Systems and methods automatically rescale an original engine model so that it models an engine of a different size. The original engine model may be coupled to an engine controller model, and the systems and methods may also rescale the original controller model to produce a rescaled controller model matched to the rescaled engine model. The original engine model may include engine parameters and engine lookup tables, and the original controller model may include controller parameters and controller lookup tables. Rescaling factors indicating the size of the new engine being modeled may be received, and ratios may be computed as a function of the rescaling factors. Original engine parameters and controller parameters may be rescaled based on the ratios. Original engine lookup tables and controller lookup tables may be reshaped based on the ratios.

According to various embodiments, described herein are methods and systems for collecting data for training a load inference regression model for use in an ADV. According to one exemplary method, an ADV is manually driven on a segment of a road for a number of periods of time. During each period of time, a set of features of the ADV are recorded, including one or more features at a first time prior to a gear shift from a first gear position to a second gear position, and one or more features at a second time after the gear shift. For each of the number of periods of time, a weight of the ADV is also recorded using a weight sensor. The recorded features and the total weight of the ADV for each of the periods of time are then used to train a neural network regression model for inferring a load of the ADV in real time.

A steering force module on a vehicle test bench including a first main body and a transverse force actuator that is displaceable relative thereto. A transverse force being generated by a displacement of the transverse force actuator relative to the first main body, by means of which the transverse force can be applied to the steering system. Furthermore, a drivetrain module is present which consists of a second main body and a drive actuator that is rotatable relative thereto, the drive actuator being rotationally fixable by means of a second mechanical interface to a drive axle of the drivetrain, a torque that is independent of the transverse force being generated by a rotation of the drive actuator relative to the second main body is applied to the drive axle.

An automotive arithmetic system includes a main arithmetic unit that generates a travel route using deep learning based on an output from a vehicle external information acquisition device; an auxiliary arithmetic unit that generates a rule-based travel route in a free space without using deep learning; a safe route generation unit that generates a safe route, which is a route that the motor vehicle takes until the motor vehicle stops at a safe stop position; and an override processing unit that prioritizes one of the travel route generated by the main arithmetic unit, the rule-based travel route generated by the auxiliary arithmetic unit, or the safe route generated by the safe route generation unit, and determines a target motion of the motor vehicle so that the motor vehicle travels on the prioritized route.

Systems, methods, and non-transitory computer-readable media can determine mission data associated with a scenario encountered during operation of a vehicle. A first evaluation of the scenario can be determined by evaluating the mission data using a simulation behavior model based on simulated driving data. A second evaluation of the scenario can be determined by evaluating the mission data using an observed behavior model based on observed driving data. Vehicle performance of an autonomy system of the vehicle can be evaluated based on the first evaluation and the second evaluation.

A method for generating training data is disclosed. The method may include executing a simulation process. The simulation process may include traversing a virtual, forward-looking sensor over a virtual road surface defining at least one virtual railroad crossing. During the traversing, the virtual sensor may be moved with respect to the virtual road surface as dictated by a vehicle-motion model modeling motion of a vehicle driving on the virtual road surface while carrying the virtual sensor. Virtual sensor data characterizing the virtual road surface may be recorded. The virtual sensor data may correspond to what a real sensor would have output had it sensed the road surface in the real world.

Techniques are disclosed for performing hybrid simulation operations with an autonomous vehicle. A method of testing autonomous vehicle operations includes receiving, by a computer, a pre-configured scenario that includes one or more simulation parameters and one or more initial condition parameters, sending, to the autonomous vehicle and based on the one or more initial condition parameters, control signals that instruct the autonomous vehicle to operate at an operative condition, and in response to determining that the autonomous vehicle is operating at the operative condition, performing a simulation with the one or more simulated objects and the autonomous vehicle to test a response of the autonomous vehicle.

An automotive arithmetic device includes circuitry that calculates a first candidate route based on a vehicle external environment estimated using deep learning; sets a static safe area (SA2) based on a result of recognition of a target object outside a vehicle according to a predetermined rule; and determines a target motion of the motor vehicle. The circuitry selects the first candidate route as a travel route of the motor vehicle under a condition the first candidate route is entirely within the static safe area (SA2), and does not select the first candidate route as the travel route of the motor vehicle when the first candidate route at least partially deviates from the static safe area (SA2).

A method for controlling an actuator system of a motor vehicle includes utilizing a model predictive control (MPC) module with an MPC solver to determine optimal positions of one or more actuators of the actuator system. The method further includes receiving a plurality of actuator system parameters, and triggering the MPC solver to generate one or more control commands from plurality of actuator system parameters. The method further includes applying a cost function to reduce a steady-state tracking error in the one or more control commands from the MPC solver and applying the one or more control commands to alter positions of the one or more actuators, and applying a penalty term to the steady-state predictions of positions of the plurality of actuators to limit a difference between a steady-state prediction of the actuator system and a solution from the MPC solver.

According to one embodiment, there is provided a computing device and method for evaluating driving risk. The computing device includes an input circuit and a processor. The input circuit is configured to receive data from a vehicle. The data includes at least one of GPS data, acceleration data or image data of views external of the vehicle or inside the cabin. Thereafter, the processor is configured to identify a plurality of risks based on the data received from the vehicle, determine a plurality of weightages which are assigned to the plurality of risks, and generate a score based on the plurality of weightages for the plurality of risks.