electrified vehicles as platforms for complex system...

This information is the exclusive property of DENSO CORPORATION. Without their consent, it may not be reproduced or given to third parties.DENSO INTERNATIONAL EUROPE

Electrified Vehicles as Platforms for Complex System Control

DENSO INTERNATIONAL EUROPETechnical Research Department

Hasan Esen03.09.2012HYCON2 WORKSHOP ON ENERGY


2 / 16AGENDA

1. DENSO Corporation Company Profile

2. Increasing Complexity of Automotive Systems

3. Example Studies

4. Outlook


3 / 16

global supplier of automotive technology, systems and components



4 / 16



� Established : 1949

� Employees (31.3.2010) : 120,812

� Subsidiaries and affiliates:

� Japan (80)

� Asia and Oceania (57)

� North & South America (39)

� Europe (36)

Headquarters: Nagoya, Japan


5 / 16



� Established : 1949

� Employees (31.3.2010) : 120,812

� Subsidiaries and affiliates:

� Japan (80)

� Asia and Oceania (57)

� North & South America (39)

� Europe (36)

Headquarters: Nagoya, Japan

DENSO AUTOMOTIVE Deutschland GmbH

European Technical Research Department

• E/E Platform & System Technologies

• ICT & Safety, Powertrain, E-mobility & infrastructure


6 / 162. Increasing Complexity of Automotive Systems

Information &

Energy Grid

Integration

�More ECUs, larger program size, larger scale

� Powertrain Trend: Increasing diversity

Internal

Combustion

Engine

Electric

Vehicle (EV)

Efficiency Increase

Biofuels

Hybrid Vehicles, Plug-in Hybrids

Range extender

Battery electric vehicle

Fuel cells

Lowest Global

Emission

�Increasing powertrain diversity

Vehicle & Grid

IntegrationNetworked &

Integrated

System

� Vehicle System Trend:



� Energy Flow (conventional vehicle):

Cabin Temperature

Electric Load

Driving

coolant

A/C

refrigant

Engine

Battery

fuel

Mechanical

Electrical

Thermal

1 function – 1 component

Heater Core (H/C)



� Energy Flow (hybrid, low fuel consumption vehicle):

Rankine

coolant

Exhaustheat

recovery

recuperation

Motor /

Generator

refrigant

Mechanical

Electrical

Thermal

� 1 function – multi component

Cabin Temperature

Electric Load

Driving

fuel

Battery

A/C

Heater Core (H/C)

Engine

Elec. Heater(PTC)


9 / 162. Increasing Complexity of Automotive System

Architectural problems

High control freedom & optimality

Fault-tolerance

…

3.1 Functional Architecture

based Control System Design

3.3 Fault Tolerant Battery

Control in Electrified Vehicles

� DENSO Example Studies

3.2 Model Predictive Control

for Cabin Heat Thermal Man.”

Challenging control problems today as a result of increasing complexity & scale

New control design concept and methodology are necessary for automotive systems


Vehicle

Coordinator

Vehicle

Motion

Powertrain

・・・ElectricalEnergy,

Body,Information,

CabinClimate,・・・

TransmissionOutputshaftTorque


･･･

･･･

Powertrain

Coordinator

PropulsionReference

ISG

Clutch

Trans-mission

Engine

Starter

Distribution

ISGTorque


EngineTorque

EngineSpeed

StarterON/OFF

ClutchEngagementSignal

GearRatio

VehicleMotion

Coordinator

MotionReference

VehicleStability

Optimization

LongitudinalAcceleration Tire

ForcesLateralAcceleration

YAWRate

VehicleSpeed

Translation

VehicleStability

Coordinator

Diff.AWD

Steering

Brake

Distribution BrakeTorque

AWDTransferRatio

TireSteerAngle

DifferentialRatio

10 / 163.1 Functional Architecture based Control System Design*

*T. Tashiro, S. Akiyama (DENSO Corp.) Global Powertrain Conference 2003, Ann Arbor, MI, 2003

� Aim: structuring the control system that gets larger and more complex

Layer 0: Vehicle Domain Layer 1: Motion Domain

Layer 2: Stability Domain

Layer 1: Powertrain Domain

Main features:

- Restructure vehicle control

under simple rules.

- Redefine and allocate all of

control function appropriately

- Standardize interface of each

component.

- Hide localized information for

outside of the component.

- Parallel development of each

component

� This hierarchical structure serves as a framework. It has been gradually embodied


11 / 16

� Energy Flow (hybrid, low fuel consumption vehicle):

Rankine

coolant

Exhaustheat

recovery

recuperation

Motor /

Generator

refrigant

Mechanical

Electrical

Thermal

Cabin Temperature

Electric Load

Driving

fuel

Battery

A/C

Heater Core (H/C)

Engine

Elec. Heater(PTC)

distribute the workload between components to heat the cabin in order to achieve real-world (driving, electric and thermal domains) fuel efficiency

3.2 Model Predictive Control for Cabin Heat Thermal Man.


12 / 16

Manipulate: Some options of engine, Electrical heater, (PTC), Heater Core (H/C)

Observe/Evaluate: Battery, engine, M/G, other electrical and thermal loads

3.2 Model Predictive Control for Cabin Heat Thermal Man.*

� Cabin Heat Problem Covering Range

Manipulated

Observed

Not included

* In collaboration with Prof. A. Bemporad (IMT Lucca) & his team

Rankine

coolant

Exhaustheat

recovery

recuperation

Motor /

Generator

refrigant

Cabin Temperature

Electric Load

Driving

fuel

Battery

A/C

Engine

Elec. Heater(PTC)

Heater Core (H/C)


13 / 16

� Control Problem

MPCCabin

Heat

Sys.

u1ym

routput

Complex relations are identified & simplified as linear-time variant (LTV) models

Multi-objective MPC is formulated s.t. LTV prediction models and constraints

For given measurements (ym) and references (r),

manipulate H/C (u1), PTC (u2), and engine power

(u3) to achieve the control objectives:

� heat power reference tracking

� maintaining battery SOC in its limits

� minimizing fuel consumption

u2

u3

� Simulation Results (NEDC, Ambient Temp 5oC)

0 200 400 600 800 1000 12000.3

0.4

0.5

0.6

0.7

0.8Battery SoC [%]

time [s]0 200 400 600 800 1000 12000

0.5

1

1.5

2

2.5

3

Sim Time [s]

He

at P

ow

er

[kW

]

Heat Power Tracking

Fuel savings

w.r.t. baseline:

~ 3%

Benefit:

Better competitiveness

3.2 Model Predictive Control for Cabin Heat Thermal Man.


14 / 163.3 Fault Tolerant Battery Control in Electrified Vehicles*

� Different Battery Topologies

Type Single String Double String Multi-String

Complexity

Sensitivity to

1-cell failure

fuse

☺ �

� ☺

1 parallel x n series 2 parallel x n series m parallel x n series

improve the reliability of simpler topologies (single string) using

active fault detection and control mechanisms

* In collaboration with Prof. J. Stoustrup (Uni Aalborg) & his team. Published in Safeprocess’12 Conference


15 / 163.3 Fault Tolerant Battery Control in Electrified Vehicles

� Software modifications: Active Fault Detection (AFD)

1. State space battery model

Input : current I

Output : voltage Vo

States : voltage across bulk and

surface capacitorsVcb, Vcs

increase the sensitivity of parameter changes

Non-convex optimization problem

4. Compare the estimates with nominal values for fault detection

3. Design the input current signal I

2. Estimate the bulk capacity Cb and the terminal

resistance Rt using Extended Kalman Filter (EKF)

� Results

� Reliability of single-string topology increases with AFD

� Early warning to the driver, before severe failures occur


16 / 164. Outlook

Zero emission, zero fatality

Utilizing advanced

computing technologies Standardization

Challenging control problems tomorrow as a result of increasing complexity & scale

cyber-physical systems, large-scale distributed systems, systems of systems

Vehicle as a component of

system of systems

Absolute fault-tolerance

Information management

DENSO EU Technical Research strategy: support and actively develop the technology together with

diversified EU partners

Secure V2G communication


17 / 16

electrified vehicles as platforms for complex system...

Documents