An apparatus of controlling a gear shifting of a vehicle, and a method thereof, to improving shift quality by minimizing the jerk generated in a shifting process of the vehicle, includes storage that stores a Gaussian process (GP) model on which machine learning is completed, and a controller that detects a change amount of engine torque and a change amount of an engagement-side clutch torque based on the GP model, and controls the gear shifting of the vehicle according to the change amount of the engine torque and the change amount of the engagement-side clutch torque.

A line pressure control method for a double clutch transmission (DCT) includes estimating a line pressure, which decreases with stoppage of an electric oil pump, based on a linear regression model using state variables of the DCT that are related to a line pressure change, and driving the electric oil pump when the line pressure estimated based on the linear regression model reaches a predetermined lower limit.

A bench test calibration method for generating wet clutch torque transfer functions includes obtaining in-vehicle clutch torques at a set of shift conditions; performing a series of bench tests at various clutch pack clearances and lubrication oil flow rates at the set of shift conditions; adjusting clutch pack clearances and lubrication oil flow rates during the series of bench tests in response to a difference between a bench test measured clutch torques and the corresponding in-vehicle clutch torques exceeding a threshold; and recording relationships between first bench test measured torques and force profiles of a clutch actuator relative to the adjusted clutch pack clearances and lubrication oil flow rates for each of the set of shift conditions as a first transfer function.

A transmission calibration tool automatically generates a detailed gearbox model based on a user input transmission topology description. During transmission calibration, the tool accepts inputs from transmission speed and torque sensors and estimates component torques for each gear element and each shift element. Following a shift or other transmission event, the calibration tool plots the component torques as a function of time, permitting the calibration engineer to better understand what is occurring during the event, and thus reducing the time required for calibration. The calibration tool also adapts several transmission component models and outputs the adapted models to provide insight into actual transmission component behavior.

A transmission controller of a toroidal continuously variable transmission includes a gain setting unit that adjusts a gain of closed-loop control for calculating a target value of a roller position in accordance with a change in a rotation speed of a disc in a first rotation speed range and a second rotation speed range higher than the first rotation speed range. The gain setting unit changes, in the first rotation speed range, the gain so that sensitivity of the closed-loop control decreases with an increase in the rotation speed, and changes, in the second rotation speed range, the gain so that a rate of decrease in the sensitivity of the closed-loop control with the increase in the rotation speed is smaller than that in the first rotation speed range.

An apparatus and a method for controlling a transmission of a vehicle includes storage that stores a dual clutch transmission (DCT) dynamic model and a machine learning-based Gaussian process (GP) model, and a controller configured for determining a first engine torque used for optimal shifting according to the DCT dynamic model, determines an engine torque compensation value according to the machine learning-based GP model, and controls a shifting of the vehicle according to the first engine torque compensated by the engine torque compensation value.

An apparatus of controlling a gear shifting of a vehicle, and a method thereof, to improving shift quality by minimizing the jerk generated in a shifting process of the vehicle, includes storage that stores a Gaussian process (GP) model on which machine learning is completed, and a controller that detects a change amount of engine torque and a change amount of an engagement-side clutch torque based on the GP model, and controls the gear shifting of the vehicle according to the change amount of the engine torque and the change amount of the engagement-side clutch torque.

A line pressure control method for a double clutch transmission (DCT) includes estimating a line pressure, which decreases with stoppage of an electric oil pump, based on a linear regression model using state variables of the DCT that are related to a line pressure change, and driving the electric oil pump when the line pressure estimated based on the linear regression model reaches a predetermined lower limit.

A method for determining a drag torque coefficient of a transmission includes performing rotational speed synchronisation involving application of synchronisation torque to a first transmission component, and obtaining an initial rotational speed of the first transmission component before the rotational speed synchronisation, and a final rotational speed of the first transmission component after the rotational speed synchronisation, and time period for performing the rotational speed synchronisation. Also, obtaining information relating to a level of the synchronisation torque applied to the first transmission component during the rotational speed synchronisation, and information relating to a total moment of inertia associated with the first transmission component. In addition, determining the drag torque coefficient based on the obtained initial rotational speed, the obtained final rotational speed, the obtained time period, the level of the applied synchronisation torque and the total moment of inertia associated with the first transmission component.

A method for determining a state of a transmission for a vehicle, including providing an input for a first generative model depending on a route information, a vehicle speed, a probabilistic variable, and an output of a second physical model, and determining an output of the first model characterizing the state in response to the input for the first model. The first model comprises a first layer trained to map input to an intermediate state. The first model comprises a second layer trained to map the intermediate state to the state depending on the output of the second model. The method includes providing an input for the second physical model depending on at least one vehicle state and/or the route information, and determining an output of the second model in response to the input for the second model. The output of the second model characterizes limit(s) for the intermediate state.