Research And Application Of Construction Method Of Vehicle Driving Condition In Complex Environment

Vehicle driving cycle construction plays an important role in the field of urban vehicle fuel consumption calibration and intelligent transportation.With the rapid development of Chinese economy,vehicle technology is developing in the direction of intelligence and new energy.With the rapid transformation of vehicle technology and the constant change of urban road characteristics,the number of cars is increasing and the traffic is becoming more and more complicated.Vehicle driving cycle is the basic of calibrating urban vehicle fuel consumption,which needs to meet the requirements of accuracy,granularity and real environmental state.The algorithm of vehicle driving cycle construction usually consists of two parts: short-trip clustering algorithm and short-travel selection algorithm.In the existing driving cycle construction methods,it is often unable to effectively construct driving cycle curve meeting the accuracy conditions,and there are often problems such as insufficient precision,insufficient granularity,unable to reflect the real road state of the city.Therefore,this thesis investigates the short-trip clustering algorithm and short-travel selection algorithm,and proposes a vehicle short-trip clustering model and a vehicle driving cycle construction model.The above model can be used to construct the driving cycle curve that meets the conditions.The main work of this thesis is as follows:1.In this thesis,clustering model for vehicle short trip(CMVS)in complex urban roads is proposed.The short stroke was divided,the special diagnostic parameters of the short stroke were extracted,the principal component analysis was carried out,and proposes the Self-Organizing feature Map of Vehicle Short trip(SOMVS)for clustering short trips,and finally the short stroke clustering result was obtained.It provides data sources for the subsequent driving cycle construction model and effectively improves the clustering accuracy.2.A fine-grained driving cycle construction model(FDCCM)is proposed.Short trip sets classification(STSC)is proposed to classify the driving cycles in a finer-grained manner.The short trips are classified and the short trips are divided into trip regions.The FDCCM model is trained,and the state sequence generate(SSG)and driving cycle generate(DCG)algorithms are proposed to construct driving cycles,which improves the accuracy of the constructed driving cycles.3.In this thesis,combining the above CMVS and FDCCM,this thesis designs and implements the construction of vehicle driving cycle system in a complex environment.Five aspects of the system are described in detail: requirements analysis,system architecture design,system development environment,design and implementation of system modules,and system testing.The system module design and implementation are described in detail,the interrelationship of all aspects,the underlying implementation is given,and the system is presented.The system was tested and the results proved the feasibility and rationality of the system.