Backlash Estimation for Heavy Truck Steering Systems

Sammanfattning

Mechanical backlash in the steering gear of heavy-duty trucks introduces nonlinearitiesand dead zones that degrade steering precision and complicate the control algorithmsrequired for Advanced Driver Assistance Systems (ADAS) and autonomousoperation. As components wear over their operational lifespan, backlash increases,motivating the need for online estimation to enable adaptive control and predictivemaintenance. While dedicated load-side angle sensors (located on the side ofthe steering gear subject to the road load) could directly quantify this wear, theircost and reliability in harsh environments make a software-based solution preferable.This thesis proposes a Switched Kalman Filter (SKF) for estimating both the sizeand position of the steering gear backlash using only sensors already existing on thetruck. The method adapts the framework of Lagerberg and Egardt to the topologyof a heavy-duty truck steering system, where no physical sensor exists on the loadside of the steering gear. A key contribution is the reformulation of the backlash offsetstates into a center position and a half-size state, which decouples slowly varyingmeasurement biases from the physically meaningful backlash size estimate and resolvesa systematic directional error present in the original formulation. The absentload-side measurement is addressed by estimating the pitman arm angle from yawrate using a steady-state bicycle model with lead compensation. The observabilityand practical stability of the switched system are analyzed theoretically.The estimator is validated against highway driving data from a heavy-duty truckwith known backlash levels. Using a physical pitman arm sensor, the worst-caseestimation error is 11% with an absolute error of 0.50◦. Using the yaw-rate-basedangle estimate on the same hardware configuration, the worst-case error increasesto 35% with an absolute error of 0.70◦. Parameter sensitivity analysis shows thesteering model to be robust to large parameter variations, while the bicycle modelrepresents the primary source of estimation uncertainty.