WO2014177380A1

WO2014177380A1 - Method for predicting instability in a vehicle-trailer combination

Info

Publication number: WO2014177380A1
Application number: PCT/EP2014/057723
Authority: WO
Inventors: Giovanni Strano
Original assignee: Jaguar Land Rover Limited
Priority date: 2013-05-01
Filing date: 2014-04-16
Publication date: 2014-11-06
Also published as: GB2513616A; GB2513616B; EP2991867A1; GB201307886D0

Abstract

A method of predicting instability in a vehicle-trailer combination, the method comprising: monitoring a vehicle direction related parameter; monitoring a trailer direction related parameter; comparing the vehicle direction related parameter with the trailer direction related parameter to determine a response of the trailer direction to the vehicle direction over time in order to define a response trend; and determining an instability event in dependence on a change in the response trend.

Description

Method for predicting instability in a vehicle-trailer combination

TECHNICAL FIELD

The present invention relates to a method for predicting instability in a vehicle-trailer combination and particularly, but not exclusively to a vehicle-trailer combination comprising a motor vehicle towing a trailer attached to the vehicle at a hitch point. Aspects of the invention relate to a method, a control module, an engine control unit, a system and a vehicle. BACKGROUND

When a vehicle, such as a car, van or truck, tows a trailer there is a tendency for the trailer to oscillate or "fish-tail" behind the vehicle. This can be particularly problematic at high road speeds and during critical manoeuvres, such as lane changes and sharp cornering. If left unchecked such oscillations can be dangerous and are a common source of road accidents.

Vehicle-trailer combinations typically act as damped systems in which the magnitude of a damping co-efficient associated with the system dictates the rate at which oscillations of the trailer decay after it has been displaced from a neutral position behind the towing vehicle, e.g. by a gust of wind. In such a system, the greater the vehicle's road speed, the lower the damping co-efficient of the vehicle-trailer system becomes. Thus, the system becomes increasingly unstable as velocity increases. This means that the faster the vehicle travels, the greater the tendency for dangerous and uncontrollable trailer oscillations to occur. This problem is exacerbated by the fact that, in recent years, increasingly stringent vehicle emissions regulations have resulted in a decrease in average vehicle weight and studies have shown that this decrease in vehicle weight has had an adverse effect on vehicle stability, particularly when towing a trailer.

Many vehicles are provided with Electronic Stability Control (ESC) systems (also known as Electronic Stability Programs (ESP) and Dynamic Stability Control (DSC)) which help to improve vehicle stability. Such systems typically detect excessive vehicle yaw using gyroscopic sensors or similar arrangements, and then counteract this by selectively braking individual wheels of the vehicle. When such a vehicle tows a trailer, ESC can help to improve the stability of the vehicle-trailer system in a similar fashion. However, due to the greatly increased instability and weight of the vehicle-trailer system (relative to the vehicle on its own), a greater magnitude and frequency of braking must be applied to maintain stability, which has a tendency to prematurely wear and overheat the vehicle's brakes. Furthermore, controlling trailer instability using this method relies on the trailer oscillating to such an extent that the movement of the trailer begins to cause the vehicle to yaw. If the trailer is lightweight relative to the vehicle, this may not occur until the trailer is oscillating at an undesirable level which may cause damage to the trailer and/or its load. Furthermore, in the case of a relatively light trailer, oscillations which may cause damage to goods being transported in the trailer may occur without the driver of the towing vehicle being able to detect through the steering wheel that such oscillations are occurring. Also, it may be much harder to rectify the problem if trailer oscillation builds up too much prior to detection. This is a particular problem at high road speeds, where the oscillations of the trailer may take longer to reach such a magnitude that they induce sufficient vehicle yaw so as to cause the ESC to apply the vehicle's brakes. In this case, by the time the ESC is activated it may be too late to take sufficient corrective action, with the result that a critical instability in the vehicle-trailer system develops, causing a road accident.

More recently it has been proposed to address this problem by measuring trailer yaw directly, using a camera mounted to the rear of the vehicle for example. This arrangement allows for an earlier response to trailer oscillations reaching undesirable levels. However, although such arrangements offer an immediate response to trailer oscillation becoming unstable, it would be preferable to offer the ability to predict the development of an instability problem before it occurs.

Therefore, against this background it is an aim of the present invention to provide an improved system for prevention of trailer instability.

SUMMARY OF THE INVENTION

According to an aspect of the invention there is provided a method of predicting instability in a vehicle-trailer combination, the method comprising monitoring a vehicle direction related parameter and monitoring a trailer direction related parameter. The method further comprises comparing the vehicle direction related parameter with the trailer direction related parameter to determine a response of the trailer direction to the vehicle direction over time in order to define a response trend, and determining an instability event in dependence on a change in the response trend. In embodiments, the method advantageously detects trailer instability as soon as it starts to develop, and in advance of a critical instability event. The method provides a computationally simple means for detecting trailer instability which relies on detection of only two parameters in real time. The method is therefore a far simpler alternative to modelling the vehicle-trailer system as a whole. This enables, for example, an alarm to be generated in advance of critical instability, which allows time for corrective action to be taken by either the driver or by automatic processes, to prevent critical instability from developing which may otherwise result in a loss of control of the vehicle. For example, the alarm may be used to activate an ESC system in order to prevent a critical instability event.

The method may comprise outputting an alarm signal in dependence on making a positive determination of an instability event. In this embodiment, the method may comprise outputting the alarm signal to a driver of the vehicle to indicate the onset of a trailer instability.

Comparing the vehicle direction related parameter with the trailer direction related parameter may comprise defining an artificial vector based on the vehicle direction related parameter and the trailer direction related parameter.

Determining the response of the trailer may comprise calculating the argument of the vector.

Defining a response trend may comprise monitoring the argument of the vector over time.

The instability event may be determined as a result of an unexpected change of direction of the argument.

Conveniently, the vehicle direction related parameter may be the steering wheel angle.

The trailer direction related parameter may be the trailer yaw. A vehicle brake may be activated in response to an instability event being determined. A vehicle mounted camera may be used to monitor the trailer direction related parameter. According to another aspect of the invention, there is provided a control module for predicting instability in a vehicle-trailer combination. The control module comprises means to receive a signal indicative of a vehicle direction related parameter, and means to receive a signal indicative of a trailer direction related parameter. The control module further comprises processing means arranged to compare the vehicle direction related parameter with the trailer direction related parameter, determine a response of the trailer direction to the vehicle direction over time in order to define a response trend, and determine an instability event in dependence on a change in the response trend. According to a further aspect of the invention, there is provided a system for predicting instability in a vehicle-trailer combination, the system comprising a control module according to the second aspect, at least one sensor arranged to monitor the vehicle direction related parameter, and at least one sensor arranged to monitor the trailer direction related parameter.

The at least one sensor arranged to monitor the vehicle direction related parameter may comprise a steering wheel angle sensor. Alternatively, the at least one sensor arranged to monitor the vehicle direction related parameter may comprise a vehicle yaw sensor. The at least one sensor arranged to monitor the trailer direction related parameter may be a vehicle mounted camera.

The invention also extends to an engine control unit comprising a control module according to the second aspect, a vehicle comprising a control module according to the second aspect, and a vehicle comprising a system according to the third aspect. The invention further extends to a vehicle comprising said engine control unit.

Within the scope of this application it is expressly intended that the various aspects, embodiments, examples and alternatives set out in the preceding paragraphs, in the claims and/or in the following description and drawings, and in particular the individual features thereof, may be taken independently or in any combination. Features described in connection with one embodiment are applicable to all embodiments, unless such features are incompatible. BRIEF DESCRIPTION OF THE DRAWINGS Embodiments of the present invention will now be described, by way of example only, with reference to the accompanying drawings, in which like components are assigned like numerals, and in which:- Figure 1 is a schematic plan view diagram of a vehicle-trailer combination in a neutral position;

Figure 2 is a schematic plan view diagram of the vehicle-trailer combination of Figure 1 in a trailer yaw position;

Figure 3 is a schematic illustration of an image viewed by a rearward facing vehicle-mounted camera where the trailer is in a neutral position;

Figure 4 is a schematic illustration of an image viewed by the rearward facing camera where the trailer is in a yawed position;

Figure 5 is a graph plotting steering wheel angle against trailer yaw during an oscillation;

Figure 6 is an illustration of a set of results from a simulation of a vehicle-trailer combination traversing a slalom in an open-loop scenario;

Figure 7 is an illustration of a set of results from a simulation of a vehicle-trailer combination traversing a slalom in a closed-loop scenario;

Figure 8 is an illustration of a set of results from a simulation of a vehicle-trailer combination traversing an adverse road in an open-loop scenario;

Figure 9 is a flow diagram illustrating a method for predicting trailer instability according to an embodiment of the present invention;

Figure 10 is an illustration of a set of results from a test of the method in Figure 9; Figure 1 1 is a schematic illustration of a control module which may be used to implement the method in Figure 9; and

Figure 12 is an illustration of a vehicle having the control module in Figure 1 1 .

DETAILED DESCRIPTION

Referring to Figure 1 , a vehicle-trailer combination 10 comprises a vehicle 12 towing an attached trailer 14. In Figure 1 , the vehicle-trailer combination 10 is in a neutral state where the longitudinal axis L_v of the vehicle 12 is substantially aligned with the longitudinal axis L_T of the trailer 14. In this condition the yaw angle, Θ, of the trailer 14 relative to the vehicle 12 is at or near 0°. Referring to Figure 2, as the vehicle 12 tows the trailer 14, there may be a tendency for the trailer 14 to yaw away from the neutral position represented in Figure 1 thereby resulting in a yaw angle Θ which is greater than 0°. This tendency to yaw may result from many factors, for example, the dynamics of lane-change manoeuvres, wind gusts, sharp cornering, adverse cambers etc. Furthermore, in such damped systems, once the trailer 14 has departed from its neutral position, it will normally tend to return back towards the neutral position as the vehicle-trailer combination 10 moves forwards. However, once the trailer 14 reaches the neutral position its momentum will also cause it to pass through the neutral position in a pendulum effect which results in a swaying motion of the trailer 14 relative to the vehicle 12. The rate at which the trailer 14 moves between a yawed position on one side of the vehicle 12 to an opposite yaw position on the other side of the vehicle 12 is the yaw frequency.

Movement of the trailer 14 relative to the towing vehicle 12 may be determined by any suitable means. For example, relative movement of the trailer 14 may be sensed by optical sensing means such as a rearward facing camera mounted on the vehicle 12 for example. In the case that the vehicle 12 is provided with one or more rearward facing parking assist cameras, these may be employed for the purpose of sensing movement of the trailer 14, advantageously obviating the need for an additional camera. Additionally, in the present embodiment, the relative movement of the trailer 14 may also be detected using ultrasonic sensors, which may conveniently be existing park-assist sensors mounted on the rear of the vehicle 12. Thus, advantageously, the trailer yaw angle Θ, i.e. the deviation of the trailer 14 away from the neutral position, can be detected by contactless sensors on-board the vehicle 12. This has the advantage that the sensors are not subjected to any mechanical wear and tear. In further embodiments of the present invention, other types of sensing means may be employed to detect the relative movement between the vehicle 12 and trailer 14 e.g. radar and/or a mechanical arrangement provided at the trailer hitch point such as an encoder wheel arrangement. Alternatively, or in addition, movement of the trailer 14 may be sensed directly by a yaw sensor, for example an accelerometer or a gyroscope, mounted on the trailer and the sensed yaw movement communicated to the vehicle 12.

Figures 3 and 4, illustrate how the dimensions of the trailer 14 may be used in a method of determining the yaw angle of the trailer's longitudinal axis L_T relative to the vehicle's longitudinal axis L_v in an embodiment where a rearward facing camera mounted on the vehicle 12 is used to determine trailer yaw. A registration process may be first undertaken by the driver in order to ascertain the dimensions of the trailer edges as viewed by the camera. Although these relative dimensions may be input manually into the system by the driver if desired, the camera can instead be used to register these dimensions automatically based on the image viewed. In the embodiment shown, the dimensions registered are edges A, B, C and D of the trailer 14; however, it will be appreciated that any other dimensions of trailer 14 could instead be registered as desired.

As shown in Figure 4, as the trailer 14 departs from the neutral position in a yawing movement, the dimensions of edges A', B', C, D' of the trailer 14 viewed by the camera will change relative to those that were viewed by the camera when the trailer 14 was in the neutral position. This change in the relative dimensions viewed by the camera can be readily correlated to the change in yaw angle Θ of the trailer 14 relative to the vehicle 12. This effect is amplified by the proximity of the rearward facing camera to the trailer 14 since even relatively small angular movements will result in significant changes in the relative dimensions of the trailer 14 perceived by the camera. For example, as the yaw angle of the trailer 14 increases, the upper and lower side edges B', D' on the right-hand side of the trailer 14 (as viewed by the camera) appear to be lengthened as compared to how they appear in Figure 3. Similarly, the corresponding upper and lower side edges on the left-hand side of the trailer 14, appear shortened in Figure 4 compared to how they appear in Figure 3.

The above method of determining the trailer yaw angle is based upon a comparison of relative perceived dimensions of the trailer 14. Therefore, it may not be necessary to perform the registration when the trailer 14 is in the neutral position but it may instead be possible to perform the registration procedure when the trailer 14 is at any initial angle relative to the vehicle 12.

In this arrangement more than one camera may be provided, and in addition to the camera(s), the vehicle 12 is also provided with additional sensors, such as ultrasonic sensors, which also detect a value for the yaw angle Θ of the trailer 14. In the case that the ultrasonic sensors are existing park-assist sensors mounted on the vehicle 12, such sensors are typically deployed at spaced locations across the rear of the vehicle 12, e.g. at optimal positions along a rear bumper of the vehicle 12. Each ultrasonic sensor may therefore be used to detect changes in the distance to an adjacent portion of the trailer 14 as the trailer 14 oscillates from side to side about the neutral position. In more detail, as the trailer 14 yaws to one side, an ultrasonic sensor disposed on one side of the hitch point detects that an adjacent portion of the trailer 14 is getting closer to the vehicle 12, whereas an ultrasonic sensor disposed on the opposite side of the hitch point detects the trailer 14 is getting further away. When the trailer 14 yaws in the opposite direction, the reverse is true.

The respective signals output from the camera and the ultrasonic sensors each provide a means for measuring the yaw angle of the trailer 14 and these measurements of the yaw angle are optionally input to processing means which performs filtering and analysis of the raw yaw outputs from the sensors.

As the vehicle-trailer combination 10 moves forwards and undergoes various manoeuvres, for example following bends in a road, there is a delay each time between a demand for the vehicle 12 turning and the trailer yaw angle increasing as explained below.

The demand for the vehicle 12 to turn may be defined by the driver turning a steering wheel, for example. When the driver turns the steering wheel, the wheel is turned to a maximum angle of the turn initially (although not necessarily full-lock), and then gradually returned to a neutral position which is defined when the front wheels of the vehicle 12 are aligned with L_v and the steering wheel angle is 0° as the vehicle 12 approaches the desired direction of travel. In contrast, the trailer yaw angle builds gradually to a maximum as the vehicle 12 turns, with the trailer 14 returning to alignment with the vehicle 12 a short time after the vehicle 12 settles into the desired direction of travel. Thus, there is effectively a delay between a change in steering wheel angle and a change in trailer yaw angle, such that the trailer yaw angle is effectively a response to the steering wheel angle.

If the vehicle-trailer combination 10 is subjected to conditions which move the vehicle 12 away from the desired direction of travel, such as a sudden gust of wind or a bump in the road, this may cause both the vehicle 12 and the trailer 14 to move in an oscillatory manner as the driver tries to recover the correct direction of travel of the vehicle 12. Assuming the driver is able to recover the vehicle 12 and return the vehicle-trailer combination 10 into stable alignment, such movement of the vehicle 12 may be represented as a damped oscillation, as the vehicle 12 overshoots the correct driving direction by less each time and gradually settles back into the correct driving direction. It is noted that the overshoot of the vehicle 12 may be exacerbated by the movement of the trailer 14 behind it, particularly if the trailer 14 is relatively heavy with respect to the vehicle 12.

One way in which this behaviour may be represented is by recording the steering wheel angle throughout this period. The steering wheel angle oscillates between, for example, 45° and -45°, as the driver turns from side to side to try to recover control of the vehicle 12. Eventually, the steering wheel angle will settle at 0° when the vehicle 12 is travelling in the correct direction. As outlined above, the direction of travel of the trailer 14 follows a similar path to that of the vehicle 12, albeit with a delay. Thus, if both steering wheel angle and trailer yaw angle are plotted together on a graph to show their changing values over time, the graph takes the form of a pair of decaying oscillation signals which are slightly out of sync with one another, i.e. with a signal representing trailer yaw angle lagging behind a signal representing steering wheel angle.

However, if the movement of the trailer 14 builds to the point where the trailer yaw becomes unstable, there is a change in the relationship between the steering wheel angle and the trailer yaw angle; the trailer yaw angle no longer maintains a consistent trend behind the steering wheel angle. Instead, the delay may either increase, indicating that the magnitude of the oscillations of the trailer 14 have begun to exceed that of the oscillations of the vehicle 12, or the delay may decrease, indicating that the trailer 14 is beginning to oscillate faster than the vehicle 12. Alternatively, a change in the relationship between the two variables may indicate that the steering wheel angle has started to change more rapidly, indicating that the driver is turning more vigorously in an effort to control the vehicle-trailer combination 10. This change in the relationship between the steering wheel angle and the trailer yaw angle, is explained more fully below with reference to figure 5.

It is noted that such a change in relationship between the steering wheel angle and the trailer yaw angle may therefore be used as an indicator that movement of the trailer 14 has become, or is about to become, unstable. Therefore, in an embodiment of the present invention, an alarm is optionally activated upon detection of a change of relationship, to alert the driver to the fact that the movement of the trailer 14 is about to become unstable. In a further refinement of this system, the alarm may be used to activate the ESC system of the vehicle 12 in order to regain control. Figure 5 illustrates the form of an artificial vector which is used to represent the relationship between the steering wheel angle and the trailer yaw angle. The two variables are represented in Cartesian coordinates in the form of v = x+iy, where x is the trailer yaw angle, y is the steering wheel angle, and v is the artificial vector which represents the relationship between x and y. The angle a which is subtended by v and the x-axis, representing the argument of v, is referred to as the phase of v. Therefore, the value of a is calculated as arctan(y/x).

The value of a shown in the Figure 5 indicates that the steering wheel angle and the trailer yaw angle are both positive and non-zero, and the steering wheel angle is smaller than the trailer yaw angle. This suggests that the vehicle 12 has recently commenced a turning manoeuvre, and the steering wheel angle is decreasing while the trailer yaw angle is increasing, in the manner described previously. Hypothetically, if the vehicle-trailer combination 10 were to oscillate steadily and continuously, the intersection of x and y would follow the circular path indicated by the dashed line 15 in a clockwise direction. In this scenario, the value of a would decrease steadily from 90° down to 0°, and then when v crosses the x-axis the sign of the angle would switch. From this point the value of a would continue to decrease from 0° to -180°. The sign of the angle would then switch again, and the value of a would decrease from 180° to 90° and hence the process would start again. In this way, a monotonically decreases during steady oscillation. In this way, the value of a at all times is calculated to enable the phase graphs of Figures 6, 7 and 8 to be constructed.

In reality, as mentioned previously the vehicle-trailer combination 10 behaves as a damped system, thus oscillations are not steady and continuous. Therefore, rather than following a circular path, when trailer movement is stable the vector v spirals gradually inwards towards the origin. Furthermore, the phase of v changes at a substantially constant rate. When v reaches the origin, both x and y are 0, indicating that the trailer 14 and the vehicle 12 are in alignment. Therefore, the origin represents a point of stability, and a spiral which tends towards the origin therefore indicates a system which is moving towards stability. If, however, the movement of the trailer 14 becomes unstable, this manifests as a change in direction in the movement of v. As illustrated in Figures 6, 7 and 8, this manifests in a reversal in the trend of a phase vector graph line, as explained later with reference to those figures. In the graph of Figure 5 this change manifests as a point of inflection, i.e. a reversal in the path of vsuch that the spiral doubles back on itself. Figures 6 to 8 illustrate a series of simulations of the above described behaviour of the vehicle-trailer combination 10, indicating the points at which an alarm is activated. In Figure 6, a set of results of a first simulation of a vehicle-trailer combination 10 are shown. In the first simulation, the oscillating behaviour is induced by sending the vehicle-trailer combination 10 through a slalom of cones in an open-loop scenario, i.e. with no intervention by the ESC of the towing vehicle to take corrective action. As shown in the uppermost graph 16 of the figure, the trailer yaw signal follows the steering wheel angle signal as expected. For example, the first minimum for the steering wheel angle appears at -25 seconds, whereas the first minimum for the trailer yaw angle is at around 27 seconds. The two signals oscillate in this manner until at -33 seconds the trailer 14 begins to become unstable, which manifests as the steering wheel angle rising sharply in response.

Below this graph, the middle graph 18 shows the results of the upper graph after the two signals have been combined into an artificial phase vector, as explained previously with reference to Figure 5. As shown in Figure 6, there is initially a stable period of operation, in which the phase vector monotonically decreases. This trend changes suddenly at 33 seconds, as indicated by the dashed line 20, when the direction of the phase vector reverses. This illustration clearly demonstrates that the point at which the system becomes unstable coincides with the point at which the phase vector which is used to represent the system changes direction. Accordingly, an alarm is activated immediately after the phase changes direction, as shown in the lowermost graph 22. As the system is open loop, the alarm has no effect, and therefore the vehicle 12 does not regain control.

Figure 7 shows a similar set of results to those in Figure 6, with the difference being that the simulation was altered such that the vehicle was arranged to brake and decelerate in response to an alarm in order to counteract trailer instability. As such the system of Figure 7 is a closed-loop system. This is evident in the graphs (24, 26), as it is clear that although an instability develops in a similar manner to the previous simulation (at around 33 seconds), the vehicle 12 regains control and returns to stable oscillation by 40 seconds. As seen in the upper graph 24, at the point 25 when the instability initially develops, the steering wheel angle starts to fluctuate with increasing amplitude as a result of the driver of the vehicle turning the steering wheel more vigorously to try to regain control. However, this is quickly controlled so that the oscillations return to the initial level by 40 seconds. The middle graph 26 shows that in fact there are at least two phase changes in the artificial vector between 33 seconds and 40 seconds, as indicated by the two dashed lines 28, 30, which may be as a result of the braking. Accordingly, the lower graph 32 indicates that the alarm is activated twice 33a, 33b.

Figure 8 relates to an enhanced adverse road simulation, in which the vehicle-trailer combination 10 is subjected to a series of adversities, denoted as (a) through to (e), which are defined as follows: (a) a pair of sidewind gusts; (b) a stretch of wavy or bumpy terrain; (c) a beam on the road accompanied by a pair of sidewind gusts; (d) a single sidewind gust; and (e) a stretch of bumpy terrain accompanied by a single sidewind gust. In addition to this, the vehicle-trailer combination 10 travels around a 90° corner between 65s and 75s. It is noted that after the vehicle-trailer combination 10 travels around the bend, the trailer yaw value settles at 90°. This is because in this simulation the trailer yaw angle is measured relative to a fixed global axis, rather than relative to the vehicle 12, in accordance with ISO standards. Hence, the trailer yaw angle settles at 90° from 75s onwards, because the trailer 14 (as well as the vehicle 12) is travelling at 90° relative to its initial position.

In the simulation the speed of the car was chosen such that an instability would be induced in each of events (a) to (e), which is evident in the upper graph 34 from the sudden changes in the amplitude of the steering wheel angle signal (peaks generally indicated as 35, for example). The phase vector results in the middle graph 36 demonstrate that an instability results in a change of phase regardless of the cause of instability, as indicated by the set of dashed lines 38, 40, 42, 44, 46. Therefore, as shown in the lower graph 48, the alarm is triggered for every adverse event. With reference to Figure 9, an embodiment of a process 52 for predicting trailer instability is illustrated. In this embodiment, the process 52 is conducted by an on-board computer of the vehicle 12, although in other embodiments a dedicated control module (for example as shown in Figure 1 1 , which is described below) is provided for this purpose. The process 52 comprises a filtering stage 54 and an instability detection stage 56. The filtering stage 54 ensures that the process 52 does not incorrectly trigger an alarm during normal driving conditions by defining a minimum rate of change for the difference between the trailer yaw angle (x) and the steering wheel angle (y) prior to advancing to the instability detection stage 56. The process 52 begins by calculating at step 58 the rate of change (e) of the difference between the trailer yaw angle (x) and the steering wheel angle (y). Next the process determines at step 60 whether the value of e exceeds a pre-determined threshold T_f. If e is below the threshold, the instability detection stage 56 is not enabled (at Step 62), and the process returns to step 58 to reiterate. If the value of e exceeds T_f, the instability detection stage 56 is enabled at step 64, and the process 52 advances. The on-board computer then calculates at step 66 the phase of vector v in the way described above. The process then iterates at step 68 the calculation of the phase of vand determines a trend for the phase of v. The on-board computer then repeats steps 66 and 68, to see whether the trend for the phase of v remains consistent. If the trend is consistent, the process continues to iterate steps 66 and 68. If the trend changes, an alarm is raised at step 70 to indicate trailer instability. This alarm may be used either to warn the driver, or to directly initiate corrective action such as activating the ESC system.

Figures 6 to 8 illustrate how the phase of v may follow a trend which can be used as described above to determine whether to raise an alarm. For example, in the phase vector graph 18 of Figure 6, from around 22 seconds the phase follows a periodic decreasing trend, with almost three complete cycles prior to an instability developing at around 33 seconds. Therefore, during the period between 22 seconds and 33 seconds, the process 52 described above would monitor the phase vector, and would find that there is a substantially consistent trend. At 33 seconds, the trend changes, therefore the process 52 would raise an alarm at this point to indicate that instability has started to develop before a state of critical instability is reached.

It will be appreciated by the skilled person that a change of phase may be calculated by taking a derivative of the signal and identifying stationary points. The stationary points can then be categorised by taking a second derivative, which identifies local maxima, minima and inflection points, indicating a change of direction. However, such an approach may only be practical in an offline set-up in which analysis is conducted after the event.

For real-time analysis, a change of phase may be detected by, for example, continuously sampling the value of the phase, comparing each new sample with the last, and determining whether the value is increasing or decreasing. Then, when a new sample value reverses the direction of the previous samples, for example if the new sample value is higher than the previous sample value, when up to that point the sample values had been decreasing with each iteration, this indicates a change of phase. An algorithm could be used to perform this analysis automatically, in which case a buffer may be included to prevent anomalous results triggering false alarms. Such an algorithm may be referred to as an online derivative calculation.

The process 52 described above was used to create the alarms in the simulations of Figures 5, 6 and 7. Figure 10 is a graph showing the results of a test of a system implementing the process described above for a real-world test. In this test, the vehicle- trailer combination 10 was driven steadily at 40mph, and impulse steering manoeuvers were performed at random intervals. As shown in Figure 10, the system successfully triggered an alarm each time an impulse steering manoeuver was performed.

With reference to Figure 1 1 , a control module 72 which may be used to implement the process 52 of Figure 9 is illustrated. The control module 72 is provided with a plurality of input means in the form of inputs 74, processing means in the form of a processor 76, and output means in the form of an output 78. The inputs 74 are arranged to receive the input signals relating to steering wheel angle and trailer yaw. The processor 76 is arranged to take these input signals and use them to conduct the process 52 outlined with reference to Figure 9, in order to predict a critical instability. On predicting critical instability, the control module 72 may be arranged to use the output 78 to output a control signal to control an alarm, where the alarm may be used to alert the driver. Alternatively, the output may output a control signal to a vehicle system, such as an electronic stability control (ESC) system, to automatically take corrective action on behalf of the driver, as described previously. Although two inputs 74 are shown in Figure 1 1 the control module may be provided with a single input for receiving sensor data or multiple inputs. As shown in Figure 12, the invention also extends to a vehicle 80 comprising the above described control module 72. The vehicle may be further provided with sensors (not shown) for the monitoring of the steering wheel angle and the trailer yaw angle.

It will be appreciated that although the above embodiments relate to measurements of trailer yaw angle, this is primarily for convenience, and the method could equally use other trailer direction related parameters, for example a direct measurement of the direction of travel of the trailer 14, to predict instability. Equally, the skilled person will appreciate that there are many alternative vehicle direction related parameters to steering wheel angle which provide an indication of the direction of travel of the vehicle 12 which could be utilised in the method for predicting instability. For example, gyroscopic sensors in the vehicle 12 could be used to measure vehicle yaw, or a GPS system and/or a system which uses gyroscopic measurement to provide positioning information in the event of a temporary cut in a GPS signal could provide data for the direction of travel of the vehicle 12. Another possibility would be to provide a sensor in the front wheels to detect their current alignment.

The above described embodiments of the invention offer the advantage that instability of the trailer 14 may be accurately predicted based on just two input variables, namely a variable indicating a direction of travel of the vehicle 12, and a variable indicating a direction of travel of the trailer 14. This negates any requirement to model the entire system of the vehicle-trailer combination 10, which would be far more computationally demanding.

The previously described system detects the onset of oscillations in the vehicle-trailer system at a very early stage and in particular detects when the oscillations are about to become unstable. This allows action to be taken by the driver to prevent the oscillation developing further. Furthermore, this can be used to allow the driver to determine an optimum driving speed (one at which the warning does not activate) thereby minimising the need for braking. In the embodiment of the invention described above, the primary output from the system is a warning given directly to the driver in order to allow the driver to take the required corrective action; however, in an alternative embodiment of the invention, the alert signal output from the system is optionally input into the vehicle's existing ESC system. This allows the ESC to automatically take corrective action on behalf of the driver. This approach has an advantage over existing systems in that over-braking of the vehicle wheels may be avoided. This is because a system operating in accordance with this embodiment of the invention directly monitors the movement of the trailer 14 such that the alert signal is output to the ESC at a much earlier stage, and therefore corrective action may be taken earlier to prevent instability from occurring. For example, this could mean that upon detecting the early onset of oscillation, the ESC system automatically eases off the accelerator without having to apply the vehicle's brakes in order to regulate the vehicle- trailer road speed.

In contrast, with a system in which the ESC is responsive to the output of gyroscopic yaw sensors disposed on the vehicle 12, it may be the case that, by the time the yaw sensors detect significant yaw of the vehicle 12 for the ESC to instigate corrective measures, the oscillations of the trailer 14 are such that it is too late to stop a critical instability from developing. In addition to this, the present invention offers the ability to detect instability of a trailer 14 which is very lightweight relative to the vehicle 12. In previous systems, movement of a very lightweight trailer of, for example, less than 750kg, would not have enough effect on the vehicle 12 to be detected by the ESC system, thus corrective action may not be taken at all. This could lead to goods contained in the trailer 14 becoming damaged.

The process 52 of Figure 9 for predicting a critical instability event may be adapted for use with a multi-trailer system, in which a vehicle is towing two or more trailers in series. In this multi-body scenario, as well as trailer movement falling out of step with the movement of the vehicle (as in the single trailer system), it is also possible for the movement of individual trailers to move out of step with one another. Either of these situations may be detected and used to predict critical instability in a similar manner to when dealing with a single trailer. For example, articulation angles between pairs of adjacent trailers may be measured and the process 52 can take two such measurements for two different pairs of trailers as the inputs. Alternatively, one articulation measurement along with the steering wheel angle may be used in the same way as in the single trailer arrangement. An artificial vector can then be constructed in the same way as outlined above in relation to a single trailer arrangement, and the phase of the vector can be monitored to identify a change in the trend of the phase, in order to predict a critical instability event.

It will be appreciated that various changes and modifications can be made to the process 52 for predicting trailer instability described herein without departing from the present invention. Further aspects of the present invention are set out in the following numbered paragraphs:

1 . A method of predicting instability in a vehicle-trailer combination, the method comprising: monitoring a vehicle direction related parameter; monitoring a trailer direction related parameter; comparing the vehicle direction related parameter with the trailer direction related parameter to determine a response of the trailer direction to the vehicle direction over time in order to define a response trend; and determining an instability event in dependence on a change in the response trend. 2. A method according to paragraph 1 , comprising outputting an alarm signal in dependence on making a positive determination of an instability event. 3. A method according to paragraph 2, comprising outputting the alarm signal to a driver of the vehicle to indicate the onset of a trailer instability. 4. A method according to paragraph 1 , wherein comparing the vehicle direction related parameter with the trailer direction related parameter comprises defining an artificial vector based on the vehicle direction related parameter and the trailer direction related parameter. 5. A method according to paragraph 4, wherein determining the response of the trailer comprises calculating the argument of the vector.

6. A method according to paragraph 5, wherein defining a response trend comprises monitoring the argument of the vector over time to identify a direction of growth of the argument of the vector.

7. A method according to paragraph 6, wherein the instability event is determined as a result of an unexpected change to the direction of growth of the argument. 8. A method according to paragraph 1 , wherein the vehicle direction related parameter is the steering wheel angle.

9. A method according to paragraph 1 , wherein the trailer direction related parameter is the trailer yaw.

10. A method according to paragraph 1 , wherein a vehicle brake is activated in response to an instability event being determined.

1 1 . A method according to paragraph 1 , wherein a vehicle mounted camera is used to monitor the trailer direction related parameter.

12. A control module for predicting instability in a vehicle-trailer combination, the control module comprising: an input arranged to receive a signal indicative of a vehicle direction related parameter; an input arranged to receive a signal indicative of a trailer direction related parameter; and a processor arranged to: compare the vehicle direction related parameter with the trailer direction related parameter; determine a response of the trailer direction to the vehicle direction over time in order to define a response trend; and determine an instability event in dependence on a change in the response trend.

13. A control module according to paragraph 12, further comprising an output arranged to output a control signal to activate an alarm.

14. A control module according to paragraph 13, wherein the alarm is used to activate an automatic vehicle safety system. 15. A system for predicting instability in a vehicle-trailer combination, the system comprising; a control module for predicting instability in a vehicle-trailer combination, the control module comprising: an input arranged to receive a signal indicative of a vehicle direction related parameter; an input arranged to receive a signal indicative of a trailer direction related parameter; and a processor arranged to: compare the vehicle direction related parameter with the trailer direction related parameter; determine a response of the trailer direction to the vehicle direction over time in order to define a response trend; and determine an instability event in dependence on a change in the response trend; at least one sensor arranged to monitor the vehicle direction related parameter; and at least one sensor arranged to monitor the trailer direction related parameter.

16. A system according to paragraph 15, wherein the at least one sensor arranged to monitor the vehicle direction related parameter comprises a steering wheel angle sensor.

17. A system according to paragraph 15, wherein the at least one sensor arranged to monitor the vehicle direction related parameter comprises a vehicle yaw sensor.

18. A system according to paragraph 15, wherein the at least one sensor arranged to monitor the trailer direction related parameter is a vehicle mounted camera. 19. An engine control unit comprising a control module for predicting instability in a vehicle-trailer combination, the control module comprising: an input arranged to receive a signal indicative of a vehicle direction related parameter; an input arranged to receive a signal indicative of a trailer direction related parameter; and a processor arranged to: compare the vehicle direction related parameter with the trailer direction related parameter; determine a response of the trailer direction to the vehicle direction over time in order to define a response trend; and determine an instability event in dependence on a change in the response trend.

20. An engine control unit according to paragraph 19, wherein the control module comprises an output arranged to output a control signal to activate an alarm.

21 . An engine control unit according to paragraph 20, wherein the alarm is used to activate an automatic vehicle safety system. 22. A vehicle comprising a control module for predicting instability in a vehicle-trailer combination, the control module comprising: an input arranged to receive a signal indicative of a vehicle direction related parameter; an input arranged to receive a signal indicative of a trailer direction related parameter; and a processor arranged to: compare the vehicle direction related parameter with the trailer direction related parameter; determine a response of the trailer direction to the vehicle direction over time in order to define a response trend; and determine an instability event in dependence on a change in the response trend.

23. A vehicle according to paragraph 22, wherein the control module comprises an output arranged to output a control signal to activate an alarm.

24. A vehicle according to paragraph 23, wherein the alarm is used to activate an automatic vehicle safety system. 25. A vehicle comprising a system for predicting instability in a vehicle-trailer combination, the system comprising; a control module for predicting instability in a vehicle- trailer combination, the control module comprising: an input arranged to receive a signal indicative of a vehicle direction related parameter; an input arranged to receive a signal indicative of a trailer direction related parameter; and a processor arranged to: compare the vehicle direction related parameter with the trailer direction related parameter; determine a response of the trailer direction to the vehicle direction over time in order to define a response trend; and determine an instability event in dependence on a change in the response trend; at least one sensor arranged to monitor the vehicle direction related parameter; and at least one sensor arranged to monitor the trailer direction related parameter. 26. A vehicle according to paragraph 25, wherein the control module comprises an output arranged to output a control signal to activate an alarm.

27. A vehicle according to paragraph 26, wherein the alarm is used to activate an automatic vehicle safety system.

28. A vehicle according to paragraph 25, wherein the at least one sensor arranged to monitor the vehicle direction related parameter comprises a steering wheel angle sensor. 29. A vehicle according to paragraph 25, wherein the at least one sensor arranged to monitor the vehicle direction related parameter comprises a vehicle yaw sensor.

30. A vehicle according to paragraph 25, wherein the at least one sensor arranged to monitor the trailer direction related parameter is a vehicle mounted camera.

31 . A vehicle comprising an engine control unit comprising a control module for predicting instability in a vehicle-trailer combination, the control module comprising: an input arranged to receive a signal indicative of a vehicle direction related parameter; an input arranged to receive a signal indicative of a trailer direction related parameter; and a processor arranged to: compare the vehicle direction related parameter with the trailer direction related parameter; determine a response of the trailer direction to the vehicle direction over time in order to define a response trend; and determine an instability event in dependence on a change in the response trend. 32. A vehicle according to paragraph 31 , wherein the control module comprises an output arranged to output a control signal to activate an alarm.

33. A vehicle according to paragraph 32, wherein the alarm is used to activate an automatic vehicle safety system.

Claims

1 . A method of predicting instability in a vehicle-trailer combination, the method comprising:

monitoring a vehicle direction related parameter;

monitoring a trailer direction related parameter;

comparing the vehicle direction related parameter with the trailer direction related parameter to determine a response of the trailer direction to the vehicle direction over time in order to define a response trend; and

determining an instability event in dependence on a change in the response trend.

2. A method according to claim 1 , comprising outputting an alarm signal in dependence on making a positive determination of an instability event.

3. A method according to claim 2, comprising outputting the alarm signal to a driver of the vehicle to indicate the onset of a trailer instability.

4. A method according to any one of claims 1 to 3, wherein comparing the vehicle direction related parameter with the trailer direction related parameter comprises defining an artificial vector based on the vehicle direction related parameter and the trailer direction related parameter.

5. A method according to claim 4, wherein determining the response of the trailer comprises calculating the argument of the vector.

6. A method according to claim 5, wherein defining a response trend comprises monitoring the argument of the vector over time to identify a direction of growth of the argument of the vector.

7. A method according to claim 6, wherein the instability event is determined as a result of an unexpected change to the direction of growth of the argument.

8. A method according to any one of claims 1 to 7, wherein the vehicle direction related parameter is the steering wheel angle.

9. A method according to any one of claims 1 to 8, wherein the trailer direction related parameter is the trailer yaw.

10. A method according to any one of claims 1 to 9, wherein a vehicle brake is activated in response to an instability event being determined.

1 1 . A method according to any one of claims 1 to 10, wherein a vehicle mounted camera is used to monitor the trailer direction related parameter.

12. A control module for predicting instability in a vehicle-trailer combination, the control module comprising:

means to receive a signal indicative of a vehicle direction related parameter;

means to receive a signal indicative of a trailer direction related parameter; and processing means arranged to:

compare the vehicle direction related parameter with the trailer direction related parameter;

determine a response of the trailer direction to the vehicle direction over time in order to define a response trend; and

determine an instability event in dependence on a change in the response trend.

13. A control module according to claim 12, further comprising output means arranged to output a control signal to activate an alarm.

14. A control module according to claim 13, wherein the alarm is used to activate an automatic vehicle safety system.

15. A system for predicting instability in a vehicle-trailer combination, the system comprising;

a control module according to any one of claims 12 to 14;

at least one sensor arranged to monitor the vehicle direction related parameter; and

at least one sensor arranged to monitor the trailer direction related parameter.

16. A system according to claim 15, wherein the at least one sensor arranged to monitor the vehicle direction related parameter comprises a steering wheel angle sensor.

17. A system according to claim 15 or 16, wherein the at least one sensor arranged to monitor the vehicle direction related parameter comprises a vehicle yaw sensor.

18. A system according to any one of claims 15 to 17, wherein the at least one sensor arranged to monitor the trailer direction related parameter is a vehicle mounted camera.

19. An engine control unit comprising a control module according to any one of claims 12 to 14.

20. A vehicle comprising a control module according to any one of claims 12 to 14.

21 . A vehicle comprising a system according to any one of claims 15 to 18.

22. A vehicle comprising an engine control unit according to claim 19.

23. A method substantially as described herein with reference to the accompanying figures.

24. A control module substantially as described herein with reference to the accompanying figures.

25. A system substantially as described herein with reference to the accompanying figures.

26. An engine control unit substantially as described herein with reference to the accompanying figures.

27. A vehicle substantially as described herein with reference to the accompanying figures.