ZH Research On Risk Management Of Mechatronics Projects In Aviation Equipment Manufacturing Enterprise

From the first industrial revolution to today’s industrial 4.0 era,every stage of human social progress is accompanied by the industrial crystallization that can be realized at the highest industrial level at that time.From the industrial point of view,it represents the limit of human wisdom in an era.From steam engines to internal combustion engines,from T-cars to four-door cars,from propeller planes to current jet planes,aviation electromechanical integration products are undoubtedly the crown in the industrial field of our time.Corresponding to the many honors of aviation mechatronics products,it is natural to have high standards and strict requirements for related projects of such products,which brings a lot of project risks in the management process of design and manufacturing of related projects,and every response to the risks challenges the mental limit of technicians and their managers.However,due to the limited human cognitive ability,the delay of information itself,the enterprise’s own ability and other reasons,the enterprise in this paper is often in a very passive situation in the process of promoting the work of the aviation electromechanical integration product project.It is impossible to coordinate the work from the overall situation of the project,and it is only possible to encounter risks to deal with risks,and it is too late to think about the causes of risks and optimization measures.The next risk occurs again,and the project risk is so vicious circle in the daily work of the enterprise in this paper.In this paper,from the perspective of enterprise project managers,the data of human resources,technical difficulty,development cycle,risk occurrence and response related to the project are collected to form a small data set for the analysis of big data machine learning technology,and the data set is machine-learned and analyzed by using Bayesian model with strong generalization ability,supervised lightweight Gradient Boosted Decision Tree(GBDT)and multi-layer neural network model under the framework of deep learning.Produce a certain hit rate of pre-judgment results and quantitative ranking of feature importance,so that managers can pre-judge project risks with data support based on the current project configuration data,and make optimal plans in time to help managers make rational risk decisions.Using multiple linear regression model to analyze and form multiple linear polynomials to fit the risk response problem and quantify the risk response cost.Use matrix mathematics software to visualize the project data,look down at the risk occurrence and response in all stages of the project development cycle,deploy relevant work in a targeted and focused way,and refine the risk management of enterprise projects.