Selective free rolling wheels to access robust vehicle-to

Abstract

A vehicle motion management, VMM, system (360) for a heavy vehicle (100), the system comprising: at least one wheel speed sensor (106) configured to output a wheel speed signal indicative of a rotational speed of a respective wheel (102) on the vehicle (100); at least one inertial measurement unit (IMU) (110) configured to output an IMU signal indicative of an acceleration of the vehicle (100); a motion estimation function configured to estimate a vehicle motion state (s), including a vehicle-to-ground speed, based at least in part on the wheel speed signal and at least in part on the IMU signal; and a motion support device (MSD) coordination function configured to coordinate actuation of a plurality of MSDs of the heavy vehicle in accordance with a vehicle motion request and the vehicle motion state (s), where the motion estimation function is arranged to model errors in an estimated vehicle motion state (s), and outputting a free roll request to an MSD coordination function if the error of the modeling fails to meet acceptance criteria, wherein the MSD coordination function is arranged to reduce a wheel slip setpoint for one or more wheels (102) of the heavy vehicle (100) in response to receiving the free roll request from the motion estimation function.

Description

Technical Field

[0001] The present disclosure relates to a method and control unit for robust heavy vehicle motion estimation. The method is particularly suitable for use with cargo transport vehicles such as trucks and semi-trailers. However, the present invention can also be applied to other types of heavy vehicles, such as construction equipment and mining vehicles as well as cars. Background Art

[0002] Traditionally, heavy vehicles are controlled using torque request signals that are generated based on the position of the accelerator or brake pedal and sent to motion support devices (MSDs) such as service brakes and propulsion devices via a controller area network (CAN) bus. However, advantages can be gained by instead controlling the actuators using wheel slip or wheel speed requests sent from a central vehicle controller to different actuators. This brings actuator control closer to the wheel end, thereby reducing latency and controlling the MSD faster and more accurately. Wheel slip-based MSD control methods are particularly suitable for wheel-end motors of battery or fuel cell powered heavy vehicles, where the motor axle speed can be accurately controlled at a high bandwidth. Wheel slip-based vehicle motion management (VMM) and its associated advantages are discussed, for example, in WO 2017 / 215751 A1 and WO 2021 / 144010 A1.

[0003] Wheel slip-based control of heavy vehicles relies on accurate knowledge of the vehicle ground speed and the rotational speed of the wheels, as these two quantities together determine the wheel slip. The rotational speed of the wheels can be reliably obtained from sensors such as Hall effect sensors or rotary encoders. However, at least in some more challenging environments and operating conditions, such as low friction conditions, breakaway friction conditions, and during maneuvers involving large wheel forces, it can be more difficult to obtain the vehicle ground speed in a robust and cost-effective manner.

[0004] An inertial measurement unit (IMU) provides data on acceleration and rotation, which can be integrated to obtain information about the vehicle's speed over the ground.

[0005] US 2018 / 0178767 A1 discusses vehicle speed determination based on a combination of IMU and wheel speed sensors.

[0006] US 2015 / 0291178 A1 describes a system for estimating vehicle speed based on an IMU in combination with wheel speed sensors and steering angle sensors.

[0007] US 5,173,860 A discloses a method for establishing a vehicle speed reference, wherein the rear wheels of the vehicle are alternately selected to establish a vehicle reference speed by free spinning, so as to avoid simultaneous free spinning of both rear wheels.

[0008] A Global Positioning System (GPS) receiver is often able to determine the vehicle's ground speed, but satellite systems are prone to errors in environments with strong multipath radio propagation and of course require a clear view of the sky to operate, which is not always possible. Camera systems can also be used, but these are costly and less effective in certain weather conditions.

[0009] In summary, there is an ongoing need for reliable and cost-effective methods to determine vehicle ground speed applicable to heavy vehicles, especially heavy vehicles based on wheel slip control. Summary of the invention

[0010] The object of the present disclosure is to provide an improved method for determining the ground speed of a heavy vehicle and performing vehicle motion management of the heavy vehicle. This object is achieved by a vehicle motion management (VMM) system for a heavy vehicle. The system includes at least one wheel speed sensor, the at least one wheel speed sensor being configured to output a wheel speed signal indicating a rotational speed of a corresponding wheel on the vehicle; and at least one IMU, the at least one IMU being configured to output an IMU signal indicating an acceleration of the vehicle. The system also includes a motion estimation function, the motion estimation function being configured to estimate a vehicle motion state including the vehicle ground speed based at least in part on the wheel speed signal and at least in part on the IMU signal; and an MSD coordination function, the MSD coordination function being configured to coordinate actuation of a plurality of MSDs of the heavy vehicle according to a vehicle motion request and the estimated vehicle motion state. The motion estimation function is arranged to model an error in the estimated vehicle motion state and output a free rolling request to the MSD coordination function if the modeled error does not meet an acceptance criterion, and the MSD coordination function is arranged to reduce a wheel slip set point of one or more wheels of the heavy vehicle in response to receiving the free rolling request from the motion estimation function. In this way, the estimation of the vehicle ground speed can be accurate despite the fact that the wheel speed sensor signal differs from the vehicle ground speed due to wheel slip. The system monitors the state estimation error and selectively places one or more wheels into a slip-reducing state when necessary to obtain accurate data about the vehicle ground speed. This means that the determination of the current wheel slip becomes more accurate, which in turn means that wheel slip-based control of heavy vehicles becomes more efficient.

[0011] According to some aspects, the motion estimation function is configured to estimate the vehicle motion state based on the wheel speed signal with a certain delay relative to the reduction of the wheel slip set point. In this way, transient effects are taken into account, which may otherwise cause errors in the estimated vehicle motion state. The motion estimation function can also be configured to estimate the vehicle motion state as the extreme value of the wheel speed signal within a certain time period based on the wheel speed signal. In this way, transients can also be avoided and the relevance of the wheel speed data for estimating the vehicle ground speed is increased.

[0012] The motion estimation function is preferably arranged to model the error in the estimated vehicle motion state at least in part by modeling the error associated with the IMU signal as a function that increases over time. Various such time-increasing functions can be envisioned. Some examples are given below. For example, the linearity of an affine function may represent an option that does not impose a substantial computational burden.

[0013] According to some aspects, the motion estimation functionality is arranged such that the estimated vehicle motion state is primarily based on the IMU signal when torque is applied at the wheel on the vehicle, and otherwise based on the wheel speed signal. For example, this selection may be implemented as a weighting between two sensor data types, or a simple switch based on which type is deemed most reliable at any given point in time. Thus, the motion estimation functionality may also be arranged to estimate the vehicle motion state based on a weighted combination of the wheel speed signal and the IMU signal, wherein the weights of the weighted combination are configured according to data indicative of a wheel slip setpoint and / or a torque setpoint.

[0014] The MSD coordination function may be configured to coordinate actuation of multiple MSDs of a heavy vehicle based on a solution to a constrained optimization problem, wherein one or more constraints of the constrained optimization problem are arranged to be configured depending on whether the free rolling request has been received. This type of mathematical optimization allows the VMM system to account for selective and temporary free rolling of wheels without significantly affecting the overall motion of the vehicle, as other MSD actuators may be used to compensate for the intermittent free rolling of one or more wheels.

[0015] The MSD coordination function is preferably arranged to output data indicative of wheel slip set points and / or torque set points for wheels on the heavy vehicle to the motion estimation function. This allows the motion estimation function to estimate the vehicle motion state based on the data indicative of wheel slip set points and / or torque set points. For example, the data may be used to determine which wheel speed sensors are outputting reliable data and which wheel speed sensors should be avoided when determining the vehicle ground speed. It is an advantage that the motion estimation function can also use the output data indicative of wheel slip set points and / or torque set points to proactively (i.e. before a wheel starts to slip or stops slipping) adjust the motion estimate.

[0016] The MSD coordination function may also be arranged to reduce the respective wheel slip set points of the one or more wheels of the heavy vehicle in a sequence, wherein each wheel in the sequence is placed in a low slip state for a predetermined duration. This balances the effects of free rolling on the vehicle to reduce the overall effect on vehicle motion.

[0017] A control unit, a vehicle, a computer program, a computer readable medium and a computer program product associated with the advantages discussed above are also disclosed herein.

[0018] Generally, unless otherwise clearly defined herein, all terms used in the claims should be interpreted according to their ordinary meanings in the art. All references to "a / an / the element, device, assembly, member, step, etc." will be interpreted as referring to at least one instance of the element, device, assembly, member, step, etc., unless otherwise clearly stated. The steps of any method disclosed herein do not have to be performed in the exact order disclosed, unless clearly stated. Other features and advantages of the present invention will become apparent when studying the attached claims and the following description. The technician will appreciate that the different features of the present invention can be combined to form embodiments other than the embodiments described below, without departing from the scope of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The above and additional objects, features and advantages will be better understood through the following illustrative and non-limiting detailed description of exemplary embodiments, in which:

[0020] Figure 1 An exemplary heavy vehicle is shown;

[0021] Figure 2 is a graph showing exemplary tire forces as a function of wheel slip;

[0022] Figure 3 An exemplary motion support device control arrangement is shown;

[0023] FIG. 4A to FIG. 4B An exemplary error evolution model is shown;

[0024] Figure 5 An exemplary vehicle control functional architecture is shown;

[0025] Figure 6 An exemplary heavy vehicle is shown;

[0026] FIG. 7A to FIG. 7B is a graph showing a wheel speed signal;

[0027] Figure 8 is a flow chart illustrating a method;

[0028] Fig. 9 A control unit is schematically shown; and

[0029] Fig.10 An exemplary computer program product is shown; DETAILED DESCRIPTION

[0030] The present disclosure will now be described more fully below with reference to the accompanying drawings, in which exemplary embodiments are shown. However, the present disclosure may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided for the sake of thoroughness and completeness. Throughout this description, like reference numerals refer to like elements.

[0031] Figure 1 An exemplary heavy vehicle 100 is shown, here in the form of a truck. It should be understood that the methods and control units disclosed herein may also be advantageously applied to other types of heavy vehicles, such as trucks with a drawbar connection, construction equipment, buses, etc. The vehicle 100 may also include more than two vehicle units. For example, a trailer vehicle unit may be used to tow more than one trailer.

[0032] The exemplary vehicle 100 includes a plurality of wheels 102, wherein at least a subset of the wheels 102 include respective motion support devices (MSDs) 104. Figure 1 The embodiment depicted in shows an MSD 104 for each of the wheels 102, but it should be readily appreciated that, for example, a pair of wheels 102 may be arranged without such an MSD 104. Moreover, an MSD may be arranged to be connected to more than one wheel, for example via a differential drive or the like.

[0033] At least some of the wheels 102 on the vehicle 100 are equipped with wheel speed sensors 106. Wheel speed sensors are sensors that measure the rotational speed of the wheels, for example based on Hall effect sensors, rotary encoders, etc. Wheel speed sensors are well known and therefore will not be discussed in detail here.

[0034] The MSD 104 may be arranged to generate torque at a corresponding wheel of the vehicle or for both wheels of the axle. The MSD may be a propulsion device, such as an electric motor arranged to provide longitudinal wheel forces to the wheels of the vehicle 100, for example. Thus, such an electric motor may be suitable for generating propulsion torque as well as being arranged in a regenerative braking mode for charging a battery (not shown) or other energy storage system of the vehicle 100.

[0035] The MSD 104 may also include friction brakes (such as disc brakes or drum brakes) arranged to generate braking torque from the wheels 102 in order to decelerate the vehicle. In this document, the term acceleration should be interpreted broadly to cover positive acceleration (propulsion) and negative acceleration (braking).

[0036] Each MSD 104 is connected to an MSD control unit 330, which is arranged to control various operations of the MSD 104. The MSD control system (i.e., the system of MSD control units) is preferably a decentralized system running on multiple separate wheel-end computers, but centralized implementations are also possible. In addition, it should be understood that some parts of the MSD control system may be implemented on a processing circuit system remote from the vehicle, such as on a remote server 120 that can be accessed from the vehicle via a wireless link. Each MSD control unit 330 is connected to the VMM system or function 360 of the vehicle 100 via a data bus communication device 114, which may be wired, wireless, or both wired and wireless. Therefore, control signals can be transmitted between the VMM function 360 and the MSD control unit 330. The VMM function 360 and the MSD control unit 330 will be discussed in conjunction with each other below. Figure 3 and Figure 5 Describe in more detail.

[0037] The VMM function 360 and the MSD control unit 330 may include a microprocessor, a microcontroller, a programmable digital signal processor, or another programmable device. The system may also or alternatively include an application specific integrated circuit, a programmable gate array or programmable array logic, a programmable logic device, or a digital signal processor. In the case where the system includes a programmable device (such as the microprocessor, microcontroller, or programmable digital signal processor mentioned above), the processor may also include computer executable code that controls the operation of the programmable device. Fig. 9 Implementation aspects of the various vehicle unit processing circuits are discussed in greater detail.

[0038] Typically, the MSDs on the vehicle 100 may also include, for example, power steering devices, active suspension devices, etc. Although these types of MSDs cannot be used to directly generate longitudinal forces to accelerate or brake the vehicle, they are still part of the overall vehicle motion management of the heavy vehicle and therefore may form part of the method for vehicle motion management disclosed herein. It is worth noting that the MSDs of the heavy vehicle 100 are typically coordinated to achieve the desired motion of the vehicle. For example, two or more MSDs may be used in conjunction to generate a desired propulsion torque or braking torque, a desired vehicle yaw motion, or some other dynamic behavior. In combination Figure 5 The coordination of MSD is discussed in more detail.

[0039] According to SAE J370 (SAE Vehicle Dynamics Standards Committee, January 24, 2008), longitudinal wheel slip Can be defined as

[0040] in is the effective wheel radius in meters, is the angular velocity of the wheel, and is the longitudinal velocity of the wheel (in the wheel's coordinate system). Therefore, is between -1 and 1 and quantifies the degree of wheel slippage relative to the road surface. Wheel slip is essentially the measured speed difference between the wheel and the vehicle. Therefore, the techniques disclosed herein can be adapted to any type of wheel slip definition. It should also be understood that in the coordinate system of the wheel, the wheel slip value is equivalent to the wheel speed value given the speed of the wheel on the road surface. The VMM function 360 and optionally the different MSD control unit 330 maintain information about the wheel in the reference system of the wheel. The wheel speed sensor 106 can be used to determine (Rotational speed of the wheel).

[0041] Slip Angle (also called the sideslip angle) is the angle between the direction the wheels are pointing and the direction they are actually traveling (i.e., the longitudinal component of velocity). With wheel forward speed and lateral speed The slip angle generates a force, namely a cornering force, in the plane of the contact patch and perpendicular to the intersection of the contact patch and the mid-plane of the wheel. In the first few degrees of the slip angle, the cornering force increases approximately linearly, then increases nonlinearly to a maximum value, after which it begins to decrease.

[0042] Slip Angle It is usually defined as In the wheel coordinate system, is the lateral velocity of the wheel.

[0043] In this document, longitudinal ground speed may be determined relative to the vehicle, in which case the direction of speed refers to the vehicle's forward direction, or may be determined relative to the wheels, in which case the direction of speed refers to the wheel's forward direction or rolling direction. The same is true for lateral ground speed, which may be the lateral speed of the vehicle or the lateral ground speed of the wheel relative to its rolling direction. The meaning is clear from the context, and it should be understood that straightforward conversions can be applied to convert ground speed between the vehicle's coordinate system and the wheel's coordinate system, and vice versa. For example, vehicle and wheel coordinate systems are discussed in Thomas Gillespie's "Fundamentals of Vehicle Dynamics" Warrendale, PA: Society of Automotive Engineers, 1992.

[0044] In order for a wheel (or tire) to generate wheel forces that affect the motion of a heavy vehicle (such as acceleration), slip must occur. For smaller slip values, the relationship between slip and generated force is approximately linear, with the proportionality constant usually expressed as the tire's slip stiffness C x The tire is subjected to a longitudinal force F x , lateral force F y and the normal force F z . Normal force F z is the key to determining some important vehicle properties. For example, the normal force largely determines the longitudinal tire force F that can be achieved by the wheel. x , because usually, ,in is the friction coefficient associated with the road friction conditions. The maximum available lateral force for a given wheel slip can be described by the so-called magic formula as described in "Tyre and vehicle dynamics" (Elsevier Ltd. 2012, ISBN 978-0-08-097016-5, Hans Pacejka), where wheel slip and tire forces are also discussed in detail.

[0045] Figure 2 is a graph showing an example 200 of achievable tire forces as a function of longitudinal wheel slip. Fx is the longitudinal tire force, and Fy is the maximum lateral wheel force achievable for a given wheel slip ratio. This type of relationship between wheel slip and generated tire forces is often referred to as an inverse tire model, and it is well known in the field of vehicle dynamics and vehicle control. Figure 2The examples in are for positive wheel forces, i.e. acceleration. A similar relationship exists between wheel slip and negative wheel forces, i.e. braking.

[0046] An inverse tire model such as Figure 2 The model 200 shown in FIG. 200 is used to calculate the desired longitudinal tire force F x Corresponding longitudinal wheel slip . The interface between the VMM and the MSD, which is capable of delivering torque to the wheels of the vehicle, as mentioned above, has traditionally focused on torque-based requests sent from the VMM function 360 to each MSD control unit 330, without any consideration of wheel slip. However, this approach has some performance limitations. If a safety hazard or excessive slip occurs, the relevant safety functions (traction control, anti-lock braking, etc.) operating on a separate control unit will usually intervene and request torque override in order to bring the slip back under control. The problem with this approach is that since the main control of the actuator and the slip control of the actuator are assigned to different electronic control units (ECUs), the time delay involved in the communication between them significantly limits the slip control performance. In addition, the relevant actuators and the slip assumptions made in the two ECUs used to implement the actual slip control may be inconsistent, and this in turn may lead to suboptimal performance.

[0047] Significant benefits may be realized by instead using wheel speed or wheel slip based requests on the interface between the VMM function 360 and the MSD control unit 330, thereby transferring the difficult actuator speed control loop to the MSD controller which is closer to the wheels and therefore can generally operate with shorter control latency than the central VMM function 360. This type of architecture may provide better disturbance rejection and therefore improved predictability of the forces generated at the tire road contact patch compared to a torque based control interface.

[0048] Reference again Figure 2 , the exemplary longitudinal tire force Fx shows a portion 210 that increases almost linearly at small wheel slips, followed by a portion 220 with more nonlinear behavior at larger wheel slips. It is desirable to maintain vehicle operation in the linear region 210, where the available longitudinal forces in response to applied braking commands are more predictable, and where sufficient lateral tire forces can be generated if necessary. To ensure operation in this region, a wheel slip limit λ of, for example, about 0.1 may be imposed on a given wheel. lim Therefore, with accurate knowledge of the current wheel slip ratio, operation in the linear region can be ensured, which greatly simplifies vehicle motion control to improve safety, efficiency and driver comfort.

[0049] Another benefit of this wheel slip based control approach is that it can effectively handle changes in road friction. A reduction in road friction usually results in a vertical scaling of the inverse tire model, e.g. Figure 2 230 in FIG. 2. Note that the peak force remains substantially at the same level as the wheel slip. Therefore, a wheel controller that maintains a given wheel slip (e.g., λ = 0.25) will automatically adjust to changes in friction, keeping the tire force close to the optimal value despite sudden changes in friction.

[0050] When using wheel slip to actively control one or more wheels on a heavy vehicle such as vehicle 100 and when performing less complex control such as locally imposing the wheel slip limit λ mentioned above at the wheel ends lim ) may not accurately know the wheel's (and the vehicle's) ground speed. For example, if a wheel speed sensor 106 (such as a Hall Effect sensor or a rotary encoder) is used to determine the vehicle ground speed, the vehicle ground speed will be determined incorrectly if the wheel used to estimate the ground speed itself slips excessively.

[0051] As described above, satellite-based positioning systems may be used to determine the ground speed of the heavy vehicle 100 and any given wheel on the vehicle 100. However, these systems may not function properly in some environments, such as environments where there is no clear view of the sky. Multipath propagation of satellite radio signals may also cause large errors in the estimated vehicle position, which in turn translates into errors in the estimated vehicle ground speed.

[0052] Vision-based sensor systems and radar systems may also be used to determine vehicle ground speed. However, the cost of such systems is relatively high and is not always without issues with accuracy and reliability. For example, vision-based sensors may suffer from performance degradation due to sun glare and fog, while radar sensor systems may be susceptible to interference from other radar transceivers.

[0053] For these and other reasons, a combination of an IMU and wheel speed sensors is often used to estimate vehicle speed. As long as there is no torque applied to the wheels, or the vehicle does not have significant yaw motion, its associated wheel slip on at least some wheels is likely to be small, which means that the wheel speed data is likely to be an accurate representation of the vehicle's ground speed as well. During periods of high wheel slip, such as when torque is applied to the wheels, the vehicle ground speed can be temporarily estimated by relying on the IMU signal. In this way, periods of high wheel slip (and unreliable wheel speed sensor data) can be "bridged" by relying instead on the IMU signal to track the vehicle ground speed until the wheel slip becomes small enough to rely on the wheel speed signal again. The VMM function 360 described herein is arranged so that the estimated vehicle motion state s is based primarily on the IMU signal when torque is applied at the wheels 102, 310 on the vehicle 100, and otherwise based primarily on the wheel speed signal. In other words, if there is no torque applied to the wheels or only a small torque is applied, the wheel speed data can be used to determine the vehicle ground speed. During periods of high torque application, the vehicle ground speed is determined by relying instead on the IMU signal. Of course, weighted combinations of the two sensor types are also possible, as will be discussed below.

[0054] A problem with most IMUs used to determine vehicle ground speed is that the IMU output is inaccurate and biased, causing drift. To reduce the IMU drift problem, it is proposed herein to model the error generated by integrating the IMU signal to obtain the vehicle ground speed. When the modeled error becomes unacceptably large, the estimated ground speed based on the IMU signal is "reset" based on the wheel speed sensor data using data from one or more free-rolling wheels or at least from wheels with less torque applied. If no suitable low-slip wheels are available, the method temporarily reduces the torque applied at one or more wheels, samples the vehicle speed, and then reapplies the torque. Therefore, according to the techniques proposed herein, the vehicle control function in the MSD control unit 330 and / or the VMM function 360 selectively and temporarily puts one or more wheels into a free-rolling state (or at least in a state with less wheel slip) in order to obtain reliable vehicle ground speed data from the wheel speed sensors of the free-rolling wheels. Once the vehicle ground speed is determined in this way, it can be used to reset the IMU-based vehicle ground speed estimate.

[0055] In other words, the motion estimation function of the vehicle 100 obtains information indicating whether the ground speed determined based on the IMU signal output is sufficiently accurate. If the reliability of the ground speed data from the IMU is not sufficient to perform vehicle motion management, the MSD coordination function reduces the slip of one or more wheels on the heavy vehicle, for example, by introducing constraints in the mathematical optimization problem being solved to obtain an MSD coordination solution that meets the global force requirements. This reduction may be temporary or last for a longer period of time. If the slip reduction is temporary, the function is similar to an anti-lock braking function (ABS) that intermittently reduces wheel slip in a periodic manner. When the wheel slip of a given wheel is reduced, the reliability of the vehicle ground speed data obtained from the wheel speed sensor of that wheel increases.

[0056] Summarizing the discussion so far, the present invention discloses a VMM function 360 for a heavy vehicle 100. The system includes at least one wheel speed sensor 106, the at least one wheel speed sensor being configured to output a wheel speed signal indicating the rotational speed of a corresponding wheel 102 on the vehicle 100; and at least one IMU 110, the at least one IMU being configured to output an IMU signal indicating the acceleration of the vehicle 100 relative to the ground. The vehicle's motion estimation function (usually but not necessarily forming part of the VMM function 360) is configured to estimate the vehicle's motion state s including the vehicle's ground speed based at least in part on the wheel speed signal and at least in part on the IMU signal. For example, the motion estimation function may implement a sensor fusion algorithm in which data from the vehicle's wheel speed sensors and data from one or more IMUs of the vehicle are combined into an estimate of the vehicle's motion state. Such sensor fusion may be implemented by known methods, such as a Kalman filter, etc. The motion estimation function may also be less complex, such as simply switching between vehicle ground speed estimation based on one or more wheel speed signals and vehicle ground speed estimation based on integrated IMU acceleration signals.

[0057] The MSD coordination function of the vehicle 100 is configured to coordinate the actuation of multiple MSDs of the heavy vehicle based on the vehicle motion request 375 and the vehicle motion state s (e.g., based on the desired acceleration profile or curvature that the vehicle 100 is to follow). The MSD coordination function may include mathematical optimization elements to obtain the desired motion of the vehicle, or lower complexity, such as using positive torque generating actuators when acceleration is required and using negative torque generating actuators when deceleration is required. Various types of MSD coordination functions are known in the art, so this topic will not be discussed in detail herein. It is worth noting that the complexity of the MSD coordination function can vary, from a simple connection between the vehicle's control input devices (steering wheel, pedals, etc.) and the MSD actuators, to more advanced control methods.

[0058] The motion estimation function is arranged to model the error in the estimated vehicle motion state s, and output a free rolling request to the MSD coordination function when the modeled error fails to meet the acceptance criteria. The complexity of the error modeling may also vary, as will be discussed below. For example, a simple linear function that increases over time may be used to model the error. More advanced error modeling methods may also consider other information sources and more than one sensor device. The MSD coordination function is configured to reduce the wheel slip set point of one or more wheels 102 of the heavy vehicle 100 in response to receiving a free rolling request 550 from the motion estimation function. For example, the MSD coordination function may be arranged to set the wheel slip request or torque request of one or more wheels 102 of the heavy vehicle 100 to zero in response to receiving a free rolling request 550, that is, to deactivate the torque actuator associated with a given wheel or a given wheel group. This reduction in the wheel slip set point of one or more wheels 102 of the heavy vehicle 100 causes the wheel slip to be reduced, and thus improves the accuracy of the vehicle ground speed data obtained from the wheel speed sensor of the wheel with the reduced wheel slip set point. The motion estimation function can then use this more accurate vehicle speed information to reset the IMU-based vehicle speed estimate.

[0059] The IMU output signal indicates the acceleration of the IMU components, so an estimated vehicle ground speed can be obtained by integrating the IMU acceleration signal from the known vehicle ground speed. However, the IMU signal is often biased and usually contains errors that accumulate to cause errors in the estimated ground speed. By characterizing the IMU based on such biases and errors, a model can be constructed to model the errors in the estimated vehicle ground speed determined from the IMU signal. For example, if the IMU output signal of the vehicle longitudinal acceleration is Roughly modeled as

[0060] in is the true longitudinal acceleration of the vehicle (relative to the ground), is a constant unknown bias, and is some form of measurement noise, such as The Gaussian zero-mean noise, then the error accumulated over time can be modeled as

[0061] If the deviation and measurement noise The statistical distribution of the error is roughly known, so the statistical data of the error It can also be determined using direct statistical methods or simply by actual experimentation or computer simulation. Alternatively, a linear or quadratic function of time, or some other form of polynomial function, can be assumed and fit to laboratory experiments on the error after integrating the IMU signal over time (e.g. by doing a least squares fit to the measured data).

[0062] The acceptance criterion may be a fixed threshold of the estimated error in the IMU signal integrator, or some statistical measure of the error. For example, the system may require that the probability that the actual error exceeds some predetermined threshold be kept below a certain level. If the estimated error grows too much, i.e. if the IMU signal is relied upon for estimating the vehicle ground speed for too long, a free rolling request is triggered, which results in free rolling of at least one wheel (or at least reduced wheel slip), and thus in the availability of more reliable wheel speed data, which can be used to reset the IMU signal integrator used to estimate the vehicle ground speed. When resetting the IMU signal integrator in this way, using reliable data from one or more wheel speed sensors, the error is also reduced.

[0063] Figure 4A An example 400 of the error in the estimate of the vehicle's ground velocity when the IMU signal is integrated over time is shown. The estimated error here follows a linear model. When the estimated error given by the model exceeds a threshold value Th, a free rolling request is triggered (at 410), resulting in a reset of the estimated error.

[0064] Figure 4B Another example 420 is shown where the error model is quadratic instead. When the error exceeds the acceptance criteria and a free roll request is generated, the error model is again reset (at 430 ), resulting in reliable data regarding vehicle ground speed being available again from the one or more wheel speed sensors.

[0065] As described above, both models 400, 420 may be pre-parameterized by computer simulation, practical experiments and / or mathematical analysis. Such parameterization may involve, for example, a comparison between an IMU signal or an estimate of the vehicle's ground velocity based on an integrated IMU signal and some form of ground truth reference ground velocity (e.g., obtained from GPS).

[0066] In general, the motion estimation functionality discussed herein may be arranged to model errors in the estimated vehicle motion state s at least in part by modeling errors associated with the IMU signals as a function 400 , 420 that increases over time.

[0067] Figure 3Schematically illustrated is a function 300 for controlling wheels 310 on a vehicle 100 by some exemplary MSDs, here comprising friction brakes 320 (such as disc brakes or drum brakes), propulsion devices 340 and steering devices 350. The friction brakes 320 and propulsion devices 340 are examples of wheel torque generating devices that can be controlled by one or more motion support device control units 330. The control is based on measurement data obtained from the wheel speed sensors 106 in combination with data from one or more IMUs 110, and optionally also based on data from other vehicle state sensors, such as radar sensors, lidar sensors, and vision-based sensors such as camera sensors and infrared detectors. The MSD control unit 330 can be arranged to control one or more MSD actuators. For example, it is common for a single MSD control unit 330 to be arranged to control two wheels on an axle.

[0068] Traffic situation management (TSM) function 370 plans driving maneuvers in a time frame of about 10 seconds. This time frame corresponds to, for example, the time it takes for the vehicle 100 to negotiate a curve, etc. The vehicle maneuvers planned and executed by the TSM function may be associated with acceleration profiles and curvature profiles that describe the desired target vehicle speed in the vehicle's forward direction and the steering to be maintained for a given maneuver. The TSM function constantly requests the desired acceleration profile a from the VMM function 360. req and the steering angle (or curvature curve c req ), the VMM function performs force allocation to satisfy the request from the TSM function in a safe and robust manner. The VMM function 360 operates on a time scale of less than one second or so and will be discussed in more detail below. The VMM function 360 then communicates with the different MSD control units 330 on the vehicle via an interface 365. The communication on the interface 365 may involve, for example, transmitting a wheel slip setpoint to the MSD control unit and receiving a capability signal from the MSD control unit 330.

[0069] Wheel 310 has a longitudinal velocity component and the lateral velocity component (in the wheel's coordinate system or in the vehicle's coordinate system, depending on the implementation). Longitudinal wheel force F x and the lateral wheel force F y and the normal force F z Acting on the wheel ( Figure 3 Unless explicitly stated otherwise, the wheel forces are defined in the coordinate system of the wheel, i.e., the longitudinal forces point in the rolling plane of the wheel, while the transverse wheel forces point perpendicular to the rolling plane of the wheel. The rotational speed of the wheel is , and the effective rolling radius is R.

[0070] The motion estimation systems discussed herein are used at least in part to determine the vehicle ground velocity, which can then be converted into wheel velocity components in the wheel's coordinate system. and / or This means that if the wheel is a steered wheel, the wheel steering angle δ is taken into account, whereas the longitudinal velocity component of a non-steered wheel is the same as the vehicle unit to which the wheel is attached, usually a truck or trailer vehicle unit.

[0071] VMM 360 can be used Figure 2 200 to generate a desired tire force at a certain wheel. Instead of requesting a torque corresponding to the desired tire force, the VMM may convert the desired tire force into equivalent wheel slip (or equivalently, wheel speed relative to ground speed) and request that slip instead. The main advantage is that the MSD control unit 330 will be able to calculate the vehicle speed by using the vehicle speed obtained from processing the data from the IMU 110. and the vehicle rotation speed obtained from the wheel speed sensor 106 Operation is maintained at the desired wheel slip to deliver the requested torque at a higher bandwidth. The one or more control units 330 may be arranged to store one or more predetermined inverse tire models in a memory, for example as a lookup table or parameterized function. The inverse tire model may also be arranged to be stored in a memory as a function of the current operating condition of the wheel 310.

[0072] According to a simple example of the techniques presented herein, as long as no torque is applied to the wheels, the vehicle speed data obtained from the wheel speed sensors 106 is considered reliable and used at the MSD control unit 330 to determine the vehicle ground speed and / or fed back to the VMM function 360 where it is used as a basis for determining the vehicle ground speed. If torque is applied, such as by the propulsion device 340 or the service brake 320, then the acceleration data from the IMU 110 will be integrated to track the vehicle ground speed in lieu of the data from the wheel speed sensors. In this way, the MSD control unit 330 can determine wheel slip during the application of torque because the vehicle ground speed can be tracked for a limited duration using the IMU signals, while the wheel speed sensors provide wheel speed information during the generation of tire forces.

[0073] The accuracy of the vehicle ground speed determined based on the IMU output signal deteriorates over time due to the accumulation of errors in the integrated IMU signal. When the estimated error magnitude (obtained from the type of model discussed above) becomes unacceptably large, the IMU integrator is corrected by placing the wheel in a free-rolling state, estimating the vehicle ground speed based on the wheel speed sensor, reinitializing the IMU integrator again, and reapplying torque to the wheel. This free-rolling of the wheel can be triggered centrally by the VMM function 360, or locally at the MSD control unit 330.

[0074] Figure 5 An exemplary vehicle control functional architecture suitable for the method disclosed herein is shown. In this exemplary architecture, the TSM function 370 generates a vehicle motion request 375, which may include a desired steering angle δ or equivalent curvature c that the vehicle is to follow. req , and may also include the desired vehicle unit acceleration a req and other types of vehicle motion requests which together describe the desired motion of the vehicle along a desired path at a desired speed profile. It should be understood that the motion request can be used as a basis for determining or predicting the required amount of longitudinal and lateral forces that need to be generated in order to successfully complete the maneuver. Of course, the TSM function 370 can also be replaced by driver input signals from the steering wheel and pedals.

[0075] The VMM system operates in a time frame of about 1 second and continuously transmits the acceleration curve a from the TSM function 370 to the VMM system. req and the curvature curve c reqThe VMM function 360 may be used to control the vehicle state or motion estimation by the motion estimation function 510 as discussed above, i.e., the VMM system continuously determines the vehicle state s including, for example, the position, velocity, acceleration and articulation angle of different units in the vehicle combination by monitoring the operation using various sensors 540 arranged on the vehicle 100, usually but not always associated with the MSDs. Important inputs to the motion estimation function 510 are signals from the IMU 110 and the wheel speed sensors 106 on the heavy vehicle 100. As discussed above, the motion estimation function 510 is configured to estimate at least the vehicle ground speed based on the wheel speed signals from the wheel speed sensors 106 and the IMU signals from the IMU 110. The motion estimation function is also arranged to model the error in the estimated vehicle motion state s (at least for the vehicle ground speed) and, if the modelled error fails to meet an acceptance criterion (e.g. a threshold or some predetermined confidence interval), output a free roll request 550 to the MSD coordination function 530 (or some equivalent software module).

[0076] The results of the motion estimation 510 (ie, the estimated vehicle state s) are input to a force generation module 520 which determines the global forces V = [V 1 ,V 2 ] so that the vehicle 100 is accelerated according to the requested acceleration curve a req and the curvature curve c req The desired global force vector V is input to the MSD coordination function 530, which distributes wheel forces and coordinates other MSDs, such as steering and suspension. The MSD coordination function outputs the MSD control distribution for the i-th wheel, which may include torque T i , longitudinal wheel slip λ i , wheel speed ω i and / or wheel steering angle δ i The coordinated MSDs then work together to provide the desired lateral force Fy and longitudinal force Fx on the vehicle unit, as well as the required moment Mz, to achieve the desired motion of the vehicle combination 100 .

[0077] According to the teachings herein, the MSD coordination function is arranged to reduce the wheel slip set point of one or more wheels 102 of the heavy vehicle 100 in response to receiving a free rolling request 550 from the motion estimation function. For example, the MSD coordination function may be arranged to set the wheel slip request or torque request to zero for one or more wheels 102 of the heavy vehicle 100 in response to receiving the free rolling request 550. In this way, when the data from the IMU deteriorates too much due to error accumulation, the motion estimation function 510 may request to temporarily increase the reliability of the wheel speed signal from the wheel speed sensor used to estimate the vehicle ground speed.

[0078] Figure 5 The exemplary VMM function 360 in FIG. 3 manages both force generation and MSD coordination, i.e., it determines what forces are needed at the vehicle units to satisfy requests from the TSM function 370, such as accelerating the vehicle according to a requested acceleration profile requested by the TSM and / or generating a specific curvature motion of the vehicle also requested by the TSM. The forces may include, for example, yaw moment Mz, longitudinal force Fx, and lateral force Fy, as well as different types of torque applied at different wheels. The forces are determined such as to generate vehicle behavior that is expected by the TSM function in response to the control inputs generated by the TSM function 370.

[0079] The MSD coordination function 530 may implement a mathematical optimization routine that finds an MSD force distribution that corresponds to the desired global force determined by the force generation module 520. The mathematical optimization routine involves constraints, i.e., limits on the forces that a given MSD may generate. Thus, the MSD coordination function 530 may be used to reduce or even eliminate wheel slip on one or more wheels 310, which facilitates more accurate determination of vehicle speed using the wheel speed sensors 106. The constraints may be imposed as wheel slip limits or as torque limits, which may be set to some small value or even to zero, where the wheels are essentially in a free-rolling state.

[0080] According to some aspects, as described above, the MSD coordination function 530 is arranged to set the wheel slip request and / or torque request to zero for one or more wheels 310 of the heavy vehicle 100 in response to receiving the free rolling request 550. Thus, no positive or negative wheel forces are generated in the longitudinal direction of the wheels, which means that the impact of the vehicle speed determination based on the wheel speed of the wheels is minimized or at least reduced.

[0081] According to other aspects, the MSD coordination function 530 is arranged to reduce the wheel slip set point of one or more wheels 310 of the heavy vehicle 100 in a certain sequence, wherein each wheel in the sequence is in a low slip state for a predetermined short duration (such as one second or half a second). In this way, the vehicle can be actuated because each wheel is in a low slip state only for a short period of time, after which the restoring force can be generated. Figure 6 This mode of operation is discussed in more detail.

[0082] The MSD coordination function 530 is optionally configured to coordinate actuation of a plurality of MSDs of a heavy vehicle based on a solution of a constrained optimization problem, wherein one or more constraints of the constrained optimization problem are arranged to be configured according to an error magnitude of an estimate associated with the vehicle ground speed based on the IMU signal, as described above, for example in conjunction with Figure 4A and Figure 4B This means that the MSD coordination function 530 can simply solve the optimization problem under the constraint that the wheel slip distributed at one or more wheels of the heavy vehicle should be kept below a certain threshold or even set to zero in order to reset the IMU integrator error and enable a continuous and reliable estimation of the vehicle ground speed.

[0083] It is worth noting that free rolling can also be performed locally, such as by the MSD control unit 330. This enables the MSD control unit to obtain a local estimate of the vehicle's ground speed, thereby allowing the MSD control unit to perform local wheel slip estimation and control. In other words, the MSD control unit can perform slip control for a limited period of time using locally available wheel slip data that uses the vehicle's ground speed determined based on the locally available IMU signal. When the locally modeled vehicle ground speed error becomes too large, the MSD control unit 330 can report a reduced capability to the VMM function 360 and then temporarily place its wheels in a slip reduction state or even a free rolling state. The VMM function 360 that receives the updated capability message in time before the MSD control unit 330 places the wheels in the free rolling state can then compensate for the actions performed locally by the MSD controller 330 (e.g., by the MSD coordination function 530).

[0084] The MSD coordination function 530 may also be arranged to output data 560 indicating wheel slip set points and / or torque set points for wheels 310 on the heavy vehicle 100 to the motion estimation function 510, i.e., a signal indicating whether a given wheel can be used to determine the vehicle ground speed. The motion estimation function 510 can then use the wheel slip indication data 560 to estimate the vehicle motion state s (particularly the vehicle ground speed) in a more reliable manner, because it now knows how the wheel will slip in the near future (when the actuator actuates the MSD set point). For example, the MSD coordination function 530 can transmit the slip limits it imposes on different wheels, and the motion estimation function 510 can then determine which wheel speed sensor signals can be used to reliably estimate the vehicle ground speed. For example, the motion estimation function 510 can estimate the vehicle motion state s based on the wheel speed of the wheel in the case of low slip, otherwise estimate the vehicle motion state based on one or more IMU signals. The motion estimation can also operate in a more active manner, avoiding the influence of transient errors caused by the onset of wheel slip.

[0085] The motion estimation function 510 estimates the vehicle motion state s, in particular the vehicle ground speed, optionally based on a weighted combination of wheel speed sensor data and IMU data, wherein the weights of the weighted combination are configured according to data indicative of wheel slip setpoints and / or torque setpoints 560. This means that the motion estimation function performs a type of sensor fusion that takes into account the accuracy of the estimates of the different sensors, resulting in improved accuracy and reliability.

[0086] Based on the error model discussed above, if the IMU integrator has been recently reset, the sensor fusion operation will assign more weight to the data from the IMU 110 than if the IMU acceleration signal has been integrated for a longer time. Parameters The estimated parameters of the weighted combination estimate (such as the vehicle's ground speed) can be written as

[0087] in , and the weight The relative size of the corresponding parameters Thus, if the IMU estimator performs very well when there is significant wheel slip, the weight of the IMU weight parameter will be close to one, but if the IMU data is deemed inaccurate and / or if there is no significant wheel slip on some of the wheels, then the relative weight of the IMU data will be reduced relative to the weight of the estimates from the wheel speed sensors, and optionally also relative to other data sources such as estimates from a GPS system.

[0088] A particular advantage may be obtained if the slip setpoints of the wheels on the heavy vehicle are temporarily reduced in a sequence such that the slip of each wheel in the sequence is temporarily reduced for a short period of time. This provides a free-rolling effect of the wheels distributed over the vehicle, avoiding excessive yaw motion, pitch motion etc. The MSD coordination function 530 is optionally arranged to reduce the respective wheel slip setpoints of one or more wheels 102, 310 of the heavy vehicle 100 in a predetermined or random sequence, wherein each wheel in the sequence is placed in a low slip state for a predetermined and limited duration.

[0089] refer to Figure 6 For example, a sequence may be selected to balance the torque losses on both sides of the vehicle, such as {A, B, C, D, E, F, G, H, I, J}. Thus, the sequence may involve periodically applying wheel slip or torque constraints on the left and right sides of the vehicle, i.e., first placing the wheels on the left side of the vehicle in a low wheel slip state, then placing the wheels on the right side of the vehicle in a low wheel slip state, and then placing the wheels on the left side in a low slip state again. This provides an opportunity to disperse the effects of placing one or more wheels in a low wheel slip state. The wheels in a low slip state used to estimate the vehicle speed may also be switched sequentially between two or more axles. The period of time that each wheel is placed in a low slip state does not need to be long, typically one second or less is sufficient to obtain an estimate of the vehicle speed from the wheel speed sensor, which estimate is not significantly affected by wheel slip.

[0090] Figure 6Also shown is an exemplary arrangement of IMUs 110a, 110b, 110c (i.e., a plurality of IMUs), each of which is mounted at a respective location on the vehicle 100. The different IMUs measure the acceleration of their respective portions of the vehicle 100. In general, when more than one IMU is available, it makes sense to base the motion estimate of a given portion of the vehicle primarily on the output of an IMU located near the portion of the vehicle. For example, to determine the vehicle ground speed of wheels A, E, J, C, G, or I, IMU 110c is most suitable, while when the vehicle ground speed of wheels H, D, F, or B is to be estimated, IMU 110b is more suitable. IMU 110a is located at the front of the vehicle, near the steering axle, and is therefore suitable for estimating the vehicle ground speed near the steering front axle, for example, during periods of large wheel slip at the front axle. Of course, a weighted combination of IMU signals may also be used advantageously, wherein the weights may be based, for example, on the distance from a given IMU device to the wheel where the vehicle ground speed is required to determine the wheel slip. Typically, the vehicle 100 may include multiple IMUs, wherein each wheel on the vehicle is associated with one or more of the multiple IMUs, and wherein the associated IMUs provide IMU signals that the VMM functionality 360 uses to estimate the vehicle ground speed during time periods when the wheels are associated with significant wheel slip (i.e., wheel slip levels that are above some predetermined threshold or do not meet some other type of predetermined acceptance criteria).

[0091] According to other aspects, the motion estimation function 510 is configured to estimate the vehicle motion state s based on the wheel speed signal with a certain delay relative to the reduction of the wheel slip set point. This delay allows transient phenomena to stabilize before the vehicle ground speed is "sampled" using the wheel speed sensor. Fig. 7A An example of such transient behavior in the wheel speed sensor signal output is shown at 700. At 710, negative torque is removed from the wheel, i.e., braking is suspended. This causes the wheel to accelerate to a wheel speed close to the vehicle ground speed, which occurs at about the time instant marked 720 after a delay 730. Therefore, by allowing for a delay, the effects of such transient behavior may be mitigated or even avoided entirely. Similar behavior may be observed if positive torque is removed from the wheel, in which case the wheel speed will decrease to more closely follow the actual vehicle ground speed.

[0092] The motion estimation function 510 can also be configured to estimate the vehicle motion state s based on the wheel speed signal as the extreme value points of the wheel speed signal within a given time period (maximum value in braking situation and minimum value in acceleration situation). The principle is that the maximum or minimum wheel speed signal is closest to the vehicle ground speed. Figure 7BAn example 750 is shown where two wheel speed signals 770, 780 are used to estimate vehicle ground speed. The true vehicle ground speed is also plotted as a dashed line 760 for reference. It is noteworthy that as torque decreases, the wheel speed signals increase until a maximum value is reached that is close to the true vehicle ground speed. Each time a wheel speed maximum is observed, the IMU integrator can be reset, thereby maintaining accurate vehicle ground speed information throughout the vehicle maneuver.

[0093] Figure 8 1 is a flow chart showing a method summarizing some of the key concepts discussed above. A computer-implemented method for performing VMM functions on a heavy vehicle 100 is shown. The method includes: configuring S1 at least one wheel speed sensor 106 to output a wheel speed signal indicating a rotational speed of a corresponding wheel 102, 310 on the vehicle 100; configuring S2 at least one inertial measurement unit IMU 110 to output an IMU signal indicating an acceleration of the vehicle 100; estimating S3 a vehicle motion state s including a vehicle ground speed based at least in part on the wheel speed signal and at least in part on the IMU signal by a motion estimation function 510; configuring S4 a motion support device MSD coordination function 530 to coordinate actuation of multiple MSDs of the heavy vehicle according to a vehicle motion request 375 and the vehicle motion state s; modeling S5 an error in the estimated vehicle motion state s, and triggering generation of a free rolling request 550 if the modeled error fails to meet an acceptance criterion; and reducing S6 a wheel slip set point of one or more wheels 102, 310 of the heavy vehicle 100 in response to the free rolling request 550. Therefore, the method covers the above combined Figure 1 Of course, different aspects can also be used with Figure 8 The methods shown are applied together.

[0094] Fig. 9 Components of the control unit 900 according to the embodiments discussed herein are schematically shown in terms of a number of functional units, such as any of the MSD control unit 330 or the VMM function 360. The processing circuitry 910 is provided using any combination of one or more of a suitable central processing unit CPU, a multiprocessor, a microcontroller, a digital signal processor DSP, etc., capable of executing software instructions stored, for example, in a computer program product in the form of a storage medium 930. The processing circuitry 910 may also be provided as at least one application specific integrated circuit ASIC or a field programmable gate array FPGA. Specifically, the processing circuitry 910 is configured to cause the control unit 900 to perform a set of operations or steps, such as in conjunction with Figure 8And generally the methods discussed herein. For example, the storage medium 930 may store a set of operations, and the processing circuit 910 may be configured to retrieve the set of operations from the storage medium 930 to cause the control unit 900 to perform the set of operations. The set of operations may be provided as a set of executable instructions. Thus, the processing circuit 910 is thereby arranged to perform the methods as disclosed herein.

[0095] The storage medium 930 may also include a persistent storage device, which may be, for example, any one or combination of magnetic storage, optical storage, solid-state storage, or even remotely mounted storage.

[0096] The control unit 900 may also include an interface 920 for communicating with at least one external device. Thus, the interface 920 may include one or more transmitters and receivers, including analog and digital components and a suitable number of ports for wired or wireless communication.

[0097] Processing circuit 910 controls the general operation of control unit 900 (e.g., by sending data and control signals to interface 920 and storage medium 930, by receiving data and reports from interface 920, and by retrieving data and instructions from storage medium 930). Other components of the control node and related functionality are omitted to avoid obscuring the concepts presented herein.

[0098] Fig.10 A computer readable medium 1010 is shown carrying a computer program comprising a program product for executing a program when the program product is run on a computer. Figure 8 The computer readable medium and the code means may together form a computer program product 1000.

Claims

1. A vehicle movement management VMM system (360) for a heavy vehicle (100), the system comprising at least one wheel speed sensor (106) configured to output a wheel speed signal indicative of a rotational speed of a corresponding wheel (102, 310) on the vehicle (100), at least one inertial measurement unit (IMU) (110) configured to output an IMU signal indicative of an acceleration of the vehicle (100), A motion estimation function (510) configured to estimate a vehicle ground speed based at least in part on the wheel speed signal and at least in part on the IMU signal. ), and a motion support device MSD coordination function (530) configured to coordinate actuation of a plurality of MSDs of the heavy vehicle based on a vehicle motion request (375) and the vehicle motion state(s), wherein the motion estimation function (510) is arranged to model an error in the estimated vehicle motion state (s) and output a free roll request (550) to the MSD coordination function (530) if the modelled error fails to meet an acceptance criterion, The MSD coordination function (530) is arranged to reduce a wheel slip set point of one or more wheels (102, 310) of the heavy vehicle (100) in response to receiving the free rolling request (550) from the motion estimation function (510).

2. The VMM system (360) of claim 1, wherein the MSD coordination function (530) is arranged to set a wheel slip request of one or more wheels (102, 310) of the heavy vehicle (100) to zero in response to receiving the free roll request (550).

3. A VMM system (360) according to claim 1 or 2, wherein the MSD coordination function (530) is arranged to set the torque request of one or more wheels (102, 310) of the heavy vehicle (100) to zero in response to receiving the free rolling request (550).

4. A VMM system (360) according to any preceding claim, wherein the motion estimation function (510) is configured to estimate the vehicle motion state (s) based on the wheel speed signal with a certain delay (730) relative to the reduction of the wheel slip set point.

5. A VMM system (360) according to any preceding claim, wherein the motion estimation function (510) is configured to estimate the vehicle motion state (s) based on the wheel speed signal as an extreme value of the wheel speed signal within a certain time period (790).

6. A VMM system (360) according to any preceding claim, wherein the motion estimation function (510) is arranged to model the error in the estimated vehicle motion state (s) at least in part by modeling the error associated with the IMU signal as a function (400, 420) that increases over time.

7. A VMM system (360) according to any preceding claim, wherein the motion estimation function is arranged (510) such that the estimated vehicle motion state (s) is primarily based on the IMU signal when torque is applied at the wheels (102, 310) on the vehicle (100), and otherwise based on the wheel speed signal.

8. A VMM system (360) according to any preceding claim, wherein the MSD coordination function (530) is configured to coordinate the actuation of the multiple MSDs of the heavy vehicle based on a solution to a constrained optimization problem, wherein one or more constraints of the constrained optimization problem are arranged to be configured based on whether the free rolling request (550) has been received.

9. A VMM system (360) according to any preceding claim, wherein the MSD coordination function (530) is arranged to output data (560) indicating wheel slip set points and / or torque set points of wheels (102, 310) on the heavy vehicle (100) to the motion estimation function (510), wherein the motion estimation function (510) is arranged to estimate the vehicle motion state (s) based on the data (560) indicating wheel slip set points and / or torque set points.

10. A VMM system (360) according to any one of claims 9, wherein the motion estimation function (510) is arranged to estimate the vehicle motion state (s) based on a weighted combination of the wheel speed signal and the IMU signal, wherein the weights of the weighted combination are configured according to the data (560) indicating a wheel slip set point and / or a torque set point.

11. A VMM system (360) according to any preceding claim, wherein the MSD coordination function (530) is arranged to reduce the corresponding wheel slip set points of the one or more wheels (102, 310) of the heavy vehicle (100) in a certain sequence, wherein each wheel in the sequence is placed in a low slip state for a predetermined duration.

12. A heavy vehicle (100) comprising a VMM system (360) according to any preceding claim.

13. A computer-implemented method for performing vehicle motion management (VMM) functions on a heavy vehicle (100), the method comprising configuring (S1) at least one wheel speed sensor (106) to output a wheel speed signal indicative of a rotational speed of a corresponding wheel (102, 310) on the vehicle (100), configuring (S2) at least one inertial measurement unit (IMU) (110) to output an IMU signal indicative of an acceleration of the vehicle (100), estimating (S3) a vehicle motion state (s) including a vehicle ground speed based at least in part on the wheel speed signal and at least in part on the IMU signal by a motion estimation function (510), configuring (S4) a motion support device MSD coordination function (530) to coordinate actuation of a plurality of MSDs of the heavy vehicle according to a vehicle motion request (375) and the vehicle motion state (s), modeling (S5) an error in the estimated vehicle motion state (s) and triggering the generation of a free roll request (550) if the modeled error fails to meet an acceptance criterion, and In response to the free rolling request (550), a wheel slip set point of one or more wheels (102, 310) of the heavy vehicle (100) is reduced (S6).

14. A computer program (1020) comprising program code means for performing the steps of claim 13 when said program is run on a computer.

Images