Research On The Estimation Of Mass For Truck And Road Slope

By the application for trucks continuing to increase,promoting the study of its automation and causing the extensive research on its control.Whether the development of control system,the implementation or improve ment of the control strategy are all carried out according to the input parameters.According to the characteristics that the variation range of mass for truck is great and the work road for truck is various,mass and road slope are two important parameters in the study of its control on automatic transmission,security system and so on.Accurately and real-time getting the two important input parameters is the premise and the key to realize truck control research,using the method of parameter estimation can not only accurately and real-time getting them,and also its estimation algorithm can be directly used in the subsequent control research.In view of this,this paper will research on the estimation of mass for truck and road slope,laying a foundation for the research of truck performance control.To mass for truck,different from the past research that simply based on the vehicle longitudinal dynamics model,this paper put forward the method of installing a low cost acceleration sensor to decoupling mass and road slope,and using the forgetting factor recursive least square method to estimate it,integrated the two points,this paper presents the vehicle mass least square estimation model that based on the combination of vehicle longitudinal dynamics model and acceleration sensor signal.To road slope,on the basis of vehicle mass is known,based on vehicle longitudinal dynamics,respectively using the forgetting factor recursive least square method and kalman filtering method to estimate it;based on acceleration sensor signal,first,the forgetting factor recursive least square method is adopted to estimate the road slope change rate,and then input the estimation as a factor using the kalman filtering method to estimate the road slope;thus,in this paper,a total of three kinds of road slope estimation models are proposed: the road slope least square estimation model based on vehicle longitudinal dynamics,the road slope kalman filtering model based on vehicle longitudinal dynamics,the road slope estimation model based on acceleration sensor.Then set up the TruckSim/Simulink co-simulation platform,through the simulation tests under different conditions to verify the estimation models that put forwarded in this paper,except the estimate effect of the road slope least square estimation model based o n vehicle longitudinal dynamics in time-varying slope road is not good,the rest of the models are all obtained accurate real-time estimation results.Through the simulation results of the three kinds of road slope estimation model found that: compared with the least square method,the models that used kalman filtering method are more suitable for estimate the real-time slope;compared to vehicle longitudinal dynamics,the models that based on acceleration sensor signal are more resistant to the effects of shift operation;compared to the situation that without considering road slope change rate,the road slope estimation model based on acceleration sensor signal with considering road slope change rate in this paper has better real-time performance and accuracy.Based on the verification results,in view of the problems of road slope least square estimation model,this paper put forward the improvement and did the verification,the verification results s how that the improved model has superiority on the estimation of real-time slope.Finally,this paper simple discussed the prophase work of the real vehicle experiment data that used for the model verification.