METHOD AND SYSTEM FOR RECOGNIZING THE SIGN OF THE VELOCITY OF A VEHICLE AND FOR ESTIMATING THE ROAD SLOPE

0Associated
Cases 
0Associated
Defendants 
0Accused
Products 
33Forward
Citations 
0
Petitions 
1
Assignment
First Claim
1. Method of estimating the slope of a road on which there is positioned a motor vehicle provided with rotational velocity sensor means for measuring the modulus of the velocity of the vehicle and with acceleration measuring means for measuring the acceleration in the longitudinal direction of the vehicle, characterized in that it comprises the following operations:
 a) providing a signal indicating the measurement of the velocity of the vehicle (V1, NCltFilV) by means of the said velocity sensor means;
b) providing a signal indicating the measurement of the acceleration in the longitudinal direction of the vehicle (AxVehFilV) by means of the said acceleration measuring means;
c) determining an estimated value of the slope of the road (StimaPend), the said determination of the estimated value comprising, subject to a condition of enabling the estimation, the processing of the velocity measurement signal (V1) and of the acceleration measurement signal (AxVehFilV) by means of a mathematical model based on the relation
A_{x}={dot over (ν
)}+g sin α
, where A_{x }represents the measurement of acceleration in the longitudinal direction of the vehicle, {dot over (ν
)} indicates the component of longitudinal acceleration of the vehicle along the road, g is the gravitational acceleration and a represents the angle of inclination of the road profile;
d) processing the velocity measurement signal (V1, NCltFilV), the acceleration measurement signal (AxVehFilV) and the estimated value of the slope (StimaPend) in such a way as to determine the sign of the velocity of the vehicle (Segno); and
e) repeating operations a) to d) recursively according to a predetermined rule, in which the sign of the velocity (Segno) determined in operation d) of a specific recursive step is used as an additional input for operation c) of a successive recursive step to update the estimated value of the slope of the road (StimaPend).
1 Assignment
0 Petitions
Accused Products
Abstract
A method of estimating the slope of a road surface on which is positioned a motor vehicle provided with a velocity sensor for measuring the modulus of the velocity of the vehicle and with an accelerometer for measuring the acceleration in the longitudinal direction of the vehicle comprises the following operations: a) providing a signal indicating the measurement of the velocity of the vehicle (V1) by means of the velocity sensor; b) providing a signal indicating the measurement of the acceleration in the longitudinal direction of the vehicle (AxVehFilV) by means of the accelerometer; c) processing the velocity measurement signal (V1) and the acceleration measurement signal (AxVehFilV) by means of a mathematical model based on the relation A_{x}={dot over (ν)}+g sin α, where A_{x }represents the measurement of acceleration in the longitudinal direction of the vehicle, {dot over (ν)} indicates the component of longitudinal acceleration of the vehicle along the road surface, g is the gravitational acceleration, and α represents the angle of inclination of the road surface, in such a way as to determine an estimated value of the slope of the road surface (StimaPend); d) processing the velocity measurement signal (V1), the acceleration measurement signal (AxVehFilV) and the estimated value of the slope (StimaPend) in such a way as to determine the sign of the velocity of the vehicle (Segno); and e) repeating operations a) to d) recursively according to a predetermined rule, in which the sign of the velocity (Segno) determined in operation d) of a specific recursive step is used as an additional input for operation c) of the successive recursive step to correct the estimated value of the slope (StimaPend).
38 Citations
View as Search Results
Method for determining the speed of a motor vehicle  
Patent #
US 7,970,525 B2
Filed 05/21/2009

Current Assignee
BMW AG

Sponsoring Entity
BMW AG

Method For Determining The Speed of A Motor Vehicle  
Patent #
US 20090265073A1
Filed 05/21/2009

Current Assignee
BMW AG

Sponsoring Entity
BMW AG

ACCELERATION DETECTION DEVICE, ACCELERATION DETECTION METHOD, AND NEUTRAL CONTROL DEVICE EMPLOYING SAME  
Patent #
US 20080140291A1
Filed 12/05/2007

Current Assignee
Nissan Motor Co. Ltd.

Sponsoring Entity
Nissan Motor Co. Ltd.

Acceleration detection device, acceleration detection method, and neutral control device employing same  
Patent #
US 8,340,860 B2
Filed 12/05/2007

Current Assignee
Nissan Motor Co. Ltd.

Sponsoring Entity
Nissan Motor Co. Ltd.

METHOD AND APPARATUS FOR GPS BASED SLOPE DETERMINATION, REALTIME VEHICLE MASS DETERMINATION, AND VEHICLE EFFICIENCY ANALYSIS  
Patent #
US 20130166170A1
Filed 12/18/2012

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

SYSTEM AND METHOD FOR CONTROLLING A TRANSMISSION  
Patent #
US 20130173124A1
Filed 12/28/2011

Current Assignee
Caterpillar Incorporated

Sponsoring Entity
Caterpillar Incorporated

ESTIMATION OF ROAD INCLINATION  
Patent #
US 20140032068A1
Filed 04/03/2012

Current Assignee
Scania CV AB

Sponsoring Entity
Scania CV AB

ROAD GRADIENT ESTIMATION ARBITRATION  
Patent #
US 20140067153A1
Filed 01/07/2013

Current Assignee
Ford Global Technologies LLC

Sponsoring Entity
Ford Global Technologies LLC

System and method for controlling a transmission  
Patent #
US 8,849,528 B2
Filed 12/28/2011

Current Assignee
Caterpillar Incorporated

Sponsoring Entity
Caterpillar Incorporated

3axis magnetic field sensor, method for fabricating magnetic field sensing structure and magnetic field sensing circuit  
Patent #
US 9,069,033 B2
Filed 06/02/2013

Current Assignee
Industrial Technology Research Institute

Sponsoring Entity
Industrial Technology Research Institute

Estimation of road inclination  
Patent #
US 9,200,898 B2
Filed 04/03/2012

Current Assignee
Scania CV AB

Sponsoring Entity
Scania CV AB

Method and apparatus for GPS based slope determination, realtime vehicle mass determination, and vehicle efficiency analysis  
Patent #
US 9,280,435 B2
Filed 12/18/2012

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Method and apparatus for GPS based slope determination, realtime vehicle mass determination, and vehicle efficiency analysis  
Patent #
US 9,384,111 B2
Filed 12/18/2012

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Using social networking to improve driver performance based on industry sharing of driver performance data  
Patent #
US 9,412,282 B2
Filed 12/21/2012

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Kinematic road gradient estimation  
Patent #
US 9,454,508 B2
Filed 01/07/2013

Current Assignee
Ford Global Technologies LLC

Sponsoring Entity
Ford Global Technologies LLC

Method and apparatus for 3D accelerometer based slope determination, realtime vehicle mass determination, and vehicle efficiency analysis  
Patent #
US 9,489,280 B2
Filed 12/21/2012

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

AUTOMOTIVE CONTROL UNIT PROGRAMMED TO ESTIMATE ROAD SLOPE AND VEHICLE MASS, VEHICLE WITH SUCH A CONTROL UNIT AND CORRESPONDING PROGRAM PRODUCT  
Patent #
US 20160332633A1
Filed 07/11/2014

Current Assignee
C.R.F. Societa Consortile Per Azioni

Sponsoring Entity
C.R.F. Societa Consortile Per Azioni

Static road gradient estimation  
Patent #
US 9,517,774 B2
Filed 01/07/2013

Current Assignee
Ford Global Technologies LLC

Sponsoring Entity
Ford Global Technologies LLC

Vehicle performance based on analysis of drive data  
Patent #
US 9,527,515 B2
Filed 01/25/2016

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Automatic incorporation of vehicle data into documents captured at a vehicle using a mobile computing device  
Patent #
US 9,563,869 B2
Filed 05/27/2014

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Method for controlling a power train of a vehicle, and corresponding device and vehicle  
Patent #
US 10,023,191 B2
Filed 01/12/2018

Current Assignee
Renault Sas

Sponsoring Entity
Renault Sas

Road gradient estimation arbitration  
Patent #
US 10,042,815 B2
Filed 01/07/2013

Current Assignee
Ford Global Technologies LLC

Sponsoring Entity
Ford Global Technologies LLC

Using telematics data including position data and vehicle analytics to train drivers to improve efficiency of vehicle use  
Patent #
US 10,056,008 B1
Filed 03/14/2014

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Method and apparatus for changing vehicle behavior based on current vehicle location and zone definitions created by a remote user  
Patent #
US 10,099,706 B2
Filed 11/18/2016

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Method and apparatus for GPS based Zaxis difference parameter computation  
Patent #
US 10,102,096 B2
Filed 07/01/2016

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Automotive control unit programmed to estimate road slope and vehicle mass, vehicle with such a control unit and corresponding program product  
Patent #
US 10,124,806 B2
Filed 07/11/2014

Current Assignee
C.R.F. Societa Consortile Per Azioni

Sponsoring Entity
C.R.F. Societa Consortile Per Azioni

Mobile computing device for fleet telematics  
Patent #
US 10,185,455 B2
Filed 10/04/2013

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Using telematics data including position data and vehicle analytics to train drivers to improve efficiency of vehicle use  
Patent #
US 10,223,935 B2
Filed 07/24/2018

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Method and apparatus for matching vehicle ECU programming to current vehicle operating conditions  
Patent #
US 10,289,651 B2
Filed 07/21/2017

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Virtual trainer for in vehicle driver coaching and to collect metrics to improve driver performance  
Patent #
US 10,417,929 B2
Filed 08/22/2016

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Method and apparatus for implementing a vehicle inspection waiver program  
Patent #
US 10,431,020 B2
Filed 12/05/2016

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Method and apparatus for changing vehicle behavior based on current vehicle location and zone definitions created by a remote user  
Patent #
US 10,507,845 B2
Filed 11/18/2016

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Virtual trainer for in vehicle driver coaching and to collect metrics to improve driver performance  
Patent #
US 10,565,893 B2
Filed 01/05/2018

Current Assignee
Zonar Systems Incorporated

Sponsoring Entity
Zonar Systems Incorporated

Multitool control system, method and medium  
Patent #
US 6,640,151 B1
Filed 12/22/1999

Current Assignee
Applied Materials Israel Limited

Sponsoring Entity
Applied Materials Israel Limited

Process and apparatus for producing a diagram of an installation comprising apparatuses supplied with gas  
Patent #
US 6,865,522 B1
Filed 09/29/2000

Current Assignee
LAir Liquide SA

Sponsoring Entity
LAir Liquide SA

System for detecting surface profile of a driving road  
Patent #
US 6,718,248 B2
Filed 06/19/2002

Current Assignee
Ford Global Technologies LLC

Sponsoring Entity
Ford Global Technologies LLC

Method and apparatus for determining a cross slope of a surface  
Patent #
US 6,389,345 B2
Filed 05/29/2001

Current Assignee
Caterpillar Incorporated

Sponsoring Entity
Caterpillar Incorporated

Controlling vehicular driving force in anticipation of road situation on which vehicle is to run utilizing vehicular navigation system  
Patent #
US 5,832,400 A
Filed 09/01/1995

Current Assignee
Nissan Motor Co. Ltd.

Sponsoring Entity
Nissan Motor Co. Ltd.

15 Claims
 1. Method of estimating the slope of a road on which there is positioned a motor vehicle provided with rotational velocity sensor means for measuring the modulus of the velocity of the vehicle and with acceleration measuring means for measuring the acceleration in the longitudinal direction of the vehicle, characterized in that it comprises the following operations:
a) providing a signal indicating the measurement of the velocity of the vehicle (V1, NCltFilV) by means of the said velocity sensor means;
b) providing a signal indicating the measurement of the acceleration in the longitudinal direction of the vehicle (AxVehFilV) by means of the said acceleration measuring means;
c) determining an estimated value of the slope of the road (StimaPend), the said determination of the estimated value comprising, subject to a condition of enabling the estimation, the processing of the velocity measurement signal (V1) and of the acceleration measurement signal (AxVehFilV) by means of a mathematical model based on the relation
A_{x}={dot over (ν
)}+g sin α
,where A_{x }represents the measurement of acceleration in the longitudinal direction of the vehicle, {dot over (ν
)} indicates the component of longitudinal acceleration of the vehicle along the road, g is the gravitational acceleration and a represents the angle of inclination of the road profile;
d) processing the velocity measurement signal (V1, NCltFilV), the acceleration measurement signal (AxVehFilV) and the estimated value of the slope (StimaPend) in such a way as to determine the sign of the velocity of the vehicle (Segno); and
e) repeating operations a) to d) recursively according to a predetermined rule, in which the sign of the velocity (Segno) determined in operation d) of a specific recursive step is used as an additional input for operation c) of a successive recursive step to update the estimated value of the slope of the road (StimaPend).  View Dependent Claims (2, 3, 4, 5, 6, 7, 8)
 9. System for estimating the slope of a road for a motor vehicle, comprising rotational velocity sensor means for measuring the modulus of the velocity of the vehicle and acceleration measuring means for measuring the acceleration in the longitudinal direction of the vehicle, positioned on board the vehicle, and processing means, connected for operation to the said velocity sensor means and to the said acceleration measuring means, and programmed for:
a) receiving a signal indicating the measurement of the velocity of the vehicle (V1, NCltFilV) provided by the said velocity sensor means;
b) receiving a signal indicating the measurement of the acceleration in the longitudinal direction of the vehicle (AxVehFilV) provided by the said acceleration measuring means;
c) determining an estimated value of the slope of the road (StimaPend), the said determination of the estimated value comprising, subject to a condition of enabling the estimation, the processing of the velocity measurement signal and of the acceleration measurement signal by means of a mathematical model based on the relation
A_{x}={dot over (ν
)}+g sin α
,where A_{x }represents the measurement of acceleration in the longitudinal direction of the vehicle, {dot over (ν
)} indicates the component of longitudinal acceleration of the vehicle along the road, g is the gravitational acceleration and α
represents the angle of inclination of the road profile;
d) processing the velocity measurement signal (V1, NCltFilV), the acceleration measurement signal (AxVehFilV) and the estimated value of the slope (StimaPend) in such a way as to determine the sign of the velocity of the vehicle (Segno); and
e) repeating operations a) to d) recursively according to a predetermined rule, in which the sign of the velocity (Segno) determined in operation d) of a specific recursive step is used as an additional input for operation c) of a successive recursive step to update the estimated value of the slope of the road (StimaPend).  View Dependent Claims (10, 11, 12, 13, 14, 15)
1 Specification
The present invention relates to a method of estimating the slope of a road on which there is a motor vehicle provided with rotational velocity sensor means for measuring the modulus of the velocity of the vehicle and with acceleration measuring means for measuring the acceleration in the longitudinal direction of the vehicle.
The information about the road slope can be used for various purposes in a vehicle.
In machines fitted with robotized automatic gearboxes, the optimization of the performance and comfort of the gear changes and of the assisted starting is a considerable problem. This is because, if the map of gear changes is calibrated to provide specific driving performance on the flat, the same performance is not guaranteed for different road slopes. However, an estimate of the slope can be used to translate the gear change map in an appropriate way in order to provide better and uniform behaviour for different road profiles.
As in the case of automatic driveaway control, the estimation of the road gradient is used to make the driving feel uniform when the vehicle is started in different slope conditions. In particular, the driveaway control algorithm provides for the generation of references for the angular velocity of the engine and of the primary gearbox shaft, as a function of the level of the accelerator pedal depressed by the user. These angular velocity profiles are supplied to a lowlevel controller which operates the actuators for the torque supplied by the engine and the torque transmitted by the clutch, in order to obtain an evolution of the actual engine speed and the speed of the primary gearbox shaft which is as close as possible to the reference value. A more uniform driving feel is obtained by using the slope information for the appropriate correction of the reference profiles of the angular velocities of the engine and of the primary gearbox shaft.
Generally, the information on the velocity of vehicles is derived from the measurement of the angular velocities of the wheels, which are measured by sensors which normally recognize the modulus of these velocities. Such a system does not provide a correct slope estimate for negative velocities. This problem is critical, for example, when it is desired to optimize the performance of the driveaway control algorithm. This is because, at the time of driving away, the sign of the vehicle'"'"'s velocity may not be compatible with the gear which is engaged and therefore a correct estimate of the slope must include reliable information on the direction of movement.
The present invention resolves the aforesaid problems by means of a method for estimating the road slope having the characteristics defined in the claims.
The method according to the invention makes it possible to recognize the direction of movement of the vehicle and to estimate the road slope without the use of special sensors apart from the accelerometer (measuring the longitudinal motion of the vehicle) and sensors of the modulus of the rotational velocity of the transmission shafts (the wheels and the primary gearbox shaft).
This is because the slope estimator according to the invention includes an algorithm for recognizing the sign of the velocity, which can be used in various applications, particularly in the field of transmission control. This algorithm provides an estimate of the direction of movement of the vehicle.
The system for estimating the road slope and the direction of movement was implemented and tested experimentally in the driveaway control system, making the slope estimation highly reliable even when the vehicle is reversing.
The invention also proposes a system for estimating the slope of a road having the characteristics defined in the claims, and a computer program product which can be loaded into the memory of a computer and which comprises portions of software code for implementing the method according to the invention, when the product is run on a computer.
Further characteristics and advantages of the invention will be made clear by the following detailed description, provided purely by way of example and without restrictive intent, with reference to the attached drawings, in which:
FIGS. 5 to 9 are graphs showing the results of a simulation conducted by using the method of
FIGS. 21 to 27 are graphs showing the results of an experimental test of the algorithm of
The schematic diagram of the whole algorithm of the method for estimating the road slope is shown in
The road slope is estimated in the block 1, using a measurement of the longitudinal acceleration AxVehFilV, a data element indicating the velocity of the vehicle Vi, and an estimate of the sign of the movement; when the vehicle starts, the slope estimate is used to determine the sign of the movement in the block 3. The sign of the velocity determined in the block 3 is entered into the estimation block 1 through a feedback block 4. As explained more fully below, the algorithm uses other information as well as the velocity and acceleration of the vehicle, as indicated by the arrows 5 and 6 in
In the latest generation (top range) of vehicles, fitted with sensors capable of measuring the sign of the velocity of the vehicle, the algorithm according to the present invention can be used as a method of analytical redundancy, in other words as information which can be used in case of a fault in the sensors. As explained more fully below, this type of application has been used experimentally in driveaway control for a test vehicle.
The signals supplied at the output and those requested at the input of the algorithm for estimating the slope and the sign of the velocity are shown in the diagram of
 a signal V1 (originating for example from a CAN message relating to an antiskid system (ASR) of the vehicle), which represents the velocity of the vehicle, which is generally but not exclusively calculated on the assumption that the wheels are not slipping, by multiplying the mean of the angular velocities of the nondriving wheels (those least subject to slipping) by their radius;
 a variable AxVehFilV, indicative of the measurement supplied by the accelerometer, preferably filtered;
 an input NCltFilV, which represents the angular velocity of a primary shaft of the gearbox (hereafter termed the “gearbox primary”) of the vehicle, preferably filtered and expressed in RPM (revolutions per minute);
 a signal GearActV, which provides the information on the gear engaged and which can only take the form of a whole number (for example, 0 indicates neutral, 1 indicates first gear, . . . , 7 indicates reverse);
 a condition ExtCondSgn, which provides information on the direction of rotation of the primary in relation to the direction of rotation of the engine in the proximity of speed synchronization during driveaway. It is used to validate or correct the estimate of the sign of the velocity, to enable the clutch to be engaged;
 an output StimaPend, which represents the estimate of the slope, supplied for example as a percentage in the form 100 tan(α);
 a variable Segno, which indicates the sign of the velocity of the vehicle and is, for example, −1 if the vehicle is reversing, 1 if the vehicle is moving forwards and 0 if the vehicle is stationary or if the sign has not yet been recognized.
Theoretically, the information on the gear engaged and on the angular velocity of the gearbox primary, supplied by the signals GearActV and NCltFilV respectively, is not indispensable for the estimation of the slope and the recognition of the sign of the velocity. However, as explained below, the signals NCltFilV e GearActV are preferably used by the algorithm according to the present invention in order to increase the speed and reliability of the estimation of the slope and the recognition of the sign.
Table 1 below shows, by way of example, the specifications for the input and output signals used in the present description.
The first level of the block diagram of a preferred embodiment of the method according to the invention for estimating the road slope and recognizing the sign of the velocity is shown in
In logical terms, the algorithm comprises four calculation blocks, 10, 20, 30, and 40:
 1. A block 10, called Processing_Or, processes the measurements of the angular velocity of the wheels V1 and of the gearbox primary NCltFilV, and supplies at its output the information on the velocity of the vehicle and a signal SottoSoglia which indicates whether the vehicle is stationary or moving. This signal processing is carried out in order to make the whole algorithm as reliable and fast as possible.
 2. A block 40, called Estimation_Enablement, supplies the signal for enabling the calculation of the estimate of slope; in particular, this signal commands the Kalman filter not to update the estimate of slope when the vehicle is moving but the direction of movement has not yet been recognized.
3. A block 20, called Kalman_Filter, supplies the information on the estimate of the road slope on the basis of the information on the velocity of the vehicle, on the sign of the movement, and on the measurement supplied by the accelerometer.
4. A block 30, called Sign_Recognition, determines the sign of the velocity of the vehicle.
The aforementioned four logical blocks must be executed sequentially at each sampling pass in the order in which they are listed.
The road slope can be estimated by using the information on the velocity of the vehicle, on the torque transmitted to the wheels and on the resistant torque. A more accurate estimate of the inclination of the road is obtained by measuring the longitudinal component of acceleration of the vehicle, by means of a suitably orientated accelerometer (not shown). This approach is the one used by the estimation algorithm according to the invention.
Disregarding the contribution due to the pitching of the vehicle and the effect of the lateral acceleration, the measurement of the longitudinal accelerometer can be described as the sum of two contributions:
A_{x}={dot over (ν)}+g sin α (1)
where {dot over (ν)} denotes the longitudinal acceleration component of the vehicle along the road profile, g is the gravitational acceleration, and α represents the angle of inclination of the road.
The problem of estimating the road slope on the basis of the measurements made by the accelerometer and the velocity sensor of the vehicle was tackled by using a Kalman filter, based on a mathematical model described below. The measurement supplied by the accelerometer is represented by a model similar to equation (1), which additionally describes the disturbance in the sensor as coloured noise x_{3}, produced by filtering a white noise v_{2 }with a firstorder filter having a pole p_{3}:
{dot over (x)}_{1}=−gx_{2}+A_{x}+x_{3 } (2)
{dot over (x)}_{3}=−p_{3}x_{3}+ν_{2 } (3)
where x2=sin α, x_{1}=v. In turn, the variation of slope is modelled with a firstorder dynamic:
{dot over (x)}_{2}=−p_{2}x_{2}+ν_{1 } (4)
By grouping equations (2)(4) in a matrix, we obtain the model in statespace form
where the input u consists of the accelerometer measurement and m represents the measurement of the velocity of the vehicle corrupted by a Gaussian white noise w.
The Kalman filter projected on to the aforementioned mathematical model is described by the block diagram of
{dot over ({circumflex over (x)})}=A·{circumflex over (x)}+B_{u}·u+K(m−C·{circumflex over (x)})
ŷ=[0 1 0]·{circumflex over (x)} (7)
Table 2 below shows the numerical values of the Kalman filter designed ad hoc to estimate the slope in a specified test vehicle, in order to improve the driveaway control performance. These numerical values are therefore provided purely by way of example. The same numerical values were used for the experimental tests described below.
An example of a simulation providing an indication of the speed with which the Kalman filter described above estimates the road slope is described below. It is important to emphasize that an increase in the speed of the estimation of the variations of slope by the Kalman filter is accompanied by an increase in its sensitivity to disturbances such as the pitching of the vehicle or the coupling of the lateral acceleration to the measurement supplied by the longitudinal accelerometer.
In particular, the simulation assumes a point mass which is initially stationary, and which at the instant t=1 s is put into motion with a constant acceleration of 2 m/s^{2 }up to the instant t=7 s, after which the mass proceeds at constant velocity.
In order to show only the speed of convergence of the slope estimate supplied by the Kalman filter, the response dynamics of the sensors and any noise have been disregarded; the Kalman filter has therefore been made to operate with ideal velocity and acceleration. The time required for the estimator to supply 90% of the exact value of the slope is approximately 3.5 s (see
A high convergence velocity of the Kalman filter would result in lower robustness of the whole algorithm, since the disturbances affecting the accelerometer would affect the procedure for recognizing the direction of movement of the vehicle. In fact, the estimation of the slope requires correct information on the sign of the movement, which in turn is estimated by a specific subsystem. In this case, the procedure for recognizing the sign of the velocity uses the value of slope estimated at the instant at which the start of the vehicle'"'"'s movement is detected, and therefore the disturbances of the accelerometer which affect the estimation of the direction of movement are those which act before the aforementioned instant and are not completely attenuated by the Kalman filter.
The determination of the sign of the velocity requires a certain period of time, during which the Kalman filter does not update the estimate of slope; the acceleration of the movement of the vehicle during the phase of recognition of the sign is therefore not interpreted as a variation of slope and therefore does not generate errors in the estimation of the slope.
The Kalman filter described above showing the differential equations in continuous time was discretized, thus producing the block Kalman_Filter 20, the block diagram of which is shown in
{circumflex over (x)}(k)=A_{d}·{circumflex over (x)}(k−1)+B_{d}·u_{d}(k−1)
ŷ(k−1)=C_{d}·{circumflex over (x)}(k−1) (8)
where the matrices A_{d}, B_{d }and C_{d}, which are implemented by respective transfer blocks indicated respectively by 70, 80 e 90 in
The information on the road slope is supplied by means of the signals pend and StimaPend, which represent, respectively, the sine of the angle of inclination of the road and the slope as a percentage:
pend=sin α (9)
StimaPend=100·tan α (10)
A saturation block 100 has been inserted into the system of
StimaPend=100·tan(a sin(u))
The block 20 which models the Kalman filter is enabled by the signal EnStima: if this enabling signal takes a value of one, the estimation of the slope is enabled; conversely, when EnStima has a zero value, the estimate is not updated and a constant value, equal to the slope estimated at the instant of disabling, is supplied at the output. The Kalman filter is disabled only during the phase of recognition of the sign of the velocity, in other words in the phase in which it has been detected that the vehicle is in movement, but the direction of movement is not yet known.
The Kalman_Filter block 20 also comprises a block 120, called State_Initialization, which calculates the value X0 which the state vector of the Kalman filter must have at the instant of activation of the block. The initialization procedure is shown in
This type of initialization is useful if the measurement supplied by the accelerometer is reliable at the instant of activation of the system.
The method according to the invention comprises an algorithm which estimates the sign of the velocity of the vehicle, mainly by using the estimate of the slope with the vehicle stationary and the measurement of the longitudinal acceleration. It must be stressed from the outset that the information on the sign of the velocity supplied by the algorithm is an estimate, and is therefore not absolutely reliable.
The basic principle used to estimate the sign of the movement of the vehicle is the calculation of the velocity on the basis of information which also enables the sign of the velocity to be determined. As mentioned above, a known way of calculating the velocity including the information on the sign is to use the simplified dynamic model
J_{eq}{dot over (ω)}_{wheels}=C_{T}−C_{R}−Rmg sin α (11)
where J_{eq }represents the equivalent inertia of the vehicle with respect to the wheels; C_{T }is the torque transmitted to the wheels; C_{R }indicates the resistant torque at the wheels, including the contribution of the brakes, the friction of the tyres and the aerodynamic resistance of the vehicle; R represents the radius of the wheels; m is the mass of the vehicle; and ω_{wheels }indicates the mean angular velocity of the nondriving wheels. The second member of equation (11) represents the balance of the torques acting on the vehicle, which, when integrated and divided by the equivalent inertia, provides an estimate of the angular velocity of the wheels. In order to determine the sign of ω_{wheels }correctly, it is not essential to calculate the exact value of the second member, but it is sufficient for its sign to have an average degree of accuracy. The variables present in the relation (11) which are known with the greatest uncertainty are the transmitted torque C_{T}, the resistant torque C_{R }and the mass of the vehicle. This is because the torque C_{T }in the driveaway phase is derived from the torque transmitted by the clutch, which is subject to wear, and is therefore known with a relative uncertainty of up to 50%; the torque C_{R }includes the contribution of the brakes, which in general is neither measured nor estimated; the total mass can vary by several hundreds of kilograms.
Since it is desired to discriminate between the positive and negative velocities, the uncertainty with which the variables present in the relation (11) are known becomes critical for low values of angular acceleration {dot over (ω)}_{wheels}, in other words when the uncertainty can lead to an error in the sign of the total result of the balance of torques (the second member of the simplified model of the relation (11)).
A more reliable estimate of the direction of movement of the vehicle can be obtained by calculating the velocity of the vehicle by means of the model
{dot over (ν)}=A_{x}−g sin α (12)
which uses the measurement of the longitudinal accelerometer and an estimate of the road slope. The calculation of the velocity with this model is more accurate, since there are fewer variables in operation and their accuracy is greater. The estimate of slope used in equation (12) is that supplied by the Kalman filter, which in turn uses the accelerometer measurement and the information on the sign of the velocity. Thus, in order not to create critical loops in the whole algorithm, it is necessary to store the value of the slope as soon as the vehicle starts to move, and to use this value in equation (12) to determine the sign of the velocity. Until the sign is known, the slope estimation algorithm uses a zero value of velocity.
With this method according to the invention also, the most critical situations are those characterized by very small accelerations, since the uncertainties present can potentially cause a significant and persistent error in the sign of the second member of equation (12) and consequently in the estimation of the sign of the velocity. However, the critical band is much narrower than in the approach which uses the information on the torque, and in any case this problem is considerably mitigated by a number of arrangements explained below.
Using a constant value of slope, equal to the value estimated when the vehicle starts to move, is equivalent to assuming that there are no significant variations of slope throughout the period of recognition of the sign. This assumption is reasonable if we consider the delay with which the vehicle is subjected to a variation of slope (due to its physical dimensions) and the speed of recognition of the sign by the algorithm. In particular, the period of recognition of the sign is of the order of a tenth of a second for high values of acceleration {dot over (ω)}_{wheels}, while it is generally less than one second for low values of acceleration.
A simple way of determining the direction of the movement is to calculate the velocity of the vehicle by integrating the relation (12) and testing the sign as soon as the modulus of the calculated velocity exceeds a certain predetermined threshold. The presence of the threshold is necessary in order to make the algorithm robust when exposed to any disturbances acting on the accelerometer or uncertainties in the estimation of the slope.
However, the phase of determining the sign executed by the algorithm implemented in a preferred embodiment is somewhat more complicated, but is also faster and more robust than the simple test of the sign of the calculated velocity. This is because the algorithm calculates two angular velocities
where the instant t_{0 }represents the moment of detection of the start of the vehicle'"'"'s movement; and O_{r}_{—}_{proc }indicates the information on the modulus of the angular velocity of the wheels supplied by a subsystem which processes in an appropriate way the measurements originating from the sensors of the motor vehicle. The angular velocities of formulae (13) and (14) are used to calculate two variables called residuals:
ResP=O_{r}_{—}_{proc}−O_{r}_{—}_{calcP } (15)
ResN=−O_{r}_{—}_{proc}−O_{r}_{—}_{calcN } (16)
which are used directly for the determination of the sign. The sign is estimated as soon as one of the two residuals exceeds a certain predetermined threshold; at this instant, the two residuals are compared with each other, and if the modulus of ResN is greater than ResP the estimated sign will be positive; if the converse is true, the sign will be negative.
The models mentioned above with reference to
The highest level of the block 30 which estimates the sign of the velocity of the vehicle is shown in
The Sign_Recognition block 30 comprises logical subblocks 310, 320, 330, 340 and 350. The block 310 implements the HOLD function indicated by 220′ in
The residuals ResP and ResN are calculated by means of the two identical Calculation1 and Calculation2 blocks, 330 and 340, which implement, respectively, the formulae (15) and (16) (see the diagram of the Calculation2 block 340 in
The determination of the sign of the movement of the vehicle on the basis of the calculated residuals and a set of other information is executed by the Logic_Unit block 320, a state machine consisting of three states, each of which corresponds to a possible value of the sign output variable.
The block 320 which implements the state machine is shown in
 a signal gsinalfa, indicating the slope stored when the vehicle starts to move;
 a signal NValidSegno, which indicates when the initial slope is below a certain predetermined threshold or when the vehicle is moving during the initialization of the system;
 a signal GearActV, representing the gear engaged; and
 an external condition signal ExtCondSgn: if this is 1, it forces the sign to be compatible with the engaged gear.
When the system is started, the state machine is initialized to the state Start, corresponding to the stationary state of the vehicle; it changes from this state when the vehicle is found to be moving.
As mentioned previously, the most critical situations for the estimation of the sign of the velocity are those characterized by starts with very low acceleration, when the uncertainties present may cause an error in the sign of the second member of equation (12). However, the uncertainties in the measurements and in the estimates used are small during the phase in which the vehicle is driven away, and therefore the probability of incorrect estimation of the sign is relatively low. To make the algorithm more reliable, it was decided to ignore the calculation of the residues when the estimated slope was below a certain threshold (which was set at 3% for the driveaway control of the test vehicle in question) and to supply in this case a sign of velocity compatible with the engaged gear. This is because it is important for the driveaway control algorithm to recognize rapidly the reversing of the vehicle with medium to high acceleration, in other words on medium to high road slopes, and therefore the aforementioned threshold makes the algorithm more robust, since it is improbable that the vehicle is reversing in the presence of a small slope. When the road slope is below the predetermined threshold and the vehicle is not in neutral, the variable NvalidSegno takes the value 1. This is because, during the testing phase, it was found that the acceleration of the vehicle'"'"'s movement in the presence of significant slopes was always such that a correct estimate of the sign of velocity was obtained.
The variable NValidSegno also takes the value of 1 when the vehicle is moving during the phase of initialization of the system.
If the vehicle is in neutral, the sign of the velocity is determined without consideration of the calculation of the residuals, but simply on the basis of the information on the slope at the start of the manoeuvre. This makes the sign estimation faster and more robust, since in the first place it is unnecessary to wait for the residuals to exceed the predetermined threshold, and in the second place the disturbance due to pitching has less effect. There is another reason, explained below, why it is more reliable to evaluate the sign of the velocity by only considering the initial slope when the vehicle is in neutral.
Except in the cases where the vehicle is in neutral or the initial slope is below the predetermined threshold, all the other situations are controlled by determining the sign on the basis of the calculation of the residuals: as soon as one of the residuals exceeds a certain threshold, the algorithm assigns the value 1 to the output sign if the modulus of ResP is smaller than the modulus of ResN, while it assigns the value of −1 to it if the modulus of ResP is greater than the modulus of ResN. Increasing the value of the threshold used for testing the residuals increases the robustness of the measurement of the angular velocity in the presence of disturbances, but also reduces the speed of recognition of the direction of movement.
To control situations in which the movement of the vehicle is reversed but the calculation of the residuals is not enabled, or the calculation supplies an incorrect result, the external condition ExtCondSgn is used; if this has a value of 1, it forces the sign to be compatible with the engaged gear.
The conditions for changing from one state to another explained above, indicated as CONDITION 1, CONDITION 2, CONDITION 3, CONDITION 4 and CONDITION 5 in
CONDITION 1: the proposition (SottoSoglia=0) AND (gear=0 OR NValidSegno=1 OR (ResP≠ResN AND (ResP>SogliaRes OR ResN>SogliaRes))) is true and the proposition ((ResP>ResN AND NValidSegno=0 AND gear≠0) OR (NValidSegno=1 AND gear=7) OR (gSinAlfa>0 AND gear=0)) is true.
CONDITION 2: the proposition (SottoSoglia=0) AND (gear=0 OR NValidSegno=1 OR (ResP≠ResN AND (ResP>SogliaRes OR ResN>SogliaRes))) is true and the proposition ((ResP>ResN AND NValidSegno=0 AND gear≠0) OR (NValidSegno=1 AND gear=7) OR (gSinAlfa>0 AND gear=0)) is false.
CONDITION 3: the proposition SottoSoglia=0 AND ((gear≧1 AND gear≦6 AND ExtCondSgn=1)) is true.
CONDITION 4: the proposition SottoSoglia=0 AND (gear=7 AND ExtCondSgn=1) is true.
CONDITION 5: the proposition SottoSoglia=1 is true.
As explained above, for reasons of robustness the sign of the velocity is set in a way which is compatible with the engaged gear when the slope at the start of the manoeuvre is below a certain threshold and the vehicle is not in neutral.
This method of determining the sign is also used on the initialization of the algorithm, if the velocity sensors do not supply a zero value. This situation can occur if the velocity sensors require a certain stabilization time, which has not elapsed at the instant of initialization of the system; or if the vehicle is actually moving at the moment when the system is initialized. This is because, when the system is started, the sign of the movement cannot be reliably determined on the basis of the calculation of the residuals if the sensors have not been stabilized or if the vehicle is already moving.
The diagram of the Slope Threshold Test block 350 for supplying the variable which indicates when the sign has to be determined on the basis of the gear engaged is shown in
The algorithm for estimating the road slope and the sign of the velocity is based on two pieces of information: the longitudinal acceleration, measured directly by an accelerometer, and the modulus of the velocity of the vehicle, found by various methods according to the situation. The vehicles are generally fitted with four sensors which measure the angular velocity of the four wheels and with a sensor which measures the angular velocity of the gearbox primary.
These sensors show a discontinuity at low velocity, since they supply a zero output for velocities below a certain threshold and start to make measurements for velocities above the threshold.
The reading threshold of the sensor of the gearbox primary of the test vehicle in question is approximately 15 rad/s, which, when converted to the wheel value by dividing it by the transmission ratio (τ=13.7), becomes of the order of 1 rad/s; however, the reading threshold of the sensors measuring the angular velocity of the wheels is approximately 2.2 rad/s. It can therefore be deduced that the measurement of the sensor located on the gearbox primary makes it possible to detect lower velocities of the vehicle than those which can be measured directly on the wheels. It is important to point out that the experimental trials conducted to test the estimation algorithm show that the direction of movement of the vehicle is recognized, in the great majority of cases, when the angular velocity of the wheels is below the reading threshold of the corresponding sensors.
To make the estimation algorithm as reliable, fast and accurate as possible, the velocity of the vehicle is determined by multiplying the radius of the wheels by the mean angular velocity of the nondriving wheels when the corresponding sensors supply a measurement above the aforementioned threshold, while the angular velocity of the gearbox primary and the information on the transmission ratio are used when the sensors measuring the velocity of the wheels supply a zero output.
However, it must be emphasized that the velocity of the wheels found from the primary velocity measurement is affected by the fact that the dynamics of the drive line and the slipping of the driving wheels are disregarded. To diminish these disturbances, it was decided to filter the measurement of the angular velocity of the gearbox primary of the test vehicle with a lowpass filter (this filter has a cutoff frequency of 2 Hz for the driveaway control of the test vehicle).
The block 10 called Processing_or, shown in
The outputs are the modulus of the velocity of the vehicle V_lim in m/s, the modulus of the angular velocity of the gearbox primary Or_proc in rad/s, and the signal SottoSoglia which indicates whether the vehicle is stationary or moving.
In particular, the information as to whether the vehicle is stationary or moving, supplied by the signal SottoSoglia, is found by comparing the velocity of the vehicle with a predetermined threshold (SogliaLetturaOr). The velocity of the vehicle used for the comparison is calculated on the basis of the measurement of the sensor on the gearbox primary, since it is essential that the minimum detectable velocity is the smallest which can be obtained in order to start the sign determination procedure as soon as possible. The only situation in which the measurement of the velocity of the gearbox primary cannot be used is when the vehicle is in neutral, since the drive line is disconnected from the wheels. The discrimination between the two cases is provided by means of a block 410 called Switch2. It is also necessary for the velocity signal used for the aforesaid comparison not to be excessively filtered, in order to avoid the concealment of reversals of the movement by cutting off a fast double crossing of the threshold. The latter case may occur, for example, when a gear change from reverse to first or first to reverse is made while the vehicle is still moving.
The output signal Or_proc represents the information on the angular velocity of the wheels which must be used for the calculation of the variables ResP and ResN. This velocity must have the same reading threshold as the velocity signal used to detect the starting of the vehicle; it is therefore derived from the latter signal after appropriate filtering applied to eliminate any oscillations due to the drive line dynamics. A rise in the cutoff frequency of the filter increases the speed of the phase of recognizing the sign of velocity, but also decreases the robustness in the presence of oscillations and disturbances in the measurement of the angular velocity of the gearbox primary. The aforementioned filtering is carried out by a filter block 420 (having a cutoff frequency set at 1 Hz), which is “reset” to the value of the measurement whenever the angular velocity of the gearbox primary crosses the threshold SogliaLetturaOr. This reset makes it possible to eliminate any errors in the estimate of the sign of velocity due to oscillations or sudden peaks which may occur before the velocity of the gearbox primary crosses the threshold. These peaks, which frequently occur in gear changes from first to reverse or from reverse to first with the vehicle in motion, would not be entirely attenuated by the filter block in the absence of the reset, and would therefore affect the calculation of the residuals and might cause an error in the estimation of the direction of movement of the vehicle.
The output V_lim represents the velocity of the vehicle, which is used to estimate the slope with the Kalman filter, and which is determined directly from the measurement of the velocities of the wheels, when the sensors on the wheels measure velocities above the corresponding reading threshold or when the vehicle is in neutral; or from the velocity calculated by the sensor on the gearbox primary, for velocities below the reading threshold of the sensors on the wheels.
A saturation block 430, set between zero and the value SogliaLetturaVel, serves to eliminate any peaks in the measurement of angular velocity of the gearbox primary, when the measurements of velocity of the nondriving wheels are below the threshold of the corresponding sensors.
An example of the operation of the algorithm for estimating the slope and determining the sign of the movement will now be described. In particular, the experimental test relates to the driving away of the aforementioned test vehicle with reference to Table 2, with the accelerator pedal depressed by approximately 100%, from a stationary condition of the vehicle and in the presence of a slope of approximately 20%. The vehicle initially reverses for approximately one second, after which it starts to move forwards with high acceleration and finally halts on the flat.
The variable Or_proc represents the angular velocity of the wheels used to calculate the residuals, which is substantially found from the measurement of the angular velocity of the gearbox primary filtered with a firstorder filter having a pole of 2 Hz; the signal sensor_{gearbox }represents the angular velocity of the wheels found from the measurement of the angular velocity of the gearbox primary filtered with a firstorder filter having a pole of 10 Hz; the variable sensor_{wheels }is the mean of the measurements of the angular velocity supplied by the sensors located on the nondriving wheels.
The three signals shown in
The measurement supplied by the accelerometer is shown in
The evolution of the estimated slope with time is shown in
By comparing the information on the angular velocity of the wheels (
The sign is determined by comparing the moduli of the residuals at the instant at which one of the two exceeds the set threshold value SogliaRes (the dotted line in
As mentioned above, the estimation algorithm which has been described was used as a method of analytical redundancy in the driveaway control of the test vehicle, in order to increase the reliability of the information on the direction of movement (and consequently the information on the road slope). This is because measurements of the sign of velocity of the wheels are present in the aforesaid vehicle, being supplied by the controller of the antiskid system, and in case of fault these are ignored and replaced with the estimated information.
The algorithm for estimating the road slope and the direction of movement of the vehicle which was designed for use in the driveaway control of the test vehicle had the parameter values shown in the tables below. In particular, Table 4 contains the values of the thresholds, the saturations, the pole of the filter, and conversion tables used to process the various values needed for the correction operation of the whole algorithm.
Table 5 shows the values of the state matrices of the Kalman filter in continuous time; and finally Table 6 contains the characteristic parameters of the vehicle, which are used in the estimation algorithm.
Clearly, provided that the principle of the invention is retained, the details of construction and the forms of embodiment can be varied widely from what has been described and illustrated, without departure from the scope of the invention. For example, the sign detection algorithm can be implemented on the basis of the relation (11), instead of the relation (12). In this case, the angular velocities O_{r}_{—}_{calcP }and O_{r}_{—}_{calcN }are defined by the following relations:
where J_{eq }represents the equivalent inertia of the vehicle with respect to the wheels; C_{T }is the torque transmitted to the wheels; C_{R }indicates the resistant torque at the wheels, including the contribution of the brakes, the friction of the tyres and the aerodynamic resistance of the vehicle; R represents the radius of the wheels; m is the mass of the vehicle; and ω_{wheels }indicates the mean angular velocity of the nondriving wheels.
Finally, it is emphasized that the method according to the invention can be used in various applications, particularly in the field of transmission control.