University of Michigan

Toyota Motor Engineering & Manufacturing North America, Inc.

Jing Sun

Kenneth R Butts

Ilya V Kolmanovsky

Zhao Lu

Christopher R Vermillion

Mohammad Reza Amini

Huei Peng

Hao Wang

Yiheng Feng

Caihao Weng

Ken R Butts

Zhen Yang

Jacob Edward Faust

An integrated speed prediction framework based on historical traffic data mining and real-time V2I communications for CAVs. The present framework provides multi-horizon speed predictions with different fidelity over short and long horizons. The present multi-horizon speed prediction is integrated with an economic model predictive control (MPC) strategy for the battery thermal management (BTM) of connected and automated electric vehicles (EVs) as a case study. The simulation results over real-world urban driving cycles confirm the enhanced prediction performance of the present data mining strategy over long prediction horizons. Despite the uncertainty in long-range CAV speed predictions, the vehicle level simulation results show that 14% and 19% energy savings can be accumulated sequentially through eco-driving and BTM optimization (eco-cooling), respectively, when compared with normal-driving and conventional BTM strategy.

Multi-range vehicle speed prediction using vehicle connectivity for enhanced energy efficiency of vehicles

An engine and cabin thermal management system for use with a vehicle having an engine, a cabin heating system configured to thermally heat a cabin of the vehicle, a coolant system operably coupled to the engine and to the cabin heating system to thermally manage a temperature of the engine and a temperature of the cabin. The coolant system having one or more coolant thermal storage units fluidly coupled with a radiator and heater core of the coolant system forming a coolant loop. The system further having a control system configured to monitor and maintain at least a predetermined coolant temperature at the cabin heating system even during a coolant temperature decrease at the engine stops.

Actively controlled coolant tank to increase thermal storage capacity of hybrid electric vehicles

A method is provided for determining capacity of a battery. The method includes: defining a model for a battery, where the model relates terminal voltage of the battery to charged capacity of the battery; taking a plurality of voltage measures of the battery, where the voltage measures are taken through a range of states of charge and the range excludes the battery being fully charged and fully discharged; determining the parameters of the model by fitting the plurality of voltage measures to the model; determining an incremental capacity curve for the battery by taking derivative of the model; and quantifying a peak of the incremental capacity curve to thereby determine a capacity of the battery.

On-board state of health monitoring of batteries using incremental capacity analysis

The method comprises the steps of inputting a first set of data into both a physical system and into an SVR model. The method includes collecting a second set of system data from the physical system and stacking a plurality of SVR models to form an output prediction without feeding back the model output. Another aspect of the invention includes a method of hybrid modeling having delayed output measurement. This method includes the steps of inputting a first set of data into both a physical system and into an SVR model and collecting a second set of system data from the physical system. A modeling error is injected into both the second set of system data from the physical system and a third set of system data from the SVR model wherein the modeling error thereby leads to substantially improved model output.

SVR dynamic system modeling with delayed output measurements

Examples of the present invention include air charge estimation models using linear programming support vector regression (LP-SVR). Air model parameters may be updated during operation of a vehicle, allowing the air model performance to be improved in the presence of part-to-part variation and part aging.

Adaptive air charge estimation based on support vector regression

Example methods of modeling a nonlinear dynamical system such as a vehicle engine include providing a model using linear programming support vector regression (LP-SVR) having an asymmetric wavelet kernel, such as derived from a raised-cosine wavelet function. The model may be trained to determine parallel model parameters while in a series-parallel configuration, and operated in the parallel configuration allowing improved and more flexible model performance. An improved engine control unit may use an LP-SVR with an asymmetric wavelet kernel.

Asymmetric wavelet kernel in support vector learning

A support vector machine with wavelet kernel was developed for accurate modeling of nonlinear systems. A method of providing an optimized model of a nonlinear system includes using a support vector machine (SVM) having a wavelet kernel, where support vectors include a family of multidimensional wavelets. Training the SVM allows optimization of the number of support vectors, the weights of the support vectors, and the translation factors of the support vectors. Use of a novel linear programming approach reduces computational demands required for training, allowing optimized support vectors to give an optimized model of the nonlinear system. Further, on-line retraining is possible, so that the model can be adapted to changing conditions.

Linear programming support vector regression with wavelet kernel

A method for controlling an engine output with at least two actuators, including applying inputs to actuators which regulate a variable of an engine, evaluating the response of the system, determining the ability of the actuators to change the engine variable, determining the capability of the actuators to reject a disturbance, calculating an optimum actuator feedforward control function based upon the ability and capability determined and controlling the actuators using the calculated function and a feedforward control algorithm.

Systems and methods for regulation of engine variables

A system for controlling a plant process wherein the plant is over-actuated with N+1 actuators and provides N performance variables is disclosed. The system includes an outer-loop controller operable to compare N performance variable values from the plant with a reference set point and provide a desired virtual control input to an inner-loop controller. The inner-loop controller receives the desired virtual control input from the outer-loop controller and provides real control inputs to the at least N+1 actuators. The N+1 actuators receive the real control inputs and subsequently provide an actual virtual control input to the plant. Upon receiving the actual virtual control input from the N+1 actuators the plant is operable to produce N updated performance variable values in an optimized manner.

Growing community of inventors

Superior Township, MI, United States of America

Jing Sun