 |
Description  |
|
|
This invention relates to vehicle suspension systems and more particularly
to actuator control of variable force semi-active suspensions. The subject of this application is related to following patent applications: 07/702,873 now U.S. Pat. No. 5,276,621 entitled "Semi-Active Suspension System with Electromechanical Damping"
and 07/702,875 now U.S. Pat. No. 5,434,782 entitled "Suspension System State Observer," both filed concurrently with this specification and assigned to the assignee of this invention. The disclosures of patent applications 07/702,873 and 07/702,875
are incorporated into this document by reference.
BACKGROUND OF THE INVENTION
In the field of vehicle suspensions, the phrase "quarter car suspension" refers to the components of the vehicle suspension relating to one of the four wheels of the typical automotive vehicle. These components include the particular wheel with
a tire that is in contact with the road, a spring that transfers the road force to the vehicle body (sprung mass) and suspends the vehicle body, and a damper or actuator that reduces undesirable relative movement between the vehicle body and wheel. The
complete suspension system of an automotive vehicle comprises four quarter car suspensions.
In recent years, vehicle manufacturers have dedicated significant effort to developing suspension systems responsive to the driving conditions of the vehicle. This effort is triggered by desire to incorporate the best features of soft and stiff
suspension systems into a single vehicle suspension system. The best feature of a soft vehicle suspension is the smooth ride it provides for the vehicle passengers. The best feature of a stiff vehicle suspension is the increased handling performance it
provides for the vehicle.
The theory of semi-active suspension systems is to selectively switch between stiff suspension and soft suspension in response to the particular driving conditions of the vehicle. Selection between stiff suspension and soft suspension may be
obtained by altering the damping force of the suspension system, e.g., a greater damping force for a stiffer suspension and a lower damping force for a softer suspension. With correct control of suspension damping force, a vehicle can provide both
optimum driving comfort and optimum handling performance. Semi-active suspension systems (along with active systems) can be referred to as variable force suspension systems.
Difficulties in designing variable force suspension systems lie partially in system controls. For example, the state of the suspension at each wheel of the vehicle is affected not only by road disturbance on the wheel, but by the rigid body
characteristics of the vehicle body.
What is desired is a suitable system for integrating four semi-active suspension systems into a vehicle to provide optimum vehicle handling and ride comfort.
SUMMARY OF THE PRESENT INVENTION
This invention provides a means and method of controlling four quarter car semi-active suspension systems in a vehicle. The invention controls the variable force damper of each quarter car suspension in response to both the effect of the road on
the particular wheel for that quarter car suspension and the rigid body characteristics of the vehicle body. The control method of this invention enables the semi-active suspension system to eliminate the effects of road disturbances and rigid body
motion on vehicle ride and handling. The rigid body motions, such as body heave, pitch, roll and yaw caused by forces transferred to the vehicle body through the suspension and like motions caused by forward and lateral accelerations of the vehicle, are
reduced, increasing performance in vehicle ride and handling.
The method of control of this invention comprises the steps of (i) developing a quarter car command for each quarter car suspension in response to the state of that quarter car, (ii) developing a semi-rigid body command in response to a signal
representative of the motion-state of the suspended vehicle body, and (iii) developing a force command controlling force between sprung and unsprung masses in each quarter car suspension, the force command equal to the greater of the quarter car command
and the semi-rigid body command.
The state of each quarter car can be represented by at least one signal representing at least one of a group of parameters, including absolute position of the sprung mass, absolute position of the unsprung mass, absolute velocity of the sprung
mass, absolute velocity or the unsprung mass, relative position of the sprung and unsprung masses and relative velocity of the sprung and unsprung masses. The signal representing the state of the vehicle body may be any signal indicating body motion, or
impending body motion, including, signals indicating forward acceleration, lateral acceleration, braking, steering wheel angle, and vehicle velocity.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is an equivalent schematic diagram of a variable force quarter car suspension.
FIG. 2 is an equivalent schematic diagram of a quarter car suspension with a controller.
FIG. 3 is an example damper for a quarter car semi-active suspension system.
FIG. 4 is a schematic diagram of example circuitry for controlling the damper in FIG. 3.
FIG. 5 is a flow diagram showing a preferred control structure for developing a quarter car semi-active suspension command.
FIG. 6 is a sliding surface plot relating to the observer of FIG. 5.
FIG. 7 is a graph of the output characteristics of the controller 200 shown in FIG. 5.
FIG. 8 is a flow diagram for a computer implementation of the control structure shown in FIG. 5.
FIG. 9 is a schematic diagram showing four quarter car semi-active suspension systems integrated into one vehicle, for use with the control method of this invention.
FIG. 10 is a more detailed diagram of one example of the apparatus of this invention.
FIG. 11 is a flow diagram of one implementation of the control method of this invention.
FIG. 12 is a diagram of a hydraulic variable force actuator for use with this invention.
FIG. 13 is a damping force diagram of a two state hydraulic damper.
FIG. 14 is a graph of the output characteristics of the controller of FIG. 5 when used with a two state hydraulic damper.
DETAILED DESCRPTION OF THE INVENTION
A suspension system of the type of which four may be installed in a vehicle with the control method of this invention may be understood with reference to the model diagram of FIG. 1. In the figure, reference numeral 12 generally designates the
sprung mass, having mass M.sub.s, which is the vehicle body supported by the suspension. The sprung mass has a position, x.sub.2, and a velocity, x.sub.2 ', both represented by line 14. The sprung mass 12 is supported by the spring 16, having a
constant k.sub.s. The spring 16 is also connected to the unsprung mass 24, which represents the vehicle wheel. The unsprung mass 24, having mass M.sub.u, has a position, x.sub.1, and a velocity, x.sub.1 ', both represented by line 26. The tire of the
vehicle is modeled as a spring 28, having a spring constant k.sub.u. The road is represented by reference numeral 32 and affects a displacement R (line 30) on the tire 28.
Variable force between the sprung and unsprung masses 12 and 24 is provided in the suspension system by actuator 22. Actuator 22 may be an adjustable damper or an actuator capable of both damping and providing a force independent of damping on
the suspension system. Actuator 22 may be an electromechanical machine, including a linear electromechanical machine, hydraulic shock with a flow control or bypass valve, or any other means of providing variable force to the suspension. The actuator 22
is attached between the unsprung mass 24 and a rubber bushing 18, which is modeled as a nonlinear spring in parallel with a damper. Rubber bushing 18 is similar to bushings used in engine mounts and is optional. If the rubber bushing 18 is omitted, the
actuator 22 is attached directly to the sprung mass 12. In general, the actuator 22 exerts a force on the unsprung mass 24 and an equal and opposite force on the sprung mass 12, through rubber bushing 18, in proportion to the relative speed of the
sprung and unsprung masses 12 and 24 and/or an input control signal.
In the suspension system, the road 32 affects a displacement R on the tire 28, which in turn applies a force on the unsprung mass 24. The unsprung mass 24 transfers force to the spring 16 which in turn applies force on the sprung mass 12. The
actuator 22 applies force on the sprung mass 12 (through bushing 18, if used) and unsprung mass 24 in the direction opposite the relative direction of travel of the two masses.
The bushing 18 is optional but may be preferable to help reduce the effect of high frequency road surface disturbance on the system. In the model of the suspension system set forth below, the effect of the rubber bushing is ignored and accounted
for as a system uncertainty. Suspension systems of the type represented by FIG. 1 are easily implemented by those skilled in the art.
Referring to FIG. 2, apparatus for control of the suspension system shown in FIG. 1 includes the controller 34 comprising a state estimator 35 (also referred to as the observer) and the actuator command generator 36. The controller 34 receives a
signal on line 37 representative of the relative system state. The signal on line 37 may be developed by actuator 22 or an independent sensor, such as a LVDT (not shown). In response to the signal on line 37 and previous state estimations, the state
estimator 35 estimates the current suspension system state (a more detailed explanation is set forth below with reference to FIG. 5). The actuator command generator 36 develops a force command, controlling the force between sprung and unsprung masses 12
and 24 applied by actuator 22, in response to the state estimation and provides that command on line 38 (an example implementation is also explained below with reference to FIG. 5).
Actuator 22 may be a damping electromechanical actuator which variably generates and dissipates power or a damping hydraulic actuator with means for variable flow control.
The bus 33 represents signals of various other vehicle parameters that may, at times, be taken into account to develop the actuator command at box 36. These signals may include semi-rigid body characteristics of a vehicle body and example
implementations according to this invention are set forth in detail below.
Referring to FIG. 3, one example of the actuator 22 may be the unit 22a, including an electromechanical machine 42. In the figure, the electromechanical machine 42 comprises a linear to rotary motion converter 60 and a rotary multi-phase
alternator 52. The linear to rotary motion converter 60 includes a ball screw cage 58, hollow connector 62, screw 56 and lower connector 66. The rotary multi-phase alternator 52 is rotatably mounted through bearings 44 and 54 to the upper connector 40. The lower connector 66 is mounted to the unsprung mass 24 (FIG. 1) and the upper connector 40 is mounted to the sprung mass 12 (FIG. 1), through rubber bushing 18, if used.
Through the relative movement of the sprung mass 12 and the unsprung mass 24 acting on the connectors 40 and 66, the ball screw 56 is forced to rotate, rotating the rotary multi-phase alternator 52 and creating electric potential on lines A, B,
and C, which are connected to the rectifier and chopper apparatus 48. In response to the controller 50, which generates a pulse width modulated control signal, corresponding to the force command, on line PWM, the rectifier and chopper 48 selectively
dissipates the power generated by alternator 52 through load resistor 46, providing the desired damping force for actuator 22a. During high frequency movements of the unsprung mass 24, e.g., on a very bumpy road, the rubber bushing 18 (FIG. 1)
attenuates the inertial effect of actuator 22 on the suspension system performance. Optionally, the actuator 22a may be used as a brushless DC motor by including three hall effect sensors and used with an inverter circuit. Another option is to use a
linear electromechanical actuator.
The rectifier and chopper 48 and the controller 50 can be better understood with reference to FIG. 4. Coils 120, 121, and 122 of the alternator 52 are connected to a rectifier bridge 140 which rectifies the three phase voltage on lines A, B, and
C. In response to a control signal on line PWM, the circuit comprising operational amplifier 136, transistors 130 and 146, and resistors 132, 134, 138, 144, 148, and 150 control MOSFET 128, selectively closing a DC circuit between resistors 46 and 142,
dissipating the power generated by alternator 52. The duty cycle of the signal on line PWM determines the amount of damping force. Example values for the resistors and capacitor are as follows: resistor 46, 1.67 .OMEGA., 150 W; resistor 142, 2
m.OMEGA.; resistor 124, 50 .OMEGA.; capacitor 33, 0.1 uF; resistors 134,138, and 150, 10 K; resistor 148, 100 K; resistor 132, 1.5 K; and resistor 44, 100 .OMEGA..
Line I may be implemented as an option to provide a damping force feedback loop for electromechanical implementations. There will probably be sufficient inductance in the circuit of alternator 52, rectifier bridge 140 and load resistor 46 that
the duty cycle modulation of MOSFET 128 produces an average DC current with a small ripple. If so, the current signal I read into microcomputer 174 is already averaged. If any additional averaging is required, it can be done with a standard digital
averaging algorithm in microcomputer 174 applied to successive values of I.
It is desirable to detect the relative velocity of the sprung and unsprung masses 12 and 24, otherwise known as the rattle space velocity, or the relative position of the sprung and unsprung masses 12 and 24. Either implementation is acceptable.
The rattle space velocity may be determined a variety of ways. One implementation is to determine the frequency of the zero crossings of the voltages on lines A, B, and C.
The circuit comprising transformer 152, operational amplifier 158, resistors 154, 156, 160, 170, and 172, capacitors 162 and 164, and zener diodes 166 and 168 provide a pulse to the microcomputer 174 on line F with every zero crossing of the
voltage between lines B and C. Preferably, identical circuits are connected between lines A and C and between lines A and B to provide zero crossing pulses on lines F' and F", respectively. The frequency of the signals on lines F, F' and F" determines
the magnitude of the rattle space velocity and the direction of the rattle space velocity is determined by the order of the signals on lines F, F' and F". The calculation of the rattle space velocity is performed by microcomputer 174 through a computer
routine easily implemented by one skilled in the art. Example values for the capacitors and resistors are: resistors 154, 156 and 170, 10 K; capacitor 162, 0.0015 uF; capacitor 164, 0.33 uF; resistor 160, 3 K; and resistor 172, 470 K. A more detailed
description of the actuator 22a and related circuitry is set forth in U.S. Pat. No. 4,815,575, to Murty, assigned to the assignee of this invention, and will not be set forth herein.
If the relative position of the sprung and unsprung masses 12 and 24 is to be determined, an LVDT-type sensor (not shown) is attached between the sprung and unsprung masses 12 and 24. The LVDT sensor provides an output signal linearly related to
distance between the sprung and unsprung masses and the output signal is provided to an A/D converter (not shown), the output of which is connected to the microcomputer 174 for processing. Any suitable position sensor may be used as an alternative.
One example of a means for determining the state of the suspension system includes the observer 195 shown in FIG. 5 including a linear Luenberger term and a nonlinear signum term. The observer 195 is also set forth in copending patent
application Attorney Docket No. G-7238, entitled "Suspension System State Observer," filed concurrently with this application, assigned to the assignee of this invention, and incorporated into this specification by reference. The observer 195 estimates
the entire system state X.sup.e' (comprising x.sub.1.sup.e, x.sub.1.sup.e', x.sub.2.sup.e and x.sub.2.sup.e') and computes the estimated relative system state, y.sup.e (t).
The observer 195 estimates the state of the quarter car suspension system according to the model:
where A and B are standard model matrices for a suspension system with control, X.sup.e is the previous estimated system state, u is the control (representing the actuator force of actuator 22 and may be either the quarter car command or the
semi-rigid body command, whichever is applied), L is a linear Luenberger matrix that provides stabilizing linear feedback (the function L(y(t)-y.sup.e (t)) is referred to below as the linear correction term), y(t) is the measured relative system state of
the sprung and unsprung masses 12 and 24, N is a proportionality constant, and the term, .PHI.(y(t)-y.sup.e (t)) (referred to below as .PHI.(.)), is a saturation function and provides a stable nonlinear element to the model that guarantees that state
estimations progress in the direction of a stable sliding surface, y(t)-y.sup.e (t) =0, on an X, X' stability plot.
For purposes of simplification of the model, the characteristics of the rubber bushing 18 (FIG. 1) are not modeled, but accounted for by the non-linear term as error. The estimated relative system state, y.sup.e (t), is related to the state,
X.sup.e, as follows:
where | | |
