Design Method Of Driving Cycle With Road Grade Based On Markov Chain Model

Driving cycles are used to describe the characteristics of a vehicle in a particular traffic environments.Driving cycles will provide the basis for the inspection of the emissions level or fuel consumption for a certain vehicle in a particular area.Heavy duty truck,new energy vehicles and hybrid electric vehicles are significantly impacted with road grad.The road grade in traditional driving cycles are set to zeros,which can't meet the requirements of energy-saving and environmental friendly automotive technologies in auto industry.This paper is supported by projects with China FAW Group Corporation R&D Center,and studied a driving cycle design method with road grade based on Markov Chain Model.The main contents in this paper are as flows:Firstly,the Markov property of driving cycles in the status space which was defined by velocity,acceleration and road grade(referred to VAG status space)were analyzed by data statistical analysis.By comparing the phase space scatter diagram and phase trajectory of driving cycles between VA(velocity and acceleration)status space and VAG status space,the similarity of shapes of driving cycles in different status space was displayed.The encode rules for VAG status space was formulated and the experiment data were encoded to state sequences.And then subsequences were constructed with different time interval.The correlation coefficient of subsequences with different time interval was calculated as the criterions of driving cycles.The Markov property of driving cycles were demonstrated with small scale time interval.Secondly,a iterative algorithm for the calculation of road grade was designed based on GPS signal due to the lack of title sensor,or acceleration sensor,of experimental equipment in large scale experiments.According to the data of a small scale experimental(which a special truck with an acceleration sensor was used for calculating accurate the road grad),road grade versus time and road grade versus distance with the same route were compared.And combined with the requirements of the design of vertical curve in road alignment design,road grade versus distance of a certain route was 'unchanged'.So,road grade versus distance was chosen as filter object.According to the feature of not real-time,a forward and backward zero phase digital filter was selected to filter the preliminary calculation of road grade.By means of optimization theory,the cutoff frequency of the filter was determined through the filter effect evaluation that whether or not the changes of road grade versus distance meet the road alignment design.4 was chose as the filter order by Single variable test.Finally,the iterative algorithm was given with engineering constraints.Thirdly,a driving cycle was designed by the new method and verified by the simulation.After the target database was selected,the design can be divided to 3 processes:the induction of state transition matrix,the deduction of new sequence and the statistical indicators check.In the induction of state transition matrix,state transition matrix was stored and computed as sparse matrix so as to reduce the storage space and improve calculation efficiency.The designed driving cycle was verified with dynamic simulation and fuel economy simulation.Then the driving cycle design method with road grade based on Markov Chain Model was proved completely.