Research On Intelligent Management Method Of Vehicle Energy Flow Based On Road Driving Conditions

With the development of new round of scientific and technological innovations,such as artificial intelligence,intelligent manufacturing,big data,and blockchain technology,there are huge changes in the automotive industry.The new technology era of electrification,intelligence,networking,and sharing have already been arrived.In addition,China 6 emissions standard has been implemented the standard of the fuel consumption and emissions have become stringent,thus,the emissions inspection site has been expanded from the laboratory environment to outdoor roads environment,which has made vehicle operating conditions more widely distributed,there are more emissions optimization points and more factors that need to be comprehensively considered.Vehicle energy flow management technology is one of the key technologies for the development of traditional vehicles/new energy vehicles.It is also one of the most active,broadest and deepest research directions in the proc ess of automobile development.Its research scope ranges from static conditions of specific working conditions to the parameter identification,prediction of the real driving cycle and the on-line adjustment and dynamic optimization of control strategy under the random road conditions.What is more,the research focus has evolved from single-objective fuel economy optimization under steady-state conditions to the multi-objective optimization control of fuel economy and emissions under transient conditions.Therefore,this paper takes the vehicle's transient performance online detection method coupling of advanced sensing technology and numerical simulati on,research on the quantitative relationship between engine con trol and performance parameters and multi-level energy management technology as the basis.It is important and necessary to solve the scientific question,such as the o n-line detection method of transient performance,operation and control parameters and the development of vehicle-energy-thermo-electric vehicle energy management strategies based on typical working conditions.Thus,in this paper,the format of experiment-data-model of vehicle energy management technology is developed,the actual and typical road driving cycle is presented based on the combination of statistics methods and genetic evolution algorithms,the on-line optimal algorithm based on Lyapunov optimization method is stated and verified in the vehicle performance simulation platform.The main conclusions of this paper are presented as follows:(1)The simplied RGF model is developed and then the on-line fuel and torque model are developed,besides,a test method for the actua l road state of vehicle is presented.The torque and fuel consumption model based on the two-steady-state sensor method has good test accuracy,the error is within 5% and the method is simple.It can be used to be implemented in online test application based on real driving roads and play a vital role in optimizing the fuel consumption under transient conditions.(2)A road condition recombination scheme based on the combination of statistical methods and genetic evolution algorithms is proposed,and a benc h test of the constructed typical cycles and original data is carried out.The results show that the typical road driving cycle is closer to real road conditions,which makes the performance of vehicle development process is closer to the this of real roads.(3)The multi-physics integrated modeling and optimization of vehicle energy flow is carried out,vehicle/engine performance,combustion characteristics and energy distribution under RDE conditions are analyzed,and the influencing factors of energy terms is revealed.The results show that(a)the effective work output ratio of NEDC and WLTC is greater than this of RDE,the largest combustion losses of driving cycle are WLTC,RDE and NEDC,respectively,which indicates that the more drastic changes in vehicle speed,the greater the combustion loss in the cylinder;(b)The exhaust enthalpy of NEDC and WLTC is higher than this of RDE,but those proportions are higher than 20%,meantime,the proportion of coolant heat transfer energy in three driving cycles are higher than 18%;(c)The transmission shifting strategy is optimized and the improvement pf fuel economy of the vehicle under NEDC and WLTC has reached up to 3.0% and 2.78%,respectively.(4)A transient fuel consumption model based on engine speed and torque is constructed and applied to the hybrid power system,a hybrid performance analysis and optimization platform is built,an online energy management algorithm based on Lyapunov optimization is proposed and is compared and analyzed along with DP and A-ECMS.There is no difference in performance between the standardized forward DP algorithm based on the combination of markov chain,working mode and the basic DP algorithm,but the required running time of former algorithm is shorter.In addition,compared with the A-ECMS algorithm,the fuel comsumption is reduced by 13% based on Lyapunov's algorithm.Therefore,the development of multi-mode algorithm and application of genetic evolution algorithm are expected to build a unified and standardized algorithm platform,with the help of multi-disciplinary energy management models.Besides,it can be used to solve many scientific problems,such as the markov nature of vehicle test cycles,control/constraint variables,uncertainty of objective function and l arge-scale,diversified evolutionary algorithms.(5)By adjusting the compression ratio and VV T variables in the traditional Otto cycle engine,on this basis,the introduction of a high-pressure EGR system can achieve the dual effect of simultaneously reducing engine fuel consumption and NOx emissions.In addition,the cumulative fuel consumption of series hybrid vehicle with optimized engine(Atkinson cycle engine)is reduced by 4.58%,4.98%,4.31% in NEDC,WLTC,and RDE conditions,respectively,and the cumul ative NOx volume fraction is reduced by 73.34%,71.82% and 58.11%,respectively.What is more,in terms of series-parallel hybrid vehicle,regardless of whether they are in CD or CS process,the decrease extent in cumulative fuel consumption is higher than that of the series model,but the decrease extent in cumulative NOx is lower than that of the series model.The research in this paper promotes the development of interdisciplinary subjects such as thermodynamics,testing methods,and machine learning,which realizes the application of artificial intelligence in vehicles field.At the same time,due to the application of transient online technology,the basic disciplines of traditional combustion,thermodynamics,and heat transfer can be promoted from steady state to transient conditions,which would further improve the basic theories of traditional disciplines and is expected to multi-objective on-line performance control under transient conditions,and then improve the vehicle performance in the actual operating environment,and finally achieve the goal of energy conservation and emissions reduction,which have great engineering application prospects.