AN INTELLIGENT CENTRALISED CONTROL SYSTEM FOR AUTOMOTIVE APPLICATIONS

AN INTELLIGENT CENTRALISED CONTROL SYSTEM FOR AUTOMOTIVE APPLICATIONS

Copyright © 2002 IFAC 15th Triennial World Congress, Barcelona, Spain www.elsevier.com/locate/ifac AN INTELLIGENT CENTRALISED CONTROL SYSTEM FOR AUT...

139KB Sizes 15 Downloads 136 Views

Copyright © 2002 IFAC 15th Triennial World Congress, Barcelona, Spain

www.elsevier.com/locate/ifac

AN INTELLIGENT CENTRALISED CONTROL SYSTEM FOR AUTOMOTIVE APPLICATIONS Gareth Thomas * and Christopher P. Jobling ** *

**

Calsonic Technology Centre Europe. Llanelli, Carms, SA14 8HU, UK. [email protected] Centre for Communications and Software Technology, University of Wales Swansea, Singleton Park, Swansea SA2 8PP, UK. [email protected]. Abstract. This paper discusses the use of distributed and centralised control systems as applied to an automotive vehicle. It examines the positive and negative aspects of each of the control philosophies as discussed in published literature. The paper then goes on to discuss an intelligent control system (ICS), with increased flexibility and advanced safety critical features. A comp arison is also given of the benefits of the application of the ICS compared to the current distributed and centralised control systems. Copyright © 2002 IFAC Keywords . Automotive, Control, Centralised Control, Communication, Bus, Distributed, Networks, Power, Safety. 1. INTRODUCTION.

one of the main reasons for the introduction of the new 42V vehicle electrical system1 .

With regards to the automotive industry, the number of functions being offered to potential customers is increasing. These functions are normally realised by a mixture of electronic modules driven by software. As an example, if we observe the North American market (Fig. 1), we see that over a 15-year period between 1994 and 2009, the estimated growth of electronics fitted to a vehicle is around 43% (Kobe 2000).

With the vehicle manufacturers striving to gain market advantage through increased user functionality and reduced emissions, their main aim is to limit the amount of negative effects such as weight, complexity etc. on the vehicle as a whole. One of the biggest areas that affect the vehicle is the number of interconnections between sub-systems. These interconnections normally take the form of either a direct wiring harness between units, or by one of the available communications buses (for example LIN, CAN2.0B, J1850 (Shultz 1997)). A major effect on the choice of which interconnection technology to use, is determined by how the individual control systems are laid out around the vehicle, and how they interact with one another (Ruston & Merchant 2000).

Electronics $ Per Vehicle

($) 2000 1500 1000 500

Generally in the automotive sector there are currently there are two trains of thought on the best philosophy of implementing vehicle control systems: the distributed control systems (DCS) introduced in the next section and the centralised control systems (CCS) introduced in Section 3. We analyse the advantages and disadvantages of DCS and CCS in Section 4, before introducing a new Intelligent Control System (ICS) in Section 5. Advantages of ICS over DCS and CCS are given in the conclusions to the paper.

0 1994 1999 2004 2009 Source: Freedonia group (Cleveland Ohio)

Fig. 1. Growth of Electronics as Fitted to Automotive Vehicles. This increasing use of electronics has some hidden effects on the vehicle as whole. For example, increased functionality results in a potential increase in the number of components, hence an increased vehicle weight. This additional weight results in an increase in fuel consumption and exhaust emissions. The second effect is the added loading on the vehicle's electrical system. Millier, Nicastri and Zarei (1992) argue that the peak electrical power of a typical 12Volt generation system will level out around 3KW. Above this rating, as the alternator size grows to meet demand, its cost becomes disproportionately high. Increasing power demand is

2. DISTRIBUTED CONTROL SYSTEM The philosophy behind a distributed control system (DCS) for automotive applications is that no single electromechanical unit should have complete control over the majority of the vehicle's control systems. Each controller operates as a stand-alone unit. In certain circumstances there is a requirement for the 1

403

1st vehicle will be Toyota 2001.

individual controllers to interact with one another. For example, process variables such as coolant temperature and vehicle speed need to be shared between the climate control system and the instrumentation cluster. A typical vehicle DCS is shown in Fig. 2.

LOAD

LOAD

CONTROLLER

LOAD

CONTROLLER

LOAD

CONTROLLER

Direct Connection

Communications Bus CONTROLLER

LOAD

LOAD

LOAD

CONTROLLER

CONTROLLER

LOAD

LOAD

LOAD

Fig. 4 Centralised Control System

CONTROLLER

LOAD

Sometimes CCS units are referred to as Smart Junction Boxes (see Fig. 5 and 6). Some intelligence is built in for power handling: for example for overvoltage and short-circuit fault detection. The CCS communicates via direct wiring connections or through a communications bus.

LOAD

Fig. 2. Block Diagram of a Distributed Control System. The implementation of a DCS starts with the power distribution sub-system. This normally takes the form of several units containing multiple relays, fuses, timers etc. These power distribution units are typically located under the bonnet and inside the dashboard of the vehicle. Each sub-control system (wipers, lights, ABS, airbags etc.) has its own independent control system. Fig. 3 Illustrates a vehicle with a DCS. Items 1-46 represent the individual controllers, loads etc. Items 8-10 represent the power distribution units, these supply power to each of the other control units. 5

31

Fig. 5 Example of a Smart Junction Box (Scramm, Bouda and Brand, 2001)

6 32

23

10

12

45

20

28

29 44 13 14

39

22

2

1

9

37

17 7

19

18

15 25

36

21 35

11 30

24

46

16

27

26

8 3

33

4

38

34

Fig. 3. Illustration of vehicle electronic modules.

3. CENTRALISED CONTROL SYSTEM

Fig. 6 Example of a Smart Junction Box

Unlike a DCS, a centralised control system (CCS) has a single central controller. Typically these units consist of a thermoplastic base housing with integrated metal electrical conductors and contacts. Some contain a microprocessor, but they are mainly made up of relays, fuses, and timers (Fig. 4).

It is interesting to note that many vehicle manufactures, use a centralised approach, but split the control features between two or three CCS units (Fig. 7). So at present the use of a truly centralised control system, that is a single unit controlling a vehicle, has not yet been developed.

404

Smart Junction Box

33

3

7



24 26

30

34

16 17 10

5

22

31

11 12

6

6

15

9

2 1

21 14

8 4

4.2 Disadvantages of CCS over DCS

Smart Junction Box 13

27

18 19 20

28 33

23

29

32

25



Fig 7. Example of a vehicle with multiple central controllers

4. COMPARISON OF DCS AND CCS APPROACHES

5. INTELLIGENT CONTROL SYSTEM

The advantages of the CCS approach over DCS are given in the following subsections.

The philosophy behind the introduction of an Intelligent Control System (ICS) is simple. Its prime aim is to enhance the functionality of the current DCS and CCS systems by further integrating components and subsystems. The ICS offers all the benefits of the current systems with only minor disadvantages.

4.1 Advantages of CCS over DCS •







Since the CCS is the only means of controlling vehicle functionality, if a system failure occurs there is a higher risk that the vehicle will become inoperable. By contrast, if a controller or sub-system of a DCS fails, the total system still has the possibility to function correctly. On a functional level, due to the integration of functionality in the CCS, if only one of the functions fails, the whole unit may need replacing. This results in a cost penalty compared to changing an individual unit from a DCS. If failure does occur in a DCS it will be easier to diagnose, since the individual units can normally be replaced with little effort. However if communications buses are used, it can be very difficult to diagnose a fault, if the control signals are communicated via the bus. (Shultz 1997)

Due to the lower number of individual components and the reduced number of interconnections, a CCS has a simpler vehicle wiring harness than a DCS. Since the concentration of development effort is on a single controller, a CCS requires fewer resources, and benefits from a reduced amount of validation testing and design/ development time. (Scramm, Bouda and Brand, 2001) As only one unit is needed only a single set of tooling is required, leading to a much lower tooling cost. The manufacturing assembly line is more straightforward: for example the complexity of supporting housing variations leads to a lower level of raw material handling. (Scramm, Bouda and Brand, 2001) The control unit is typically larger than a single DCS component; however, it takes up less packaging space within the vehicle for a total system. Additionally, because of the integration of functions, one of the biggest benefits is the deletion of individual controllers. Due to the reduction in the number of controllers and the wiring harnesses there is a total system weight reduction. The typical weight of a dashboard DCS wiring harness can be 11Kg. With a number of individual controllers, DCS suffers from a greater risk of EMI (Electro Magnetic Interference) and RFI (Radio Frequency Interference) emissions being generated around the vehicle.

The intelligent control system, can really be defined as an advanced form of a central control system. At its centre lies a Central Control Unit (CCU). LOAD

LOAD

LOAD

COMMUNICATIONS BUS CENTRAL CONTROL UNIT

LOAD

LOAD

LOAD

Fig. 8 Intelligent Control System Figure 8 illustrates the Intelligent Control System with the implementation of the CCU. Although the CCU performs all the functions undertaken by the CCS, physically the CCU is completely different in it's construction. It consists of a • • • •

Thermo plastic base carrier Multi-layer Printed Circuit Board (PCB) 2 x heat sinks. Lid with integral cooling fan.

Figure 9 is an illustration of an actual CCU.

405







Fig 9. Central Control Unit •

The PCB retains twin 16 bit microprocessors where the intelligent software algorithms are retained in flash memory. The first processor is responsible for safety-critical functions such as lights, and screen wash wipe, whilst the second controls non safety-critical functions, such as climate control system, and internal panel light dimming. All electromechanical relays, fuses timers etc., have been replaced by solid-state switches with the result that each solid state switch replaces a relay, 2 fuses and a wiring harness. (Siemens 1997)





Originally, the CCU concept was developed using a vehicle with a DCS. If we consider the total functionality of the CCU we find that a number of vehicle functions have been integrated by software.



Fig 10 illustrates the actual individual controller's systems, which have been replaced by software routines.



6. COMPARISON OF CCS AND ICS CONTROL STRATEGIES



The advantages of the ICS approach over CCS are: •

All control functionality can be totally reconfigured by software. The CCS is restricted to its physical internal connections and limitations of its electromechanical devices.

• •

406

All control function characteristics are stored in memory. For example, for the turn indicator, the logic sequence and number of flashes per minute are defined in software. Safety critical and normal functions are split between microprocessors. These are dedicated to controlling and monitoring each of the specified functions. The CCU offers a higher level of quality and reliability, as it contains no electromechanical fuses and relays. It has been stated by Siemens (1997), that semiconductor switches provide a much higher level of failure protection: typically by a factor 10 over the equivalent relays. Under software control the, functions can easily be modified without changes to hardware. For example, a simple ON/OFF signal can be changed to pulse-width modulation (PWM) (for proportional action) with very little effort. The CCS would require physical changes in components to achieve the same functionality. Each output device is capable of running loads from 0.5A up to 30A, without hardware modifications. The CCS would again require physical hardware modifications. Additional intelligence is built into the CCU; each of the microprocessors continually monitors its partner. If a fault is detected with one of the processors, its partner will take control of its inputs / output's and at the same time switch off its power supply. This allows the functionality for the user to be unaffected by the failure of the CCU. If faults occur with the systems loads, the solid-state switches are capable of detecting open circuits, short circuits, over voltages etc. This is a similar function as the CCS except that the solid-state switches send a signal back to the controlling microprocessor. The microprocessor can then switch off the effected solid-state switch and isolate the system load. The benefit of this is that the microprocessor can, after a period of time switch the load back on to see if the fault has cleared. To undertake this task on a CCS additional sensing circuits would be required. The CCU has the ability to record all faults detected; this takes the format of date, time, fault code etc. All information can be accessed at the time of the vehicles service, or when repairs are required. The CCU uses dual power supplies, so if a major fault occurs with the power inputs the CCU, can isolate the effected supply and still remain functional. The cost of the CCU is cheaper than multiple control units. The control system only uses single wires from the CCU to each load; return loops are achieved via the vehicles chassis.

• •







One controller can be used for a whole range of vehicles and models, by loading new software routines into flash memory. The CCU has the benefit of multiple communications buses (CAN 2.0B), hence a further reduction in the number of wiring harnesses required. The CCU uses PWM control signals for all its inductive loads i.e heater blower motor etc. this gives the benefit of soft start. Which results in the vehicle electrical system not being loaded to such a level as the CCS. I.e. if a motor runs at 10 Amps, then on a relay ON/OFF system the motor could give a peak starting current of 30 - 40 Amps. This is unlike the ICS where the minimum current is drawn to get the motor to rotate, i.e. typically 3 Amps. There is a major reduction in the number of harness interconnects between components and systems. Typically on the development vehicle (DCS) 60% of all harnesses either reduced in wire gauge or were deleted altogether. It has been estimated when compared to the CCS system a saving of 20% is achieved. (Scramm, Bouda and Brand, 2001) When applied to an actual vehicle, a 55% weight saving has been achieved over the equivalent DCS.

These benefits, combined with the intelligent control system's ability to manage the power applied to the loads, results in improvements in the vehicle's total power loading. This has the knock-on benefit of reduced fuel consumption and exhaust emissions. The ICS also offers the vehicle user the benefit of being able to use the vehicle without reduced features, even when 50% of the control system has failed. There are definite advantages in the implementation of an intelligent control system, when compared to the more traditional DCS and the more recently introduced CCS vehicle control systems. The vehicle manufacturers will be the limiting factor in the implementation of an ICS system. The individual component cost of the CCU, although higher, gives a total system saving compared to the CCS approach. Costs will be reduced through design iteration and the multiple project implementation of the ICS system. This development has proved that an ICS is technically feasible and gives a number of advantages, which outweigh any of its negative aspects. REFERENCES Kobe, G. (2000) "What’s Driving Electronics Growth?" Automotive Industries, August 2000 page 29.

Its disadvantages are • • • • •

The design as it is currently implemented is limited to 22 loads, each of up to 30 A per load. The CCU still requires the main vehicle isolation components to protect the power supply entering the CCU. The CCU has to be carefully located in the vehicle, because, under full load conditions, the heat-sink surface becomes hot. There are some complex manufacturing processes required to manufacture the unique heat sinks. Due to the large heatsinks the weight of the CCU is greater than a single CCS unit

Millier, Nicastri and Zarei (1999) ‘Future System Architectures.’ Power Conversion & Intelligent Motion (PCIM). Online Internet Journal – http://ww.pcim/miller1/Ford Co. Dearbourn, Michigan Rushton, G. and Merchant, R. (2000). "Auto-Neural Systems. Vehicle Electrical/Electronic System Design Considerations." 2000 Conference Proceedings Society of Automotive Engineers. SAE No. 2000-01-0131 Schramm, Dr D., Bouda H, Brand, R. (2001) Tyco Europe. Central Electrical Power Distribution Modules. AutoTechnology Volume No.1 August 2001. Publishing Power Limited.

7. CONCLUSIONS

Shultz, G. (1997) Portable On -Board -Diagnostic (OBD) II CAN Scan Tool. Siemens Components Inc. 1997 Conference Proceedings Society of Automotive Engineers. SAE No. 971126.

The ICS offers several advantages over the traditional DSC and CCS. These have a major effect on the vehicle's functionality and control systems as a whole. • The ICS gives the benefit of being reconfigurable through software. • The ICS gives improved weight savings compared to a complete CCS.

Siemens (1997) ‘Sense Highside Switch in Smart Power Technology takes over Fuse Function’ Special Subject Book – Technical Publication B112-H7088-X-X-7600 Siemens AG.

407