blog

With Isuzu Advanced Engineering Center conducting research on elemental technologies, the Isuzu group is developing demonstration experiments for automated trucks and buses. To achieve tracking to the target orbit, one of the components of automated driving, the lateral displacement and angular error must be minimized as much as possible. One of the control methods is to generate a target yaw rate from the target value and sensor inputs and manipulate the steering angle using a lateral motion controller. Model Predictive Control Toolbox™ made it easy to implement model predictive control for target yaw rate generation—a process that can be otherwise challenging. On the other hand, to design a “lateral motion controller” that determines the steering angle so that the vehicle follows the target yaw rate, it is necessary to derive a model of the vehicle to be controlled. Since it is difficult to model the lateral motion characteristics of a truck, Isuzu Advanced Engineering Center is working on designing a control system using data-driven control. Using MATLAB®, they developed a prototype tool to automatically generate controller gains through data-driven control, creating a seamless environment from data measurement to controller design and implementation. They were also able to verify the effectiveness of the prototype tool by quickly creating a truck model using the Vehicle Dynamics Blockset™ library blocks before implementing the lateral motion controller.