,and the ability points required to master.At the same time,a personalized cognitive skills map was constructed according to different designers.Then introduce new boost research tasks and gradually increase the nodes in the knowledge graph.Finally,adopt the Pearson correlation method to develop a switching power supply design intelligent guidance system,and according to the designer’s shortcomings of knowledge or ability,personalized push experimental projects to achieve personalized intelligent guidance.