GB2466479A

GB2466479A - A method of inhibiting stop-start control using gear selector position

Info

Publication number: GB2466479A
Application number: GB0822390A
Authority: GB
Inventors: Themi Philemon Petridis; Ian Halleron; Peter George Brittle
Original assignee: Ford Global Technologies LLC
Current assignee: Ford Global Technologies LLC
Priority date: 2008-12-09
Filing date: 2008-12-09
Publication date: 2010-06-30
Anticipated expiration: 2028-12-09
Also published as: CN101749125A; CN101749125B; DE102009047233A1; GB2466479B; GB0822390D0; DE102009047233B4

Abstract

A method for controlling automatic starting and stopping of an engine (2, fig 1) of a micro hybrid vehicle, the method comprises in step 810 carrying out a validity or a plausibility test of a transmission state sensor TSS (7), in step 820 checking whether an odd and even plausibility test been completed, in step 830 determining if the plausibility tests have been passed and if yes in step 840 enabling automatic stop-start control of the engine (2). If the answer in step 830 is no then in step 850 stop-start control is inhibited and in step 860 a user is warned of a system error. The transmission state sensor TSS (7) provides an indication of a position of a gear selector and includes an odd gear selector position and an even gear selector position. The method reduces a risk of unintended movement of the micro-hybrid vehicle (1) due to erroneous transmission state sensor (7). An apparatus for controlling automatic starting and stopping of an engine is also disclosed, the apparatus comprises a transmission state module which receives a signal from the senor (7) and provides an output to a stop-start controller to inhibit stop-starting if the sensor (7) output is confirmed not to be valid for either or both odd and even gear selector positions.

Description

A Method and Apparatus for Controlling Automatic Starting and Stopping of the Engine of a Vehicle This invention relates to a method for controlling automatic starting and stopping of the engine of a vehicle and in particular to determining whether the output from a transmission state sensor used to determine the engagement state of a manual transmission fitted to a Micro-Hybrid vehicle can be trusted.

It is known that the fitting of Micro-Hybrid technology to manual transmission vehicles enables a reduction in fuel consumption by invoking automatic engine stops and starts when the vehicle is stationary. Different engine shut-down and restart strategies such as Stop-in-Neutral (SIN) and Stop-in-Gear (SIG) can be utilised. In both SIN and SIG configurations there are circumstances when a reliable signal indicating that the transmission is in neutral is required. This signal is used by the engine management system Stop-Start logic as a condition to determine whether an engine shut-down or restart is permissible. This is a safety critical requirement to avoid unintended vehicle movement caused by cranking the engine whilst the driveline is engaged.

In a SIN system, the engine is typically shut down when the vehicle is stationary, the transmission is in neutral and the clutch pedal is released. To restart the engine following a shutdown, the driver typically triggers a restart by pressing the clutch pedal if it is detected that the transmission is in neutral.

SIG stops are typically invoked if both the clutch and brake pedals are depressed, possibly with the transmission in gear and a SIG restart is typically invoked when the brake pedal is released while the clutch remains depressed.

Furthermore, system-induced restarts when the transmission is in neutral are a possible requirement for both SIN and SIG applications to prevent the driver being stranded due to low battery voltage or to ensure cabin comfort during a prolonged stop event.

In general therefore, transmission neutral sensing is a fundamental requirement for operating a SIN strategy and it is also required for SIG systems if system induced restarts are to be utilised.

Robust and reliable transmission engagement state sensing is not a trivial concept to implement due to the tolerance build-up/ tolerance chain of the mechanical parts used in a transmission, coupled with sensor and magnet tolerances and inaccuracies and external noise factors.

Additionally, the definition of neutral is not a straightforward concept. If neutral is defined as the transmission selector turret rotational region in which zero torque is transmitted, then this rotational region is typically too small to sense accurately given the tolerance stack up, measurement inaccuracies and noise factors.

Furthermore, the transmission mechanical parts tolerance stack up affects the selector turret rotational neutral resting position. Rotational movement of the selector is measured from zero degrees rotation, zero degrees being the neutral resting position in that particular transmission.

This complicates the calibration of fixed thresholds to determine bounds of any neutral window or in-gear zones which would be valid for all transmissions. Although it is possible to learn-out the transmission to transmission differences in the neutral resting position by some form of end of transmission-line zero-offset learning (neutral offset learning) or end of vehicle-line zero-offset learning, this process introduces a risk that the learning will not be performed correctly or, when a transmission is replaced in-service, the learning might not be updated and hence an error is introduced in the learnt offset. This can result in serious consequences in the form of unintended vehicle movement during Stop-Start operation and so must be avoided.

It would be an advantage to provide a method for controlling the operation of a Micro-hybrid vehicle and in particular to a method for confirming whether the output from a transmission state sensor used to determine the engagement state of a manual transmission can be used to safely control automatic starting and stopping of an engine of the vehicle.

According to a first aspect of the invention there is provided a method for controlling the automatic starting and stopping of the engine of a vehicle, the vehicle having an engine driving a manual transmission and a sensor to provide an indication of an engagement state of the transmission by sensing the position of a gear selector, wherein the method comprises preventing automatic starting and stopping of the engine if an output from the sensor has not passed a validity check for both an odd gear selector position and an even gear selector position.

The method may further comprise indicating to a user of the vehicle when automatic stopping and starting is being prevented.

The validity check for an odd gear selector position may not be passed if the output from the sensor disagrees with an independent indication of an odd gear transmission state for a test period of predetermined duration, provided that the independent indication of an odd gear transmission state remains the same for the duration of the test period.

An indication that the output from the sensor is invalid may be indicated by the setting of a test failed flag.

The validity check for an odd gear selector position may be passed if the output from the sensor indicates that the state of the transmission is in an odd gear for the test period and the independent indication that the transmission is in an odd gear state persists for the test period.

An indication that the output from the sensor is valid may be indicated by the setting of a test passed flag.

The validity check for an odd gear selector position may further comprise checking agreement of the independent indication of an odd gear transmission state with the output from the sensor after the independent indication of an odd gear transmission state has been stable for a predetermined period of time.

The validity check for an even gear selector position may not be passed if the output from the sensor disagrees with an independent indication of an even gear transmission state for a test period of predetermined duration, provided that the independent indication of an even gear transmission state remains the same for the duration of the test period.

The validity check for an even gear selector position may be passed if the output from the sensor indicates that the state of the transmission is in an even gear for the test period and the independent indication that the transmission is in an even gear state persists for the test period.

The validity check for an even gear selector position further may further comprise checking agreement of the independent indication of an even gear transmission state with the output from the sensor after the independent indication of an even gear transmission state has been stable for a predetermined period of time.

According to a second aspect of the invention there is provided an apparatus for controlling the automatic starting and stopping of the engine of a vehicle, the vehicle havinq an engine drivingly connected to a manual transmission having a selector, the position of which determines whether the transmission is in one of an odd gear, an even gear and neutral, a sensor to monitor the position of the selector, and a transmission state module to receive a signal from the sensor and provide an output signal to a stop-start controller, wherein the transmission state module is operable to prevent automatic starting and stopping of the engine if the sensor output has not been confirmed to be valid for either or both of an odd gear selector position and an even gear selector position.

The sensor output may not be confirmed to be valid for an odd gear selector position if the sensor output disagrees with an independent indication of an odd gear transmission state for a predetermined period of time, and the sensor output may not be confirmed to be valid for an even gear selector position if the sensor output disagrees with an independent indication of an even gear transmission state for a predetermined period of time.

The transmission state module may be operable to receive inputs from a number of sensors located on the vehicle indicating the current operating state of the vehicle and use the inputs from two or more of the sensors to determine each independent indication of transmission state.

It is an advantage of the invention that unsafe operation of a micro-hybrid vehicle is prevented in the case where zero-offset learning has not been carried out for the transmission or has been carried out incorrectly.

The invention will now be described by way of example with reference to the accompanying drawing of which:- Fig.1 is a diagrammatic representation of a Micro-Hybrid motor vehicle according to an embodiment of the invention; Fig.2A is a scrap view of part of a transmission of the motor vehicle shown in Fig.1 showing the location of a transmission state sensor and magnetic target; Fig.2B is a pictorial view showing the motion of a transmission turret selector cylinder, the rotational position of which is sensed by the transmission state sensor; Fig.3A is a first pictorial view of a turret selector cylinder follower; Fig.3B is a second pictorial view of the turret selector cylinder follower shown in Fig.3A; Fig.4 is a block diagram of the data flow between the transmission state sensor and a Micro-Hybrid Stop-Start Module shown in Fig.1; Fig.5 is a chart showing the relationship of output signal from the transmission state sensor and selector cylinder rotation; Fig.6 is a chart showing various factors affectinq the determination of neutral and In-Gear threshold values for the transmission shown in Fig.1; Fig.7A is a flow chart of a method or validity check for confirming whether the output from a transmission state sensor used to determine a neutral engagement state of the transmission shown in Fig.1 is valid; Fig.7B is a flow chart of a method or validity check for determining whether an in-gear state indicated by the transmission sensor is valid; Figs.8A and 8B is a flow chart of a method for determining neutral thresholds for the transmission shown in Fig.1; Figs.9A and 9B is a flow chart of a method for determining In-Gear thresholds for the transmission shown in Fig. 1; Fig.1O is a time line of a neutral signal check test according to an embodiment of the invention; Fig.11 is a time line of an in-gear signal check test according to an embodiment of the invention; Fig.12 is a table showing the logic employed to confirm whether a neutral signal can be trusted; Fig.13 is a table showing the logic employed to confirm whether an in-gear signal can be trusted; Fig.14 is a chart showing the relationship of output signal from the transmission state sensor and selector cylinder rotation for a situation where zero offset learning is required; Fig.15 is a chart showing the relationship of output signal from the transmission state sensor and selector cylinder rotation for raw and zero-offset signals; Fig.16 is a chart showing the relationship of output signal from the transmission state sensor and selector cylinder rotation showing how validity checking of both odd and even in-gear states can be used to identify a potentially unsafe operating condition; Fig.17 is a chart showing the relationship of output signal from the transmission state sensor and selector cylinder rotation showing an unsafe operating condition produced by replacement of a transmission without new zero offset learning; and Fig.18 is a high level flow chart showing a method for reducing the risk of unsafe operation of a micro-hybrid due to sensor zero offset errors.

Referring firstly to Figs 1 to 5 there is shown a micro-hybrid motor vehicle 1 having an engine 2 drivingly connected to a manual gearbox/ transmission 3 via a clutch (not shown) . An electronic controller 4 is provided to control the operation of the engine 2 and includes a stop-start controller 6 to automatically stop and start the engine 2 and a transmission state module 5 to determine the operating state of the transmission 3.

The electronic controller 4 is arranged to receive a number of inputs or signals from sensors 9 including one or more of engine speed from an engine speed sensor, vehicle speed from a vehicle speed sensor, clutch pedal position from a pedal sensor, accelerator pedal position from a pedal sensor, brake pedal position from a pedal sensor and may also receive information regarding other components on the vehicle such as the state of charge of a battery (not shown) and the operating state of an air conditioning unit (not shown) Some or all of the inputs from the sensors 9 may be used by the stop-start controller 6 to determine when it is safe to stop and start the engine 2. It will be appreciated that the stop-start controller 6 and the transmission state module 5 could be separate units or be formed as part of a single electronic controller 4 as shown.

The transmission state module 5 is arranged to receive a signal from a transmission state sensor 7 attached to a casing 3B of the transmission 3. The transmission state sensor 7 is a magnetic PWM sensor and provides a signal based upon variations in flux between the transmission state sensor 7 and a magnetic target 8 associated with a turret selector cylinder 3A.

Fig.2A shows a typical H-gate' transmission configuration consisting of a shifter turret selector cylinder 3A located inside the main transmission casing 3B.

The shifter turret selector cylinder 3A rotates when a gear lever (not shown) is moved forwards and backwards to select respectively odd and even gears and it moves axially when the gear lever is moved left and right to change the plane in which the gear lever moves. Reverse gear can be configured as an odd gear or an even gear depending upon the configuration of the transmission 3.

The magnetic target 8 is attached to the shifter turret selector cylinder 3A and, in the example shown, the transmission state sensor 7 is located on the outside of the -10 -transmission housing 3B and detects rotational movement of the magnetic target 8. However, it will be appreciated that the transmission state sensor 7 could be mounted inside the transmission casing 3B.

Figure 2B shows the movement of the magnetic target 8 when different gears are selected. Although in this case the magnetic target 8 is fixed to the selector cylinder 3A so that it moves with the selector cylinder 3A this need not be the case and, in some applications, it is possible to attach the magnetic target 8 so that it only rotates and does not move axially.

Furthermore, in applications where the movement of the gear selector is linear between in-gear and neutral positions linear movement may be sensed rather than rotational movement.

Figs.3A and 3B show a follower 3C which is rotated by rotation of the selector cylinder 3A, the follower 3C has three detents 3E, a central detent corresponding to a neutral gear position, an odd gear detent to one side of the neutral detent and an even gear detent to the other side of the neutral detent. A spring loaded ball 3D is shown for engagement with one of the detents 3E, the ball 3D is slidingly supported by the transmission casing 3B either directly or via a bracket. It will be appreciated that the ball 3D could be replaced by a spring biased pin having a hemi-spherical end. The detents 3E define the neutral and in-gear positions for the transmission 3 and in particular the peaks located between the neutral detent and the in-gear detents determine whether upon releasing the gear lever the transmission 3 will move into gear (pull-in) or into neutral (no pull-in) as will be described in greater detail hereinafter.

-11 -Fig.4 shows the relationship between the transmission 3, the magnetic target 8, the transmission state sensor 7, the transmission state module 5 and the stop-start module 6 in greater detail and in particular the data flows therebetween.

Starting with the transmission 3 it can be seen that there is a physical link to the magnetic target 8 in the form of the mechanical connection of the magnetic target 8 to the selector cylinder 3A and a physical connection to the transmission state sensor 7 in the form of the mechanical connection of the transmission state sensor 7 to the transmission housing 3B.

There is a flux connection between the transmission state sensor 7 and the magnetic target 8 such that variations in flux can be sensed by the transmission state sensor 7 to provide a signal indicative of the rotational position of the selector cylinder 3A and hence whether the transmission 3 is in an odd gear, an even gear or neutral.

Note that the transmission state sensor 7 is only able to distinguish whether the transmission is in an odd gear (such as 1, 3, 5), an even gear (such as 2, 4 6) or neutral it cannot determine the exact gear the transmission 3 is in.

The transmission state sensor 7 outputs a signal indicating either that the transmission is in an odd gear or an even gear or in neutral and a quality signal produced by the transmission state sensor 7 itself indicating whether there are any faults with the transmission state sensor 7.

That is to say, the transmission state sensor 7 is an intelligent sensor and has a self diagnostic capability that produces a quality signal indicative of whether there are any faults associated with the transmission state sensor 7.

In Fig.4 these signals have been split into four inputs but in fact there are only two inputs to the transmission -12 -state module 5, a sensed position signal and a quality signal. To be more precise, the transmission state sensor 7 outputs a PWM signal which is either in range (between 10% and 90%) or out of range (>90% or < 10%) . The transmission state sensor 7 generates the out of range signal when there is a fault and so in fact there is only one physical output from the transmission state sensor 7. Input driver software in the transmission state module 5 interprets the PWM and, if the PWM is out of range (>90% or < 10%) the input driver software sets a quality signal to FAULT. If the PWM signal is in range (between 10% and 90%) the input driver software sets the quality signal to OK. The transmission state module 5 then compares the PWM signal to thresholds to set flags indicating whether neutral is or is not selected, an odd gear is or is not selected, an even gear is or is not selected.

The transmission state module 5 outputs a signal to the stop-start module indicating the engagement state of the transmission 3 along with a signal indicating the quality of this output. Note that in practice the transmission state module 5 compares the PWM signal to thresholds to set flags indicating whether neutral is or is not selected, an odd gear is or is not selected, an even gear is or is not selected.

Fig.5 shows a typical sensor signal plotted against shifter turret selector cylinder angular rotation on the x-axis. In this case the PWM sensor signal is shown, ranginq between 10 and 90% PWM duty cycle. The transmission 3 is at rest in neutral when at zero degrees rotation and the corresponding nominal sensor signal is then 50%. When the gear lever is moved forwards into one of the odd gears, the sensor signal decreases below 50% and conversely when one of the even gears is selected the sensor signal increases above ro )U a.

-13 -Sensor signals outside the 10-90% range are used for out-of-range failure modes of the transmission state sensor 7 to aid engine management system diagnostics. So for example a signal level of 5% would indicate a fault with the transmission state sensor 7.

It will be appreciated that the transmission state sensor 7 could also be arranged such that when the transmission 3 is in neutral the corresponding nominal sensor signal is 50%, when the gear lever is moved forwards into one of the odd gears the sensor signal increases above 50% and when one of the even gears is selected the sensor signal decreases below 50%.

Referring now to Figs.8A and 8B a method 100 for determining the odd and even neutral boundaries is shown.

After starting, the method advances to step 101 where the mechanical tolerances associated with the transmission 3 that may affect signal output are calculated. The mechanical tolerance Tmech is the mechanical transmission tolerance consisting of noises which are constant over time that affect the sensor signal when the transmission 3 is in its neutral resting position and do not vary with selector cylinder rotation. The mechanical tolerance Tmech is therefore derived from accumulated transmission mechanical tolerances and a noise factor analysis which in combination affect the transmission neutral resting position. These are noises which could be learned out by zero-offset learning, if learning was deemed as necessary. Tmech is depicted in figure 6 and represents the possible range of the neutral resting position of the transmission 3.

The method then advances to step 102 where a nominal neutral window boundary is defined. The normal definition of neutral which requires that zero torque is transmitted is -14 -not used in this method, rather a limiting criterion of transmission pull-in' is used to define neutral.

Pull-in limits are the positions where, if the engine 2 is cranked with the clutch pedal released and the transmission selector cylinder 3A is in a displaced position less than the pull-in limit, the transmission 3 will be forced back towards the neutral position but, if the transmission selector cylinder 3A is in a position that is beyond the pull-in limit, the transmission 3 will be pulled' into gear and the vehicle 1 is liable to jerk or move.

The pull-in limits are therefore considered as safe thresholds to use as the nominal neutral window boundary.

It will be appreciated by those skilled in the art that it is possible to transmit torque when the transmission 3 is positioned prior to the pull-in, but only under the following conditions: engine running, clutch not pressed, driver then applies considerable force on the gear lever.

Under these circumstances, the vehicle 1 is apt to creep.

However, in the context of stop-start operation, it is not easy to meet these conditions during a cranking event since if the engine 2 is stationary, it does not take a great deal of force on the gear lever to select a gear without the use of the clutch pedal. This means that the transmission 3 will be moved past the pull-in limit very easily and the transmission state sensor 7 would detect this. Therefore the driver would have to apply a high force on the gear lever just after engine cranking is initiated, without using the clutch pedal for there to be the slightest chance of vehicle movement. In addition, to comply with current safety standards, it is acceptable for the vehicle 1 to move forwards if the acceleration is less than 0.25 m/s2, that is to say, the vehicle 1 moves less than 0.5 meters in 2 seconds because this movement is considered sufficiently slow for the driver to react without compromising safety.

-15 -With reference to Fig.6, assuming a worst case transmission, AminPleven and AmjnPIodd represent the minimum shaft rotation from the neutral resting position to the earliest possible pull-in in the even and odd gear directions respectively.

Referring back to Fig.8A the next step is to calculate the pull-in measurement tolerances in the even and odd gear directions PITOLeVen and PIToLodd.

PITOLeven and PIToLodd are the tolerances consisting of noises affecting the sensor signal that vary with rotation of the selector cylinder 3A and these can include noises that affect the signal gradient and are constant over time and also those which vary over time. These noises cannot be learned out by zero-offset learning.

PITOLeven and PIToLodd are the necessary buffers between the earliest possible pull-in and how accurately it can be measured by the transmission state sensor 7 given all noise factors.

The method then advances to step 104 which is a check to determine whether end of line zero offset learning is required or whether the signal from the transmission state sensor 7 can be used without zero offset learning. If

Tmech < (AminPI -TQLeven) / 2 and Tmech < (AminPlodd -TQLdd) / 2 -16 -then the signal from the transmission state sensor 7 can be used without zero offset learning and the method advances to step 106.

However, if one of the tests in step 104 cannot be met then the method advances to step 105 where the transmission 3, magnetic target 8 and transmission state sensor 7 tolerances, specifications and noise factors must be reduced before the signal from the transmission state sensor 7 can be used without end of line zero offset learning. Reducinq the tolerance stack might involve tightening of component or assembly specifications or reducing or eliminating external noise factors such as temperature. After step 105 the method returns to step 101 and then steps 101 through 103 are repeated before once again executing step 104.

If the two conditions of step 104 are once more failed indicating that it is not possible to tighten component or assembly specifications sufficiently or reduce or eliminate external noise factors such as temperature then the method advances to step 107. In step 107 the method ends and end of line offset learning will be required. However, if the conditions of step 104 are then met, the method advances to step 106.

Note that step 104 in effect contains inequality conditions that in essence check that the neutral window calibration boundaries PlEvenlhresh and PlOddlhresh fall outside of the region bounded by the odd gear and even gear mechanical tolerances mecheven and mechodd (shown only as mech on Fig.6) This is because if not the position of neutral cannot be accurately determined directly from the transmission state sensor signal.

In step 106 the neutral window bounds PlEvenThresh and PlOddlhresh are calculated using the equations:- -17 -EvenThresh = -TOLeVen -TfflCCh odd OddThresh = "min'odd -TOLodd -TfflCCh even PlEvenlhresh and PlOddlhresh are the resulting safe neutral window bounds which are stored in the transmission state module 5 for use in determining whether the transmission 3 is in neutral.

Referring now to Figs.9A and 9B, a method 200 for determining odd and even In-Gear boundaries is shown.

After starting, the method advances to step 201 where the mechanical tolerances associated with the transmission 3 that may affect signal output are calculated. The mechanical tolerances Tmechodd and Tmecheven are the mechanical transmission tolerances consisting of noises which are constant over time that affect the sensor signal when the transmission 3 is in its neutral resting position and do not vary with selector cylinder rotation. The mechanical tolerances Tmechodd and Tmecheven are therefore derived from accumulated transmission mechanical tolerances and a noise factor analysis which in combination affect the transmission neutral resting position. These are noises which could be learned out by zero-offset learning, if learning was deemed as necessary. Tmechodd and Tmecheven are shown as Tmech in figure 6 and represent the possible range of the neutral resting position of the transmission 3.

The method then advances to step 202 where the minimum selector cylinder rotation to In-Gear in the odd and even directions for the transmission 3 are determined.

Assuming a worst case transmission, AminlGeven and AmjnIGodd on Fig.6 represent the minimum shaft rotation from the neutral resting position to the earliest possible In-Gear in -18 -the even and odd gear directions respectively. Note that the odd and even In-gear positions are the positions when the spring loaded ball 3D is at the bottom of the odd and even detents 3E of the follower 3C.

Referring back to Fig.9A the next step is to calculate the In-Gear measurement tolerances in the even and odd gear directions IGTOLeven and IGToL0dd.

IGTOLeven and IGToL0dd are the tolerances consisting of noises affecting the sensor signal that vary with rotation of the selector cylinder 3A and these can include noises that affect the signal gradient and are constant over time and also those which vary over time. These noises cannot be learned out by zero-offset learning.

The method then advances to step 204 where the even and odd In-Gear Thresholds IGEvenlhresh and IGoddlhresh are determined using the equations:-IGEvellTesh AminlGeven -IGTOLeven -Tmech odd IGOddTesh -IGTQLOdd -Tmech even IGEvenlhresh and IGoddlhresh are the resulting safe In-Gear window bounds which are stored in the transmission state module 5 as equivalent signal levels and which are used for comparison with the signal from the transmission state sensor 7 to determine whether the transmission 3 is In-Gear.

That is to say if:-The signal from transmission state sensor 7 is lower than the IGoddlhresh equivalent signal level; or -19 -The signal from transmission state sensor 7 is higher than the IGEvenlhresh equivalent signal level then an In-Gear present signal is sent from the transmission state module 5 to the stop-start controller 6.

Note that the above logic is for the case provided the signal goes low for odd gears and high for even gears as described and shown in Fig.6 it will appreciated that if the opposite sensor arrangement were to be used e.g. a high signal level equals an odd gear and a low signal level equals an even gear then the tests for an in-gear state above would become:-If the signal from transmission state sensor 7 is higher than the IGoddlhrsh equivalent signal level or the signal from transmission state sensor 7 is lower than the IGEvenlhresh equivalent signal level, then the transmission is confirmed to be in an in-gear state.

Referring back to Fig.9B after step 204 the method advances to step 205 which is a check to determine whether the signal from the transmission state sensor 7 can be used safely as an indication of an In-Gear state. If

IGEvenlhresh > PlEvenlhresh and IGoddlhresh > PlOddlhresh then the signal from the transmission state sensor 7 can be used and the method advances to step 207 where it ends and the in-gear odd and even thresholds can be used to -20 -determine from the transmission state sensor 7 the engagement state of the transmission 3.

Note that the above tests relate to the distance of the thresholds from the resting or zero degree position of the selector cylinder 3A and if transposed into a signal level test can be rewritten as Signal level IGEvenlhresh > Signal level PlEvenlhresh and Signal level IGoddlhresh < Signal level Ploddlhresh As before this holds true for the case where the signal from the transmission state sensor 7 goes high for even gears and low for odd gears as shown in Fig.6.

However, if one of the tests in step 205 cannot be met then the method advances to step 206 where the transmission 3, magnetic target 8 and transmission state sensor 7 tolerances, specifications and noise factors must be reduced before the signal from the transmission state sensor 7 can be used without end of line zero offset learning. Reducinq the tolerance stack might involve tightening of component or assembly specifications or reducing or eliminating external noise factors such as temperature. After step 206 the method returns to step 201 and then steps 201 through 204 are repeated before once again executing step 205.

If the two conditions of step 205 are once more failed indicating that it is not possible to tighten component or assembly specifications sufficiently or reduce or eliminate external noise factors such as temperature then the method advances to step 208. In step 208 the method ends and the In-Gear thresholds cannot be safely used for determining from the transmission state sensor 7 whether the -21 -transmission is in-gear. However, if the conditions of step 205 are then met, the method advances to step 207 as before.

Note that step 205 checks that the In-Gear Thresholds lie outside of the Neutral Thresholds because if they do not then the signal from the transmission state sensor 7 cannot provide an indication of when the transmission 3 is in gear.

Referring now to Fig.7A there is shown a method 500 according to an embodiment of the invention for confirming the output from the transmission state sensor 7 used to provide an indication of the engagement state of the transmission 3.

After starting, the method continues with the method step 100 in which odd gear and even gear neutral thresholds are determined for the transmission 3 as described above with respect to Figs.8A and 8B. As previously mentioned, the odd gear neutral threshold is the minimum safe signal level in the odd gear direction where neutral can be guaranteed to exist and the even gear neutral threshold is the maximum safe signal level in the even gear direction where neutral can be guaranteed to exist if the signal level goes high for even gears and low for odd gears.

After step 100 follows step 510 in which the transmission state module 5 monitors the output from the transmission state sensor 7. That is to say the signal from the transmission state sensor 7 is supplied to the transmission state module 5.

The odd gear and even gear neutral thresholds are stored in the transmission state module 5 and the signal being received from the transmission state sensor 7 is repetitively checked against these thresholds to determine whether it falls within these threshold limits as indicated by step 512. If the signal from the transmission state -22 -sensor 7 is between the odd gear neutral threshold and the even gear neutral threshold then this indicates that the transmission 3 is in neutral. After step 512 the method advances to step 520.

In step 520 it is checked whether the conditions for inferred neutral checking are present and, if they are, the method advances to step 522 otherwise the method loops through step 520 again to recheck whether the inferred neutral conditions are present.

Inferred neutral checking involves using the inputs from several other sensors 9 that can be used to establish whether a neutral state exists. For example, if the clutch pedal is sensed to be released so that the clutch is engaged, no faults related to the clutch pedal signal are known to exist, the engine speed exceeds a threshold, there are no known faults with the engine speed signal, the vehicle speed is below a threshold and no known faults related to the vehicle speed signal are present then the conditions for inferred neutral checking are present.

Then in step 522 it is checked whether the inferred neutral conditions are stable for a predetermined period of time such as, for example, 0.Sseconds. If the conditions are stable for the predetermined period of time then the method advances to step 524 otherwise the method returns to step 520. In step 524 it is determined whether an inferred neutral state exists before advancing to step 526 where the inferred neutral state from step 524 is compared to the transmission state determined in step 512 for a predetermined period of time such as, for example, 0.5 seconds. If at any time during the test period the output from the transmission state sensor 7 differs from the inferred neutral state while the inferred neural state remains stable then this indicates that the output from the transmission state sensor 7 is not currently valid and the -23 -method advances to step 540 where an error is indicated and then an error counter is incremented by one in step 552.

The method then advances to step 554 where it is checked whether the stored value is greater than a predetermined limit. If the counter value is above the limit then the method advances to step 560 and a transmission state sensor fault is indicated and automatic stop-start control using the transmission state sensor 7 will be inhibited.

However, if the counter is less than the limit the method returns to step 510 and the steps 510 to 526 will then be repeated the next time neutral is entered.

That is to say, the test is only run to completion (pass or fail) once each time the vehicle 1 is in neutral.

This is in order to prevent the error counter being run-up or run-down during a long idle period.

If at step 526 there is agreement between the inferred neutral state and the output from the transmission state sensor 7 for the duration of the test period then the test is passed and the method advances to step 528 where a confirmation of a valid sensor output is provided and the method advances to step 529 where the counter is decremented and then advances to step 554 discussed above.

Note that step 100 is a one time calibration step that once complete does not need to be repeated, the respective thresholds being stored in the transmission state module 5 for subsequent use.

Each time the test is run, which is once each time a neutral idle occurs, the test is treated independently, with no memory of the previous run. Each pass or fail of the test is then used to increment or decrement the fault -24 -counter and the fault counter is used to store the fault history' . All the time the fault counter is below a threshold the sensor output is considered good. When the fault counter exceeds the threshold the sensor output is considered to be faulty, a confirmed fault flag is raised, and the stop-start feature will disable stop-start.

However, neutral tests will continue to be run each time a neutral idle is detected and the fault bucket will be incremented or decremented appropriately. If the fault counter subsequently subsides below the threshold the confirmed fault flag is lowered but the stop-start feature will not re-enable until the next key-on.

Fig.12 shows an example of a logic table that is used in step 526 to determine whether the output from the transmission state sensor 7 is valid or invalid.

The first two rows show the output if a stable inferred neutral signal does not exist during the whole test period.

That is to say, the inferred state changes during the test period. In such a case the output from the transmission state sensor 7 is presumed to be valid because there is no independent means for checking it.

In the third row an inferred stable neutral exists for the test period but the output from the transmission state sensor 7 indicates that the state is not in neutral for at least part of the test period. In this case the output from the transmission state sensor 7 is considered to be invalid.

In the fourth row an inferred stable neutral exists for the test period and the output from the transmission state sensor 7 indicates that the state is in neutral for the duration of the test period. In this case the output from the transmission state sensor 7 is considered to be valid.

-25 -Referring now to Fig.7B there is shown a second part of the method 500 according to this invention for confirming the output from the transmission state sensor 7 used to provide an indication of the engagement state of the transmission 3.

After starting, the method continues with the method step 200 in which odd In-Gear and even In-Gear thresholds are determined for the transmission 3 as described above with respect to Figs.9A and 9B.

After step 200 follows step 510 in which the transmission state module 5 monitors the output from the transmission state sensor 7. That is to say the signal from the transmission state sensor 7 is supplied to the transmission state module 5 (This step is common to both Fig.7A and 7B because there is only one output from the transmission state sensor 7 that is monitored) The odd gear and even gear neutral thresholds are stored in the transmission state module 5 and the signal being received from the transmission state sensor 7 is repetitively checked against these thresholds to determine whether the transmission is sensed to be in an odd gear, an even gear or some other state as indicated by step 513.

After step 513 the method advances to step 530.

In step 530 it is checked whether the conditions for stable in-gear checking are present and, if they are, the method advances to step 532 otherwise the method loops through step 530 again to recheck whether the stable in-gear conditions are present.

Stable in-gear checking involves using the inputs from several other sensors 9 that can be used to establish whether an odd or even in-gear state exists. For example, if the clutch pedal is determined to be released and a known -26 -relationship is present between engine speed and vehicle speed, the vehicle speed exceeds a threshold, the engine speed exceeds a threshold, no faults related to the clutch pedal signal are known to exist, no faults with the engine speed signal are known to exist, no known faults related to the vehicle speed signal are present then the conditions for stable in-gear checking are present. It will be appreciated that from the relationship between engine speed and vehicle speed the current gear can be determined and therefore it is possible to determine whether the transmission 3 is in an odd gear or an even gear.

In step 532 it is checked whether the independent in-gear conditions are stable for a predetermined period of time such as, for example, 0.5seconds. That is to say, does an independent stable in-gear state exist. If the conditions are stable for the predetermined period then the method advances to step 534 otherwise the method reverts to step 530.

In step 534 the stable in-gear state (odd/ even) is determined before advancing to step 536 where the stable in-gear state from step 534 is compared to the transmission state determined in step 513 for a predetermined period of time such as, for example, 0.5 seconds. If at any time during the test period the output from the transmission state sensor 7 differs from the stable in-gear state while the stable in-gear state remains the same then this indicates that the output from the transmission state sensor 7 is not currently valid and the method advances to step 550 where an error is indicated and then an error counter is incremented by one in step 552.

The method then advances to step 554 where it is checked whether the stored value is greater than a predetermined limit. If the counter value is above the limit then the method advances to step 560 and a -27 -transmission state sensor fault is indicated and automatic stop-start control using the transmission state sensor 7 will be inhibited.

However, if the counter is less than the limit the method returns to step 510 and the steps 510 to 536 will then be repeated the next time a stable in-gear condition exists.

That is to say, the test is only run to completion (pass or fail) once each time the vehicle 1 is in a stable gear. The idea is that the test is not re-run until the driver selects another or the same gear. This is in order to prevent the error counter being run-up or run-down during a long idle period.

If at step 536 there is agreement between the inferred in-gear state and the output from the transmission state sensor 7 for the duration of the test period then the test is passed and the method advances to step 538 where a confirmation of a valid sensor output is provided and the method advances to step 539 where the counter is decremented and then advances to step 554 discussed above.

Note that step 200 is a one time calibration step that once complete does not need to be repeated, the respective thresholds being stored in the transmission state module 5 for subsequent use.

Each time the test is run, which is once each time a stable in-gear condition occurs, the test is treated independently, with no memory of the previous run. Each pass or fail of the test is then used to increment or decrement the fault counter and the fault counter is used to store the fault history' . All the time the fault counter is below a threshold the sensor output is considered good.

When the fault counter exceeds the threshold the sensor -28 -output is considered to be faulty, a confirmed fault flag is raised, and the stop-start feature will disable stop-start.

However, in-gear tests will continue to be run each time a stable in-gear condition is detected and the fault bucket will be incremented or decremented appropriately. If the fault counter subsequently subsides below the threshold the confirmed fault flag is lowered but the stop-start feature will not re-enable until the next key-on.

As described in greater detail with reference to Fig.18 the method further comprises executing the steps 510 to 536 until a valid result has been obtained for both odd and even gears.

In addition steps 552, 554 and 560 on Fig.7A and steps 552, 554 and 560 on Fig.7B may be common steps so that if the test is failed either by a mismatched neutral determination or a mismatched in-gear determination then the counter in step 552 is incremented by one.

Fig.13 shows an example of a logic table that is used in step 536 to determine whether the output from the transmission state sensor 7 is valid or invalid.

In the first row a stable odd in-gear state exists during the test period and the output from the transmission state sensor 7 also indicates that an odd gear has been selected. In such a case the output from the transmission state sensor 7 is presumed to be valid.

In the second row a stable odd in-gear state exists during the test period and the output from the transmission state sensor 7 indicates that a state other than an odd gear has been selected. In such a case the output from the transmission state sensor 7 is presumed to be invalid.

-29 -In the third row a stable even in-gear state exists during the test period and the output from the transmission state sensor 7 indicates that a state other than an even gear has been selected. In such a case the output from the transmission state sensor 7 is presumed to be invalid.

In the fourth row a stable even in-gear state exists during the test period and the output from the transmission state sensor 7 also indicates that an even gear has been selected. In such a case the output from the transmission state sensor 7 is presumed to be valid.

By using the method 500 validity checking of the transmission state sensor 7 to ensure that the sensor signal has not frozen in neutral can be performed because, if the transmission state sensor 7 has frozen in neutral the in-gear tests will be failed. It will be appreciated that, if the sensor signal has frozen in neutral, there is a risk that the stop-start controller 6 would think that the transmission 3 was in neutral, when in fact the driver might have selected a gear and subsequently an automated restart could be initiated causing unintended vehicle movement

Therefore in summary and with reference firstly to

Fig.10 the method comprises comparing the output from the transmission state sensor 7 with neutral thresholds to determine if a neutral gear is selected. The validity of the resulting neutral/not neutral indication is found by comparison with an inferred neutral' signal which indicates the neutral gear is selected if:-the clutch pedal is released, engine speed exceeds a threshold, vehicle speed falls below a threshold, there are no faults related to the clutch pedal signal and there are no faults related to the vehicle speed signal.

Once the inferred neutral signal is stable, that is to say it has unchangingly indicated neutral for a set period -30 -of time, it is compared with the neutral indication from the transmission state sensor 7. If the inferred neutral indicates neutral is selected, but the transmission state sensor 7 does not, then a neutral mismatch is indicated and latched. This latched signal is reset whenever the inferred neutral ceases to be stable. Once the inferred neutral signal is stable, a test period' timer starts. If the inferred neutral signal changes during this test period the test period' and stable period' timers are reset. Thus a stable inferred neutral signal is sought and is then compared with the transmission state sensor 7 neutral indication during a test period. At the end of the test period the latched neutral mismatch indication is examined and, if a mismatch occurred during the test period an instantaneous neutral fault is indicated.

Secondly, with reference to Fig.11, the transmission state sensor output is compared with thresholds to identify when an odd or even gear is selected. The validity of the resulting odd and even gear indications are determined by comparison with an independent estimated gear' signal which identifies the selected gear by comparing the engine speed, vehicle speed ratio with expected values for each gear.

This signal will be unusable for in-gear confirmation when the clutch is depressed or neutral gear is selected and so it is only used for validity checking of the transmission state sensor 7 when:-the estimated gear is not neutral, no faults related to the estimated gear signal or its inputs are present, the clutch pedal is released, no faults related to the clutch pedal state signal are present, the vehicle speed exceeds a threshold and engine load or propulsion torque exceeds a threshold.

Once the estimated gear is stable that is to say, it has unchangingly satisfied these conditions for a set stable period of time, it is compared with the transmission state sensor odd and even gear indications. If the estimated gear -31 -indicates an odd or an even gear, but the transmission state sensor 7 does not then an in-gear mismatch is indicated and latched. This latched signal is reset whenever the estimated gear ceases to be stable.

Once the estimated gear is stable, a test period timer is started and if the estimated gear signal changes during this test period the test period and stable period timers are reset. Thus a stable estimated gear signal is sought and then compared with the transmission state sensor odd/even gear indication during a test period.

At the end of the test period the latched in-gear mismatch indication is examined and, if a mismatch occurred during the test period an instantaneous in-gear fault is indicated.

Once a neutral or in-gear test has run to completion it is not run again until the inferred neutral or estimated gear respectively detects that the selected gear has changed and then been re-selected. At the end of each completed test period a neutral or in-gear fault integrator (error counter) is incremented if the instantaneous fault is present or decremented otherwise. When the fault levels exceed a threshold a confirmed neutral or in-gear fault is indicated. Also, a confirmed neutral fault is indicated if there is a fault with the clutch or vehicle speed signals and a confirmed in-gear fault is indicated if there is a fault related to the estimated gear.

Additionally, indications that the transmission state sensor 7 has been correctly active in the neutral, odd & even gear regions during the current key-cycle are obtained by detecting and latching that neutral, odd & even in-gear tests have completed with no mismatches. These latches are reset if a related fault occurs.

-32 -Although the invention has been described above in relation to a PWM magnetic transmission state sensor 7 in which zero offset learning is not normally required it will be appreciated that it could be applied with equal advantaqe whether or not zero offset learning has been used and to other types of transmission state sensor.

It will be appreciated that other inputs could be used to provide independent inferred neutral or stable in-gear values for comparison with the output from the transmission state sensor 7.

It will be appreciated that the signal output from the transmission state sensor could be an analogue voltage signal or a digital output.

Referring to Fig.14 there is shown the condition referred to above in relation to steps 104 and 107 on Fig.8B in which the neutral thresholds Pioddlhresh and PlEvenlhresh fall within the mechanical tolerance bands mechodd and mechEven (shown only as Imech) If it is not possible to meet the requirements to avoid end of line zero-offset learning, it is possible to learn out the measured offset between the transmission neutral resting position and the 50% sensor signal position which is nominally zero degrees rotation of the selector cylinder 3A.

That is to say, that if PlEvenlhresh > Tmecven and PlOddlhresh > Tmechodd respectively (these equations relate to distance from the zero or normal resting position of the selector cylinder 3A and in signal magnitude terms the equations become if PlEvenlhresh < TmechEven and PloddThresh > Tmechodd respectively) then zero offset learning is required and this situation is shown in Fig.14.

In this case the transmission 3 can be legitimately in neutral and is within the bounds of Tmech but is in the -33 -undetectable part of the neutral zone. In this case it would not be possible to inform the stop-start controller 6 that the transmission 3 is in neutral and hence stop-start Figure 15 shows a raw sensor signal and an adjusted signal after end of line zero-offset learning for a worst case transmission, which is one where the transmission 3 has a neutral resting position at mech The transmission state module 5 will check the adjusted sensor signal against the calibrated neutral and in-gear thresholds and will ignore the raw sensor signal.

If end of line learning has not been carried out correctly, for example if an operator's hand is resting on the gear stick during the learning process, a permanent zero-offset error would result. Furthermore, if the transmission 3 is replaced in-service and the zero-offset learning process is not carried with the newly installed transmission, the offset stored in the transmission state module will be wrong. Fig.16 shows the worst possible error that can arise which corresponds to a worst-case transmission. A worst case transmission is a transmission with the largest possible selector cylinder 3A rotational zero-offset when in its neutral resting position that is subsequently replaced with a transmission that has a largest possible selector cylinder 3A zero-offset in the opposite direction. This would result in an error of magnitude 2Tmech.

As can be seen in Fig.16 where the incorrectly adjusted' values represent a transmission that was zero adjusted for a worst case error in one direction and has been replaced by a transmission that requires to be zero adjusted to account for a worst case transmission mechanical tolerance in the opposite direction as indicated by the adjusted' values, if the first transmission is replaced with the second transmission without re-learning the zero -34 -offset then in the odd gear direction the transmission state sensor 7 would indicate that the state of the transmission is between the IGoddlhresh and the resting position whereas in fact the transmission 3 is in-gear as indicated by the adjusted value at IGoddlhresh. Therefore, if a sensor validity test were to be conducted prior to zero offset learning for the new transmission, the odd gear validity test would be failed because the independent stable indication of transmission state would indicate that the transmission 3 is in the odd gear state but the output from the transmission state sensor 7 would indicate that it is not in the odd gear state as the signal has a higher signal value than IGoddlhresh.

However, in the even gear direction the validity test would be passed because the incorrectly adjusted value falls further in the even in-gear zone and has a signal value higher than IGEvenThresh indicating that the transmission is in an even gear state. Therefore, in order to confirm that a transmission is incorrectly adjusted either due to incorrect zero-offset learning having taken place or due to the fact that the transmission has been replaced with a transmission requiring a different zero offset, it is not sufficient to carry out a validity test for only one of the odd and even gear states. Both odd gear and even gear validity testing must be carried out to identify such a problem.

Fig.17 shows the situation where the adjusted sensor signal from a previous transmission is located inside the neutral window but the correct offset for the current transmission is located past the pull-in threshold. This could lead to a situation in which the transmission state module 5 interprets the output from the transmission state sensor 7 as an indication that the transmission 3 is in neutral but the actual status is that the transmission 3 is not in neutral thus exposing the stop-start controller 6 to potentially allowing unintentionally movement of the vehicle 1.

-35 -In order to prevent this from happening, and as referred to above with reference to Fig.16, the solution is to use the in-even or in-odd gear validity checking as described with respect to Figs 7B, 11 and 13 above to check that both in-even and in-odd gear validity checks have taken place before enabling stop-start to use the gear neutral signal for automatic stop-start operation. The reason for this is highlighted in figure 16 which shows in the first instance a correctly learnt zero-offset (adjusted) at both the odd-gear and even-gear ends of the range. Here both the in-odd gear and in-even gear zones are reached by the adjusted signal. This means that validity checking will be successful for both even and odd gears prior to stop-start being activated. However, for the case that the zero-offset is wrong (incorrectly adjusted), only one of the in-even or in-odd gear zones will be reached by the adjusted sensor signal which, in this case, is the in-even gear zone but could be the in-odd gear zone if the opposite erroneous calibration were to be present. The consequence of this is that the stop-start controller 6 is not allowed to trust the gear neutral sensor signal and hence stop-start is never activated, obviating the risk of unintended vehicle movement.

A method for implementing this strategy is shown in Fig.18. After starting, which may be a vehicle key-on event, the first step 810 is to carry out validity testing of the transmission state sensor 7 as described above with reference to Figs. 7A, 7B, 10, 11, 12 and 13.

That is to say, after key-on of the vehicle 1, a neutral validity test is carried out the first time a stable inferred neutral condition exists for a short predetermined period of time such as a few tenths of a second.

Provided the inferred neutral condition continues for a further period of time referred to as a test period in which -36 -the output from the transmission state sensor 7 is compared to the inferred neutral condition then, if there is agreement between the output from the transmission state sensor 7 and the inferred neutral, the neutral validity test is said to be passed and, if the output from the transmission state sensor 7 differs from the stable inferred neutral condition at any point during the test period, the neutral validity test is said to have been failed.

Although neutral validity testing is not required for determining whether a dangerous zero offset condition is present it is required to confirm that a neutral output from the transmission state sensor 7 is valid as it is usual to use an indication of whether neutral exists provided by the transmission state sensor 7 as a primary parameter for controlling stop-start operation of the engine 2. It is therefore normal practice to carry out neutral validity testing whenever a key-on event occurs.

The transmission state module is also operable to carry out an in-odd gear validity test the first time after key-on that a stable independent indication of the selection of an in-odd gear state exists for a predetermined of time, such as a few tenths of a second, and an in even-gear validity test the first time after key-on that a stable independent indication of the selection of an in-even gear exists for a predetermined of time, such as a few tenths of a second.

Provided the respective stable independent indication of transmission state continues unchanged for a further short predetermined period of time referred to as a test period in which the output from the transmission state sensor 7 is compared to the stable independent indication of transmission state, then, if there is agreement between the output from the transmission state sensor 7 and the respective stable independent indication of transmission state, the respective in-gear validity test is said to be -37 -passed, and if the output from the transmission state sensor 7 differs from the stable independent indication of transmission state at any point during the test period, the respective in-gear validity test is said to have been failed. Note that both in-odd gear and in-even gear validity tests have to be conducted for the method to advance past step 820.

In step 820 it is checked whether the validity tests for both the in-odd gear state and in-even gear state have been carried out at least once, if they have then the method advances to step 830 otherwise the method returns to step 810 and will cycle around steps 810 and 820 until the validity tests for both the in-odd gear state and in-even gear state have been carried out. This check can be carried out by setting respective flags when each type of validity test has been performed and checking the status of the flaqs in step 820.

In step 830 it is confirmed whether both of the validity tests for the in-odd gear and in-even gear states have been passed. As before, flags can be set to indicate whether a validity test has been passed or failed, and these flags are checked in step 830 to determine their respective states.

If one of the tests has been failed, then it is not safe to use the output from the transmission state sensor 7 to automatically control start-stop operation of the micro-hybrid vehicle 1, and the method advances to step 850 where automatic stop-start control of the engine 2 is prevented by the stop-start controller 6. That is to say, the transmission state module 5 sends a signal to the stop-start controller 6 inhibiting automatic starting or stopping of the engine 2. It will however be appreciated that this does not prevent manual stopping and starting of the engine 2 by an operator of the vehicle 1. It will further be -38 -appreciated that the signal could be in any suitable form and could include, for example, the setting of a flag in the stop-start controller 6.

After step 850 the method could end or, as shown, the method could continue to step 860 where a warning is provided to a user of the vehicle 1 that automatic startinq and stopping has been prevented. This warning could be a simple warning lamp or could be an alpha-numeric output providing script with or without an error code.

Referring back to step 830, if it is confirmed that both of the validity tests for the in-odd gear and in-even gear states have been passed then it is safe for the stop-start controller 6 to use the output from the transmission state sensor 7 to control automatic starting and stopping of the engine 2 and the method advances to step 840 where automatic stop-start control is enabled.

Although the invention has been described with respect to the use of a PWM magnetic sensor or PLCD (Permanent Magnet Linear Contactless Displacement) sensor which uses a magnet and generates a PWM output for the transmission state sensor (sometimes referred to as a LVDT sensor), it will be appreciated that other types of displacement sensor could be used such as, for example, a Hall-Effect sensor which still uses a magnet and generates a PWM output. Furthermore, the invention is not limited to the use of sensors producing a PWM output; it is equally applicable for use with a displacement sensor which generate a variable voltage output instead of a PWM output signal.

Claims

-39 -Claims 1. A method for controlling automatic starting and stopping of the engine of a vehicle, the vehicle having an engine driving a manual transmission and a sensor to provide an indication of an engagement state of the transmission by sensing the position of a gear selector, wherein the method comprises preventing automatic starting and stopping of the engine if an output from the sensor has not passed a validity check for both an odd gear selector position and an even gear selector position.
2. A method as claimed in claim 1, wherein the method further comprises indicating to a user of the vehicle when automatic stopping and starting is being prevented.
3. A method as claimed in claim 1 or 2, wherein the validity check for an odd gear selector position is not passed if the output from the sensor disagrees with an independent indication of an odd gear transmission state for a test period of predetermined duration, provided that the independent indication of an odd gear transmission state remains the same for the duration of the test period.
4. A method as claimed in claim 3, wherein the validity check for an odd gear selector position is passed if the output from the sensor indicates that the state of the transmission is in an odd gear for the test period and the independent indication that the transmission is in an odd gear state persists for the test period.
5. A method as claimed in claim 3 or 4, wherein the validity check for an odd gear selector position further comprises checking agreement of the independent indication of an odd gear transmission state with the output from the sensor after the independent indication of an odd gear -40 -transmission state has been stable for a predetermined period of time.
6. A method as claimed in any of claims 1 to 5, wherein the validity check for an even gear selector position is not passed if the output from the sensor disagrees with an independent indication of an even gear transmission state for a test period of predetermined duration, provided that the independent indication of an even gear transmission state remains the same for the duration of the test period.
7. A method as claimed in claim 6, wherein the validity check for an even gear selector position is passed if the output from the sensor indicates that the state of the transmission is in an even gear for the test period and the independent indication that the transmission is in an even gear state persists for the test period.
8. A method as claimed in claim 6 or 7, wherein the validity check for an even gear selector position further comprises checking agreement of the independent indication of an even gear transmission state with the output from the sensor after the independent indication of an even gear transmission state has been stable for a predetermined period of time.
9. An apparatus for controlling automatic starting and stopping of the engine of a vehicle, the vehicle havinq an engine drivingly connected to a manual transmission having a selector, the position of which determines whether the transmission is in one of an odd gear, an even gear and neutral, a sensor to monitor the position of the selector, and a transmission state module to receive a signal from the sensor and provide an output signal to a stop-start controller, wherein the transmission state module is operable to prevent automatic starting and stopping of the -41 -engine if the sensor output has not been confirmed to be valid for either or both of an odd gear selector position and an even gear selector position.
10. An apparatus as claimed in claim 9, wherein the sensor output is not confirmed to be valid for an odd gear selector position if the sensor output disagrees with an independent indication of an odd gear transmission state for a predetermined period of time, and the sensor output is not confirmed to be valid for an even gear selector position if the sensor output disagrees with an independent indication of an even gear transmission state for a predetermined period of time.
11. An apparatus as claimed in claim 9 or 10 wherein the transmission state module is operable to receive inputs from a number of sensors located on the vehicle indicating the current operating state of the vehicle and use the inputs from two or more of the sensors to determine each independent indication of transmission state.
12. A vehicle comprising an apparatus as claimed in any of claims 9 to 11.
13. A method for indicating the engagement state of a manual transmission substantially as described herein with reference to the accompanying drawings.
14. An apparatus for controlling the operation of a micro-hybrid vehicle substantially as described herein with reference to the accompanying drawings.