Research On Fault Diagnosis Of Ship Electric Propulsion System Based On Frequency Rotation Cyclic Mode Decomposition And SAGA-FC

Ship electric propulsion system is not only the main power consumption object of ships,but also the only propulsion power source of ships.It is known as "marine lifeline".The marine electric propulsion system is affected by complex state of sea conditions and state of working conditions.In addition,it is immersed in water for a long time and works continuously.The frequency of damage to the propulsion device is high.On the other hand,ultra high power inverters and propulsion motors are widely used in marine electric propulsion systems.The capacity of these equipment can often be compared with that of single capacity marine generator set,resulting in the complexity of "mechanical,electrical and magnetic" coupling,which has a great impact on the quality of power grid and makes the electrical fault of propulsion system more complex.Therefore,this paper takes the ship electric propulsion system as the object,introduces its basic structure and fault characteristics,carries out the mathematical modeling and simulation research of key equipment and systems such as propeller,propulsion motor and power supply line,and completes the fault analysis,fault feature extraction and fault diagnosis of the propulsion system.The specific contents include:Aiming at the influence of ship electric propulsion system fault on power grid,the quasi synchronous grid connection simulation model of ship power station and a set of mathematical models of ship electric propulsion system including inverter,PMSM,propeller and hull are established.In this model,torque control is added to the two closed loops of magnetic field oriented vector control speed and current to form a three closed loop control.Finally,the anti-interference characteristics of the propulsion system under sudden constant torque load and random fluctuating torque load are analyzed,and the dynamic and static performance of the following characteristics of the propulsion system under speed switching is analyzed.The results show that the model can well simulate the transient and steady-state characteristics of ship electric propulsion system.Aiming at the problem that the permanent magnet synchronous motor is prone to inter-turn short-circuit fault due to large power consumption and load,a simulation model of PMSM is proposed.By introducing two parameters of short-circuit resistance and fault degree factor,the inter-turn short circuit model of PMSM is derived.The inductance and the loop current of inter-turn short circuit winding and of PMSM are calculated.The model can simulate the performance of on-line continuous normal motor and the fault of winding inter-turn short circuit separately only by adjusting the parameter of fault degree,so as to prepare data for fault diagnosis.Aiming at the mode aliasing problem existing in the traditional empirical mode decomposition method in fault feature extraction,a frequency rotation cyclic mode decomposition algorithm is proposed,and the mathematical signals including no aliasing,partial aliasing and full aliasing are tested.This method can not only effectively separate the aliasing modes,but also has fast separation speed and good effect.Aiming at the classification error caused by local search in traditional fuzzy c-means clustering,a fault diagnosis method of ship electric propulsion system based on simulated annealing genetic algorithm fuzzy c-means clustering(saga-fcm)is proposed.Using the complementary advantages of genetic algorithm and simulated annealing algorithm,the traditional fuzzy c-means clustering problem is improved and the success rate of clustering is improved.This method is used to classify four kinds of signals:normal current,phase to phase short circuit circuit,single-phase grounding current and three-phase grounding current.Finally,a set of experimental device for fault analysis and diagnosis of marine electric propulsion system is designed and developed under the existing marine electric power system and equipment in the laboratory.Based on this,the operation state monitoring and fault diagnosis software is developed by using Lab VIEW,VB,MATLAB and other tools.