Research On Fault Simulation And Diagnosis Method Of Inland Ship Hybrid Power System

Using hybrid power is an effective method to improve energy efficiency and reduce carbon emissions.In recent years,hybrid power systems have been vigorously promoted in the inland waterway ship industry.Inland waterway ships equipped with diesel-electric hybrid power systems have 4 modes of PTI,PTO,PTH and main engines.The propulsion method can enable inland ships to meet different power requirements and improve energy efficiency in complex and changeable inland waterways.However,compared with traditional propulsion methods,the composition of the hybrid power system is complex,and the operation and maintenance are relatively difficult.In order to realize the stable operation of the hybrid power system and ensure the safety of inland ships,the simulation of the hybrid power system of the inland ship is systematically studied in this paper.On this basis,the research on fault diagnosis of hybrid power system is carried out.The main research contents are as follows:(1)In order to accurately simulate the failure of the diesel-electric hybrid system of the inland ship,a 330 k W diesel engine is selected as the main engine,and a 180 k W three-phase asynchronous motor is used as the propulsion motor.According to the hybrid power system matching mode of the 7 500 t demonstration ship,the power matching of the hybrid power system simulation model is carried out,and the speed ratio is set according to the output speed of the system.The zero-dimensional model of the diesel engine is established,the gearbox model is constructed according to the "two-in and one-out" gearbox and its clutch and speed change functions,and the propulsion motor model based on the direct torque control method is established,and finally integrated into a diesel-electric hybrid system model.The functionality of the hybrid system model is verified in 4 operating modes.(2)In order to ensure the calculation accuracy of the diesel engine model under various steady-state operating conditions,the parameter tuning method based on reinforcement learning is studied,and the Proximal Policy Optimization algorithm is used to construct an agent,which realizes the diesel engine model under various steady-state operating conditions.The parameter tuning work overcomes the problem that the error of the least squares method increases with the increase of the tuning parameters,and the genetic algorithm increases the time-consuming problem with the increase of the tuning parameters.The parameter fitting was completed,and the diesel engine non-predictive model was converted into a semi-predictive model.(3)Based on the established hybrid power system model,the equipment faults simulation research is carried out.The diesel engine model is used to simulate the fuel quantity reduction and timing faults,and the cylinder pressure,power,and instantaneous speed are selected as evaluation indicators,and the single-cylinder performance and the performance of the whole engine under faults are analyzed respectively.Using the propulsion motor model,a single-phase open-circuit fault simulation was carried out,and the variation rules of the rotational speed,torque and stator three-phase current under the fault were studied.(4)In view of the variety of inland waterways and the dense traffic,the power performance stability of the hybrid system is highly required,and the power demand is large under the condition of countercurrent or high speed.Based on the established hybrid power system model,the influence of diesel engine and electric motor faults on system performance was studied under the four operating modes of the hybrid power system,PTI,PTO,PTH and main engine.The speed,torque and fluctuation rate of the hybrid system are used as evaluation indicators.Diesel engine fuel loss,timing abnormality,single-cylinder fuel cut,double-cylinder fuel cut and motor single-phase cut-off have a great impact on power.The research on the influence of faults is carried out,and the influence of different faults on the operation mode is analyzed.(5)Research on fault diagnosis method for diesel engine in hybrid system.The instantaneous speed is selected as the main analysis signal,and the influence rules of various faults on the time domain and frequency domain signals of the instantaneous speed is analyzed,and the experimental data of the 4120 SG diesel engine are compared and verified.And the peak-to-peak value,the maximum peak value,the instantaneous speed difference corresponding to 0.5 and 1 harmonic are used as fault features.And the fault diagnosis method based on decision tree,random forest and Xgboost is studied to diagnose the single-cylinder fuel cut-off,double-cylinder fuel cut-off and abnormal timing faults of diesel engine.And compared with the accuracy of multilayer perceptron and support vector machine methods.