Dynamic axel load estimation for an electrified vehicle : Normalkraftestimering på drivaxeln

Abstract: The brake system is critical for ensuring safe driving and has been the focus of development for many years. Pneumatic braking technology is commonly used in heavy vehicles,but it results in energy wastage and high service costs. Some countries mandate auxiliarybraking systems to assist in stopping the vehicle in addition to service brakes. One suchsystem is the regenerative braking system, which captures kinetic energy from braking andconverts it into electrical energy. Retarders are another commonly used auxiliary brakingsystem. These systems are essential due to heavy vehicle weight, which can weaken servicebrake performance.This thesis focuses on estimating net forces on the truck’s driven axle to understandhow auxiliary braking systems and vehicle traction affect the normal force on the drivenwheel axle. The expected result can assist in maintaining the slip ratio and increasing thelife span and performance of brakes.Scania uses a function to estimate forces on the driven axle and drive wheel slip. Theyneed to determine the normal force on the axle to improve performance of auxiliary brakingsystems, but tests showed that for the same specified slip ratio, the auxiliary braking forcerequired was smaller than that in an acceleration state. Scania believe that dynamic axleload transfer may be the cause, so a 6x2 electrified truck will be investigated in this thesis.The obtained results show the driven wheel axle’s behavior during different dynamicalscenarios.This research aimed to develop a model that can accurately simulate a truck’s movement and estimate the ground reaction force in response to variations in the scenario of thecontrol signal. By studying the quarter car model, bounce-pitch and half car model, theresearchers were able to obtain a model represented by 5 ODEs, which predicts the wheelaxle normal force. To verify the model, data from the CAN bus and measurements using ascale were collected and compared with the model’s output. The Mean Squared Error canbe used to evaluate and compare the model’s performance, and the results showed that themodel provides a reasonable estimate of the normal force on all axles. The study also analyzed the factors that contributed to the errors in the results. The behavior of the normalforce for each wheel axle during acceleration and braking was illustrated, explaining howthe normal force distribution becomes mirrored compared to the acceleration state duringbraking. The study’s discussions enhance the validity of the observed behavior and thereliability of the results.