Condition Monitoring And Fault Diagnosis Technology Of Generator Set On Offshore Platform

The fundamental purpose of power system is to provide customers with electricity that meets the needs of manufacture and life economically and reliably.However,due to the limitation of marine environment and cost considerations,the connecting lines between the power systems of offshore oil and gas exploration platforms are weak,while the main load capacity on the platforms is large,the start-up impact is large,the gas turbine generator set capacity is relatively small,and easy to trip,which means the system has the characteristics of large generator and small network,and its reliability is generally far lower than that of the land power grid,which affects the oil and gas production.The main purpose of this paper is to improve the reliability of gas turbine generator set used in offshore power grid.In order to realize state monitoring and fault diagnosis of generator set,the dynamic model of grid-connected analysis model and the multiphysics coupling finite element model are established,basing on which the common techniques of reducing equipment failure rate,shortening fault repair time and improving equipment reliability are proposed.The main outcomes of this paper are as follows:1.A dynamic model of gas turbine generator set for grid-connected analysis is established.According to the actual electromechanical parameters,gas speed control strategy and excitation control method,the generator model,speed control model of gas turbine and excitation control model are constructed.By simulating the steady state and dynamic states of the set in the healthy state and fault states,the linkage relationship between the different physical quantities is obtained.At the same time,the model can be used to build the offshore power grid model to improve the reliability of the power system from the operational and control levels.2.Based on the multiphysics coupling finite element model,fault characteristic analysis is carried out in different fault states.First of all,the finite element model of the two-dimensional solution domain of a gas turbine generator is established.Then,the multiphysics analysis is carried out on different fault state,which can be regarded as the basis for the fault diagnosis of generator.3.The generator fault diagnosis method based on machine learning classifier algorithm is studied.Based on the multi physical field analysis of various fault states,multi types of fault information can be obtained.After big data training,the classifier model can automatically classify the characteristic signal array of each state,realize the intelligent fault diagnosis,shorten the downtime,and improve the reliability of the equipment.4.Based on the monitoring status fault tree,a method for making the operation and inspection strategy of gas turbine generator set is proposed.Firstly,the fault causes of abnormal state are sorted out comprehensively,and the fault tree is established and simplified.Then,qualitative and quantitative analysis of fault tree is carried out.Based on the fault tree simulation,the top event reliability curve of the generator set is obtained,and then the planned outage maintenance cycle is arranged according to the reliability requirements to reduce the failure rate;on the other hand,when the abnormal state occurs,each component of the set can be checked according to the importance ranking results to shorten the troubleshooting time.