The Study On Vibration And Noise Analysis And Control Of Range-extender

As a new type of electric vehicle, extended-range electric vehicle overcomes theshortcoming of pure electric vehicle’s short trip range and improves the fuel economy byadding a range-extender. But today, the control strategy of most range-extenders isdesigned for the fuel economy as the single optimization goal. This control strategyreduces range-extender’s fuel consumption, but it ignores its NVH performance at thesame time. When the range-extender runs, its noise and vibration may be very strong thatreduce the vehicle’s comfort performance. Aim of this paper is for the study ofrange-extender’s noise, vibration and optimization of its control strategy. Based on therange-extender’s fuel economy and NVH performance, a new optimization of controlstrategy has been put forward by a multi-objective optimizing method to reducing its noiseand vibration and guaranteeing the fuel economy.First, the NVH performance of range-extender is described from the extended-rangeelectric vehicle’s operation mode, the structure of range-extender and its working principle.Traditional engine and generator’s noise and vibration control method has beensummarized. Three directions of range-extender’s noise and vibration control techonologyhave also been put forward. The existing range-extender control strategy has beendescribed that its engine’s operation points are always on the low fuel consumption rate,high rotate speed and high torque area. This control strategy easily leads to a poor NVHperformance of range-extender. A new optimizing method has been put forward that therange-extender’s NVH performance can be improved from developing its control strategy.This strategy is to find the engine’s new working area which has relatively low fuelconsumption, low noise and vibration and in this area the NVH performance can beimproved by increasing a little fuel comsumption.Simulation software Matlab/Simulink was adopted for modeling simulation accordingto the range-extender structure and control requirements. Engine model, generator modeland controller model have been established. Range-extender model was generated throughthe connection of all subsystems, which was realized by the coaxially connected physicalstructure between the engine and the generator, torque and the kinetic energy balanceequation. Through model simulation, range-extender’s working principle and controlstrategy were described and validated. The new scheme that finding an optimizedrange-extender control strategy to improve its NVH performance has also been putforward.Refer to some CNS standards and enterprise standards, a bench test for measuringrange-extender’s fuel consumption rate, noise and vibration has been designed and the testprogram has been wroten. Based on the test program, the range-extender body’sacceleration, sound-pressure of its envelope surface noise and the engine’s fuelconsumption rate have been measured. Wavelet transform filteration and some othermethods have been used to process the test data. Finally the data of range-extender body’svibration severity, envelope surface noise sound-pressure level and engine’s fuelconsumption rate have been got. BP network has been used to eatablish the nonlinearmapping relation between the engine’s fuel consumption rate, body’s vibration severity,sound-pressure of its envelope surface noise with range-extender’s rotate speed and torque.Three maps have been plotted as the reference of range-extender’s vibration and noise.Three bivariate polynomial models have been used to establish the correlationfunctions between fuel consumption rate, range-extender’s body vibration intensity andnoise sound-pressure with range-extender’s rotate speed and torque. The binary nonlinearregression algorithm has also been used to establish the correlation functions. All of thefunctions are regarded as multi-objective optimization’s target functions. Range-extender’srotate speed and effective torque are regarded as the variables and constraints. The linearweighting genetic algorithm and NSGA-II(Non-dominated Sorting Genetic Algorithm-II)have been used to solve this multi-objective optimization problem. Based on the linearweighting genetic algorithm’s optimized results, three excellent operation points have beenchoosed and they represent respectively economical efficiency, stationarity and halfmeasure. Based on the optimized results of NSGA-II, the Pareto best face has been found.In range-extender’s actual operation, the operation points can be selected based on theenvironment, required power from VMS(Vehicle management system) and other conditins.