HK1088650B - Active suspending - Google Patents

Description

Active suspension

Technical Field

The invention relates to an active suspension.

Background

A vehicle moving in a desired direction will inevitably be affected by movements in other directions. Such unwanted movement often results from disturbances in the medium in which the vehicle is traveling. For example, whether riding in a car, boat or airplane, a person may encounter bumps, waves, bubbles, etc.

At best, this occasional acceleration can cause the passengers to feel uncomfortable and bored. For some susceptible people, these occasional accelerations can lead to dizziness and nausea. However, in some cases, particularly violent accelerations can cause the operator to briefly lose control of the vehicle.

Even when the vehicle is stationary, there is some residual vibration associated with the vehicle engine. During movement, the residual vibrations can cause oppression and fatigue even if the road is flat.

Disclosure of Invention

In general, in one aspect, the invention features a method of actively suspending a real plant in a vehicle, the method including modifying a control signal based on a difference between a property of the real plant and a property of a nominal plant, the property of the real plant being represented by a response of the real plant to the control signal.

Implementations of the invention may include one or more of the following features. The physical device includes a vehicle seat. The real plant is passively supported such that an active suspension coupled to the real plant experiences zero-mean loading. Actively suspending the real plant includes actively suspending the real plant using an electromagnetic actuator.

In general, in another aspect, the invention features a method of actively suspending an actual device in a vehicle, the method comprising: detecting a change in a characteristic of the real plant, the change indicating an abnormal condition; and to buffer movement of the real plant in response to the detected change.

Implementations of the invention may include one or more of the following features. Detecting the change includes detecting a fault condition of the sensor. Damping the movement of the real plant includes clamping motor leads of an electromagnetic motor coupled to the real plant.

In general, in another aspect, the invention features a control system for actively suspending a real plant in a vehicle, the control system comprising: a reference model of the device; a plant estimator for determining a difference between a property of the real plant and a property of a reference model of the real plant; and a device compensator that modifies the control signal that dampens the actual device vibration in response to the difference.

Implementations of the invention may include one or more of the following features. The weight sensor measures a weight associated with the device.

In general, in another aspect, the invention features an active suspension system for a device in a vehicle, the system comprising: an electromagnetic actuator; an amplifier having a control input to receive a control signal, an output in electrical communication with the actuator, and an energy input to receive a power source modulated by the control signal; an active power source in electrical communication with the energy input of the amplifier, the active power source having a saturated operating state to provide constant energy to the amplifier and an unsaturated operating state to provide variable energy to the amplifier; and a passive power source in electrical communication with the energy input of the amplifier and with the active power source, the passive power source providing energy to the amplifier when the active power source is operating in the saturated state, the passive power source storing energy from the active power source when the active power source is operating in the unsaturated state.

Implementations of the invention may include one or more of the following features. The passive power supply includes a capacitor. The capacitor is arranged to disable the amplifier substantially within 55 milliseconds.

In general, in another aspect, the features of the invention include: an active suspension comprising an electromagnetic actuator coupled to a device in the vehicle; and a force bias eliminator coupled to the plant for causing the actuator to experience a zero-mean load.

Implementations of the invention may include one or more of the following features. The electromagnetic actuator includes a multi-phase actuator. The sensor provides information indicative of the state of the device. The MEM sensor provides information indicative of the state of the device. Redundant sensors provide information indicative of the status of the device. The force bias eliminator includes a pneumatic system. The force bias eliminator includes a hydraulic system. A power supply is connected to the electromagnetic actuator, the power supply including an energy storage element responsive to a peak energy demand. The power supply includes an active power supply that provides energy to the actuator; and a passive power source that supplements the energy provided by the active power source during periods of highest energy demand. The passive power source includes a capacitor that is recharged by the active power source between periods of highest energy demand. The energy storage element includes a capacitance. The force bias eliminator is configured to enable adjustment of a position of the plant.

There is also a passive suspension, the active suspension and the passive suspension being connected to the plant in parallel to each other. The apparatus also includes a passive suspension, and wherein the active suspension and the passive suspension are connected to the plant in series with each other. The active suspension is configured to dampen vibrations of the plant along one axis, and the passive suspension is configured to dampen vibrations of the plant along multiple axes. The active suspension is configured to dampen vibrations of the plant along multiple axes. The plurality of axes includes a vertical axis and an axis extending in the front-rear direction.

In general, in another aspect, the invention features an apparatus comprising: an active suspension comprising an electromagnetic actuator connected to a device in the vehicle; and a force bias eliminator coupled to the plant for causing the actuator to experience a zero-mean load; the active suspension is configured to have a plurality of modes of operation, including an active mode and a passive mode.

Implementations of the invention may include one or more of the following features. The device is arranged to automatically switch between the active mode and the passive mode in response to detection of an event. The device is arranged to automatically switch between the active mode and the passive mode in response to a selection by a user. The electromagnetic actuator has terminal leads that are shorted in the passive mode. The force bias eliminator includes a passive component coupled to the plant, the passive component having a variable spring characteristic. The passive component has a first spring characteristic in the active mode and a second spring characteristic in the passive mode. The electromagnetic actuator is arranged to adjust the position of the device in the active mode. The passive component is configured to adjust the position of the device in a passive mode. The first spring characteristic is selected such that when cooperating with the actuator to adjust the position of the device in an active mode, the passive component removes a force bias in a control signal controlling the actuator, and wherein the second spring characteristic is controlled such that the passive component adjusts the position of the device in a passive mode.

In general, in another aspect, the invention features an active suspension device, comprising: a seat; an electromagnetic actuator for applying a force to the seat; and a force bias eliminator coupled to the seat for causing the electromagnetic actuator to experience a zero-mean load.

In general, in another aspect, the invention features an apparatus for suspending a seat, the apparatus comprising: an active suspension system suspending the seat; and a fail-safe system for controlling suspension of the seat in response to detection of a triggering event.

Implementations of the invention may include one or more of the following features. The suspension system includes an electromagnetic actuator having terminal leads, and the fail-safe system shorts the terminal leads of the electromagnetic actuator in response to detecting a triggering event. The fail-safe system places the active suspension in a passive mode of operation in response to a triggering event. The fail-safe system is configured to control suspension of the seat in response to a detected sensor failure that, in operation, provides information indicative of a state of the seat. The fail-safe system is configured to control suspension of the seat in response to detecting a failure in the suspension system. The fail-safe system is configured to control suspension of the seat in response to detecting a user selection.

In general, in another aspect, the invention features an apparatus for controlling an actively-suspended plant in a vehicle, the apparatus comprising: an active suspension including an electromagnetic actuator; a vibration isolation mass for generating a control signal in response to a measured characteristic of the actively-suspended plant; a force bias eliminator coupled to the plant for causing the electromagnetic actuator to experience a zero-mean load.

Implementations of the invention may include one or more of the following features. A compensation system modifies the control signal in response to a difference between the measured characteristic of the actively-suspended plant and a nominal plant. The apparatus comprises a seat and a passenger and the compensation system is arranged to modify the control signal in response to a difference caused by a change in mass of the apparatus. The compensation system includes a feedback loop that conveys a feedback signal indicative of the acceleration of the device; the compensation system is arranged to maintain a constant bandwidth of the feedback loop.

In general, in another aspect, the invention features an apparatus comprising: an actively suspended seat containing an electromagnetic actuator; a component associated with the seat and matching movement of the actively-suspended seat with movement of the component.

Implementations of the invention may include one or more of the following features. The component is selected from the group consisting of a beverage holder, a navigation device, a display, a control unit, a data entry/retrieval device, a socket and a pedal. A support coupled to the device, the support being movable in response to a force applied by the active suspension. The support comprises a scissor support. The support comprises a four-bar linkage. The support comprises an electromagnetic actuator. The support is arranged such that movement in response to a force applied along the first axis is independent of movement in response to a force applied along the second axis. The support is arranged such that movement in response to a force applied along the first axis is dependent on movement in response to a force applied along the second axis. The vehicle is selected from the group consisting of a tractor, a truck trailer, an Recreational Vehicle (RV), an aircraft, a car, a weapons platform, a machine tool isolation table, an elevator, a boat, truck, wheelchair, ambulance, baby carriage, roller car, construction equipment, and farm machinery, and the apparatus includes a seat.

In general, in another aspect, the invention features an apparatus that includes a device in a vehicle, the device being movable in a plurality of directions; and an active suspension system for supporting the plant, the active suspension system including a plurality of electromagnetic actuators for applying a force to dampen vibration of the plant in each of the plurality of directions; a control system for controlling the plurality of actuators in response to information indicative of a state of the device; the control system includes a plurality of controllers, each controller having a bandwidth designed to accommodate vibration characteristics in each of the plurality of directions.

Implementations of the invention may include one or more of the following features. The plurality of directions includes a vertical direction and a fore-aft direction, wherein a bandwidth associated with the vertical direction is less than a bandwidth associated with the fore-aft direction.

In general, in another aspect, the invention features an apparatus, an active suspension system coupled to a seat; a control system for controlling the active suspension system; and a cushion disposed on the seat, the cushion having a resonance selected according to a characteristic of the control system.

Implementations of the invention may include one or more of the following features. The characteristic includes a bandwidth associated with the control system, wherein the seat cushion is selected to have a resonant frequency outside of the bandwidth. The characteristic includes a bandwidth associated with the control system, wherein the seat cushion is selected to have a resonant frequency above the bandwidth.

In general, in another aspect, the invention features a method of suspending a seat using an active suspension system including an electromagnetic actuator, the method including selectively switching operation of the active suspension system between an active mode and a passive mode.

Implementations of the invention may include one or more of the following features. Selectively transitioning includes automatically transitioning in response to detection of an event. The event includes a failure of the active suspension system. The event includes a user selection.

In general, in another aspect, the invention features a method of actively suspending a plant in a vehicle, the method including applying a force to the plant using an active suspension system, the active suspension system including an electromagnetic actuator; and activates the fail-safe system to suspend the plant in response to the triggering event.

Implementations of the invention may include one or more of the following features. The triggering event also includes a fault in the sensor that, during operation, provides status information indicative of the device. The triggering event also includes a fault in the active suspension system.

Other features and advantages of the invention will be apparent from the following detailed description, and from the claims.

FIGS. 1 and 4-7 illustrate actively-suspended plants;

FIGS. 2 and 3 illustrate parallel and series connections to an actively-suspended plant;

FIG. 8 illustrates a control system that controls the actively-suspended plant shown in FIG. 1 and FIGS. 4-7;

FIG. 9 illustrates an embodiment of the control system shown in FIG. 8;

FIG. 10 illustrates a vibration isolation module;

11-12 illustrate a control system having different types of plant estimators;

FIG. 13 is a flow chart of an algorithm used by the force bias eliminator;

FIG. 14 illustrates an exemplary force bias eliminator;

FIG. 15 is a block diagram illustrating the processing performed by the force bias eliminator;

FIG. 16 shows a control system for measuring the weight of the device;

FIG. 17 illustrates the algorithm of the fail-safe system;

FIG. 18 illustrates the energy requirements in the general case;

FIG. 19 shows a power supply;

FIG. 20 shows a support arrangement for a seat that enables coupling between movement along different axes;

figure 21 shows a control system modified to accommodate coupling by movement along different axes.

Detailed Description

Actively-suspended plants include a seat, or other platform, coupled to one or more active suspension elements, each providing active suspension along an axis. In many cases, it is useful, but not necessary, to use passive elements that cooperate with active suspension elements along one or more axes. In this case, the active suspension elements may be mounted in series or parallel with the passive suspension elements.

In the following description, numerous references will be made to the positions and movements of the devices. It is to be understood that for ease of discussion in the disclosed embodiments, "position" refers to the position of the device relative to the vehicle and "movement" refers to movement of the device relative to an inertial reference. Thus, references to a position signal refer to a signal having information about the position of the device relative to the vehicle. References to movement signals refer to signals that represent information about the movement of the device relative to an inertial reference, such as acceleration.

The embodiments in the following description represent movements of the apparatus along any one or more of three coordinate axes of a cartesian coordinate system. However, the control system does not require any particular coordinate system for its operation. For example, the device may be arranged to move along any one or more of two axes. Furthermore, the device may be arranged to move along one or more non-orthogonal and orthogonal axes.

The control system described below is also not limited to control of device movement. The control system may also be used to control rotational movements such as tilting (pitch), turning or deflecting. Alternatively, the control system may be used to control any combination of rotational and translational movement.

FIG. 1 shows an actively-suspended plant 10 having a vertical active suspension element 12, such as a compression, that affects movement of the plant 16 along a vertical axis z, and a longitudinal active suspension element 14 that affects movement of the plant 16 along a longitudinal axis y. The movement of the device 16 along the transverse axis x is influenced by a passive suspension element 18, such as an elastic or an elastic/damper-based system. Where the plant 16 includes a seat, as shown in FIG. 1, it is particularly useful to use the active suspension elements 14 to inhibit longitudinal movement without using a purely passive suspension plant, since the active suspension elements allow the seat to remain stationary while forces are applied to the external environment. This feature is useful, for example, in preventing the seat from moving rearward when the driver depresses the foot pedal.

As used in this specification, an active suspension is a suspension that uses an actuator as an integral part thereof. Such actuators are capable of generating forces whose magnitude and direction can be controlled independently of the position and movement of the suspension. In some embodiments, the actuator is an electromagnetic actuator, or an electromagnetic motor, linear or rotary, single-phase or multi-phase.

The term "device" is intended to include a system that receives control signals and whose position and movement are to be controlled. The apparatus may include a seat, an occupant, any fixtures associated with the seat, a support structure for the seat, power electronics, and mathematical models of active and/or passive suspension elements, insofar as the elements affect the dynamic properties of the system to be controlled.

Actively-suspended plants may be used in a variety of applications. For example, the actively-suspended plant may be a stationary part of an engine, a platform on a ship, a seat, a bed, or in the cab of any moving vehicle, such as a car, truck, cruise ship or other watercraft, a train, a bus, an amusement vehicle, an ambulance, a tractor, a truck trailer, an agricultural vehicle, a construction vehicle, a weapons platform, an airplane, a helicopter, or other aircraft, a personal transportation device such as a wheelchair or a crib. Other examples of actively-suspended devices include lathe isolation tables, interferometer tables, photolithography tables, and the like.

The device need not include a seat. The device may also be a bed for sleeping in, for example, a truck cab or in a sleeping compartment of a train. Furthermore, the device is not necessarily a people carrier. For example, fragile items (e.g., ceramics and crystals) or explosive items (e.g., explosives) that require very careful transportation. Actively suspended plants may provide a suitable way to transport such cargo.

Furthermore, the device can be used in many places. For example, on luxury boats, the device may be used in a barber's room or a seasickness recovery rest room, and thus remain stationary even when the boat is tilted and turned. The device can prevent people from being in distress when the people are stormy at sea.

Each suspension element, whether an active suspension element or a passive suspension element, may inhibit movement along at least one axis. In some embodiments, active suspension elements are provided on all axes, in which case passive suspension elements need not be used. However, other embodiments include the following: one, preferably vertical, shaft, is provided with an active suspension; and passive suspension elements are arranged on the axes except the vertical axis. Alternatively, one axle, preferably the transverse axle, may be provided with passive suspension and the other axle may be provided with active suspension. In other embodiments, rotational movement such as tilting, turning, and deflecting may be controlled. In this case, the active suspension may be configured to facilitate tilting, turning, deflecting, or a combination thereof, of the device.

An axle without an active suspension need not be provided with a passive suspension. However, if passive suspensions are not provided on the other axes, the passengers will feel uncomfortable. For this reason, it is desirable to place active and passive suspensions on one or more axles. In this case, the active and passive suspensions may be in series with the active suspension or below the passive suspension, as shown in FIG. 2, or above the passive suspension (not shown). Alternatively, the active and passive suspensions may be arranged in parallel, as shown in FIG. 3.

In some embodiments, the active suspension may be turned on and off on an indefinite or real-time basis, may act on all axes at once, or may be done on an axis-by-axis basis. In some cases, if the active suspension element associated with a certain axis is not functioning, it may be useful to provide a fail-safe system associated with the active suspension element to dampen the movement of the plant along that axis.

Alternatively, if the active and passive suspensions are in series, the power to the active suspension may be cut off. In this case, any moving parts associated with the active suspension are clamped. Once the moving parts of the active suspension are clamped, only the passive suspension will affect the movement of the plant. In other cases, the moving parts of the active suspension may remain free to move in response to the action of the passive suspension elements.

The vertical active suspension element 12 comprises: a vertical accelerometer 20; a vertical position sensor 24; and a vertical actuator 28, which in the illustrated embodiment is parallel to the vertical position sensor 24 and preferably as close as possible to the vertical accelerometer 20. Similarly, the longitudinal active suspension element 14 includes: a longitudinal accelerometer 22; a longitudinal position sensor 26; and a longitudinal actuator 30 parallel to the longitudinal position sensor 26 and preferably as close as possible to the longitudinal accelerometer 22.

Actuators 28, 30 that may be used for the active suspension elements 12, 14 include single-phase or multi-phase electromagnetic actuators, such as three-phase linear actuators, single-phase linear actuators, rotary actuators, and variable reluctance actuators. One suitable actuator is an electromagnetic linear actuator such as that disclosed in U.S. Pat. No.4,981,309, the contents of which are incorporated herein by reference.

Any position sensor 24, 26 with sufficient resolution and accuracy may be used. Examples of suitable position sensors 24, 26 include sensors with potentiometers, sensors that utilize the hall effect, and sensors that use magnetostriction. Examples of position sensors with potentiometers include sensors from Novotechnik Inc of Ostfildern, germany. Other position sensors 24, 26 that may be used with active suspension elements include sensors having encoders and limit switches to determine absolute position. Various position sensors of this type, when fixed to a reference, can be used to obtain acceleration relative to the reference. Examples of suitable accelerometers 20 include MEM (micro-electro-mechanical) based accelerometers, such as those in the KXM60 series manufactured by Kionix, inc.

To help dampen vertical movement, the actively-suspended plant 10 includes elements that remove the biasing force in the actuator command force signal, thereby subjecting the actuator to a zero-mean load. In some embodiments, the element has the dynamic characteristics of a variable low stiffness spring. The low stiffness spring characteristic ensures that the actuator does not "fight" the spring when active isolation is performed. This can reduce energy loss. Such an element, referred to as a "force bias eliminator system," can be implemented as a cylinder with an associated accumulator, as shown in FIG. 14. The force bias eliminator system provides a biasing force, thereby causing the actuator to provide no force. Such an offset may be derived from factors such as the weight of the device 16. Because the force bias eliminator system provides a biasing force, the vertical actuator 28 need only suppress excursions from a predetermined equilibrium position. In a preferred embodiment, the cylinder and associated reservoir are arranged to maintain the actuator at a zero mean load. As described below, the force bias eliminator system can also provide passive suspension with or without additional damping.

As shown in FIG. 4, another embodiment of an actively-suspended plant 16 provides a separate active suspension element 34 for inhibiting movement of the plant 16 along a vertical axis. In this case, the active suspension elements 34 are mounted in series with the multi-axis passive suspension elements 36. As shown in FIG. 4, a multi-axis passive suspension element 36 is mounted between the active suspension element 34 and the plant 16. However, the passive suspension elements 36 may also be mounted between the device 16 and the vehicle floor 37, as shown in FIG. 5.

As shown in fig. 6, the device 16 may include various features or structures in addition to the seat 40 and its occupant. These additional features or structures are of the type that greatly benefit by remaining stationary relative to the device 16. Exemplary structures include a cup holder 42, which often holds a beverage that is susceptible to spillage under the random acceleration of the vehicle, a writing surface, a data entry/retrieval device, an ashtray or other receptacle 44, a display such as a navigation display, and a controller 47, particularly one that does not require a direct mechanical connection to the vehicle. Exemplary controls include electronic controls to operate heavy equipment, such as pedals or levers, for braking and acceleration. Although features or structures are shown attached to the device 16, it should be noted that these features or structures may be located remotely from the device 16 (not shown), but are "controlled" by the movement of the device 16.

The actively-suspended plant 16 shown in fig. 1 and 4-7 includes a base 46 configured to be bolted into standard bolt holes of various motor vehicles and models thereof. However, the actively-suspended plant 16 may be supported by any of a variety of support structures, including the scissor mechanism 51, the modified scissor mechanism, the four-bar linkage, and the modified four-bar linkage shown in FIG. 7 to be used when the ratio of actuator stroke to stroke is less than 1. Furthermore, the actuator itself may be designed as part of the support structure.

The active suspension elements 12, 14 exchange data with a control system 48, as shown in detail in FIG. 8. The control system 48 receives data signals from the sensors 24, 26, such as the device acceleration a_nAnd the position p of the device 16 relative to the vehicle_r. The control system 48 then provides a control signal u via a controller 49_rFor causing the respective actuators 28, 30 (see fig. 1) to exert a force to return the device 16 to the equilibrium position; minimizing the acceleration experienced by the device 16. The data signals represent the position and acceleration of the device 16, as well as data indicative of the properties of the device 16. Force bias eliminator module 60 exchanging data with plant 16 from actuator force control signal u_rThereby maintaining a zero mean load on the actuator.

In practical situations, the corresponding characteristics of the real plant 16 may not be accurately known. Thus, in general, any controller 49 design and control signal (u) output by the controller 49_r) Will be based on assumptions about these properties. Therefore, the control signal for controlling the real plant 16(real plant) may not achieve the intended effect. Thus, the control system 48 estimates errors in assumptions about the real plant 16 and compensates for these errors. In these examples, as followsThe estimation and assumption use the same reference model.

FIG. 9 illustrates an exemplary control system 48 in which the reference model includes a mathematical reference model 50 of the nominal plant in the form of the response Pn(s) of the nominal plant to the complex frequency input s. For the sake of simplicity of presentation, this mathematical model 50 of a nominal plant will simply be referred to as "nominal plant 50". The data indicative of the nominal plant response and the actual plant position and movement are used by the vibration isolation model 52 to calculate a nominal control signal u_n。

The nominal plant 50 is thus a reference model for the real plant 16. Such a model may be defined to include one or more parameters, including: desired performance characteristics, frequency response, poles/zeros, or a combination of these parameters. For example, where the real plant 16 includes a vehicle seat, the parameter indicative of the nominal driver weight may be defined as the average weight of a large number of typical drivers.

The vibration isolation module 52 may be implemented using various methods. For example, FIG. 10 shows a vibration isolation module 52 that includes position and acceleration feedback signals. In some implementations, a position controller having a position feedback loop sends a relative position signal p_rAs an input and to cause the real plant 16 to maintain the predetermined equilibrium position r. In some cases, the equilibrium position may correspond to a midpoint of the actuator stroke. In other embodiments, an acceleration controller with an acceleration feedback loop sends an acceleration signal a_rAs an input and to control the acceleration experienced by the real plant 16.

For ease of discussion, the following description is based on a dual-loop control architecture. Generally, however, the controller uses the position and acceleration of the real plant 16 as inputs and provides a control signal as an output. The implementation method does not need to be a dual-loop controller structure. For example, the controller may have a single loop. Other embodiments include a controller having 2n inputs and n outputs, where n is the number of axes actively controlled.

The bandwidth of the position loop may be designed based on the discomfort experienced by the passenger. The design performed varies with the particular axis along which the movement of the real plant 16 is to be controlled. For example, most passengers may experience greater vibration in the vertical direction than in the fore-aft direction. Further, in general, a spectrum vibrating in the front-rear direction has a higher frequency component than a spectrum vibrating in the vertical direction. Thus, in some embodiments, the position loop for suppressing vibrations in the vertical direction has a smaller bandwidth than the position loop for suppressing vibrations in the horizontal direction. When more than one axis is actively controlled, the position loop of each active shaft may have a bandwidth adapted to the vibrational characteristics along that shaft.

In the embodiment shown in fig. 10, the measured position signal p_rRepresenting the relative displacement between the real plant 16 and the vehicle frame, which is subtracted from the ideal equilibrium position r of the plant 16. The resulting difference is input to the vibration isolation module 52. Measured acceleration signal a_rSubtracted from the ideal acceleration, which in the embodiment shown is zero acceleration. The resulting difference is input to the vibration isolation module 52. The output of the vibration isolation module 52 is the nominal control signal u_n. Suitable vibration isolation modules 52 include those described in U.S. Pat. Nos. 3,701,499 and 6,460,803, the contents of which are incorporated herein by reference.

In some cases, it may be desirable to change the ideal equilibrium position r of the real plant 16 while actively controlling the plant 16. For example, when the real plant 16 includes a seat, it may be desirable to adjust the height of the seat to accommodate different passengers. This can be achieved by changing the ideal equilibrium position r. For the control system 48 shown in fig. 9, a change in the ideal equilibrium position r will cause a control signal u_rThe offset force component of (a). The bias force component is removed by the force bias eliminator module 60.

As described above, the vibration isolation module 52 generates a nominal control signal that is subjected to the measured position and acceleration at a nominal plant 50The signal may be used to control the movement of the nominal plant 50 at a particular disturbance. To generate the nominal control signals for the vibration isolation module 52, the signal a representing these disturbances will be_r、p_rInput into the vibration isolation module 52. However, in general, the nominal plant 50 has different dynamic characteristics than the real plant 16. Thus, the output of the vibration isolation module 52 is generally not optimized to control the movement of the real plant 16 subject to those same disturbances.

However, in most cases, the nominal plant 50 and the real plant 16 have sufficiently similar dynamics such that the control signal controlling the real plant 16, referred to herein as the "real control signal", is similar to the nominal control signal.

It is important to note that there is no actual nominal plant 50 that is undergoing any actual physical movement. Only a model of the nominal plant exists. The model is selected to respond in the same manner as the real plant 16.

In effect, the control system 48 uses the nominal plant 50 to simulate the response of the real plant 16 to the control signal. The control system 48 is free to use the assumed response of the nominal plant 50 to the control signal and the actual measured response of the actual plant 16 to the control signal. The control system 48 adjusts the control signal based on the difference between the assumed response and the actual measured response.

To compensate for the difference between the real plant 16 and the nominal plant 50, the control system 48 uses the plant estimator 62 to estimate the difference based at least in part on the signal indicative of the movement of the real plant 16. The plant estimator 62 then provides an error signal e(s) indicative of the difference to the plant compensator 64. The plant compensator 64 then operates by applying a control signal u at nominal_nWhich is modified to compensate for the difference before being input to the real plant 16. The combined use of the plant estimator 62 and the plant compensator 64 is referred to as a "compensation system 65". Although the plant estimator 62 and compensator 64 are separate from each other, they are separated only to illustrate their respective functions. In thatIn practice, the functions of the plant estimator 62 and compensator 64 may be implemented by circuitry embedded in separate hardware components or software.

In practice, the plant compensator 64 uses the error signal to perturb the nominal control signal u_n. The interference may result in an actual control signal u_rThe control signal is input to the real plant 16. As shown in fig. 11, the plant compensator 64 includes an amplifier. However, the plant compensator 64 may also include a filter. Note that as used herein, "actual" means that the control signal is to be input into the actual device 16. And not its usual mathematical meaning, i.e. the signal has no imaginary component.

In FIG. 9, the plant estimator 62 is shown receiving various inputs including: actual disturbance signals, e.g. position and acceleration signals a, representing the disturbances experienced by the real plant 16_r、p_rAnd represents the nominal control signal u_nA nominal disturbance signal 51 of the corresponding disturbance experienced by a nominal plant 50 under control. Alternatively, other information obtained externally from control system 48 may be used to estimate the difference, as described below. These inputs represent possible sources of information that the plant estimator 62 can use to generate the error signal. Embodiments of the plant estimator 62 need not actually receive or use all of the information sources shown in FIG. 9, but may instead receive or use a subset of these information sources.

The details of designing the compensation system 65 depend on the control objectives to be achieved. In one embodiment, as shown in FIG. 11, the control objective of the compensation system 65 is to maintain a constant bandwidth of the open loop acceleration loop transfer function, which is determined by the frequency at which the amplitude of the open loop acceleration loop transfer function intersects the 0db line, independent of any device differences caused by the occupant's weight. The compensation system 65 achieves this by adaptively adjusting the gain of the acceleration loop transfer function in response to the difference between the actual acceleration experienced by the real plant 16 and the nominal acceleration that would be experienced by the nominal plant 50 under the same conditions.

In this case, a suitable compensation system 65 is a model reference adaptive controller. In this case, the plant estimator 62 estimates the actual acceleration signal a of the actual plant 16 based on_rAnd a nominal acceleration signal u representing the acceleration to which a nominal plant 50 is to be subjected_nGenerating an error signal e(s). In this case, the plant compensator 64 is an amplifier that will nominally control the signal u_nMultiplied by the error signal e to obtain the actual control signal u_r。

As shown in FIG. 12, the compensation system 65 implemented using the model reference adaptive controller includes filtering the nominal acceleration a that the nominal plant 50 would experience under the same conditions_nAnd filters the actual acceleration a experienced by the real plant 16_rAnd a second filter 68. When using the nominal control signal u_nAs an input, a nominal acceleration a can be obtained_nAs the output of the nominal plant 50.

The output of the first filter 66 is a filtered nominal acceleration a_jnThe output of the second filter 68 is the filtered actual acceleration a_fr. The first and second filters 66, 68 primarily process the desired crossover frequency (i.e., the frequency at 0 db).

The filtered nominal acceleration is subtracted from the filtered actual acceleration in subtractor 70 to produce an error signal. The error signal may be further processed in a variety of ways to minimize certain quantization indicators of the error e.

In the embodiment shown in fig. 12, the quantization index of the error to be minimized is the Least Mean Square (LMS) of the error e. This is done by multiplying the error signal by the filtered nominal acceleration a in the amplifier 72_fnAnd then the operation is completed. The result of this process, i.e., the derivative of the compensation signal, is then input to the integrator 74. The output of the integrator 74 is then multiplied by the nominal control signal to produce the actual control signal.

An optional feature of the compensation system 65 is that the integrator 74 provides a unity output under special circumstances. In these special cases, the nominal control signal will remain unchanged when the output of the integrator 74 is multiplied by the nominal control signal. Thus, the actual control signal will be the same as the nominal control signal. An exemplary special case includes the detection of a nominal control signal that will cause the actuator to apply a very small force, on the order of friction. Other special cases include detecting acceleration below a threshold, or any other special case combination.

Fig. 12 illustrates a particular embodiment of a compensation system 65 that may compensate for only one of many factors, in this case, a change in the dynamic properties of the real plant 16. For example, such changes result from changes in passenger weight that result in differences between the dynamic properties of the real plant 16 and the nominal plant 50. However, the compensation system 65 may be designated to compensate for other such factors. One such factor includes the deviation (drift) in power electronics parameters.

The control system 48 is an analog system using continuous time signals. However, the control system 48 may also be implemented using discrete time, in which case the integrator 74 becomes an addition block and the low pass filters 66, 68 become appropriately defined digital filters.

As described above, the vertical active suspension element 12 that inhibits movement in the vertical direction includes the vertical actuator 28 that applies the necessary force to maintain the plant 16 in an equilibrium vertical position. However, in the vertical direction, the device 16 is constantly subjected to gravity. Thus, the actuator 28 consumes a significant amount of energy to support the weight of the device 16.

In one implementation, the force bias eliminator system is used to apply a biasing force in the vertical direction that is sufficient to counteract the actual control signal u_rThereby maintaining a zero mean load of the vertical actuator 28. Using the resulting force bias eliminator system, the vertical actuator 28 is ready for application of a loadThe device 16 maintains the force at the equilibrium position. While the vertical actuator 28 only needs to apply a force to compensate for the small offset from the equilibrium position.

The force bias eliminator system as described above is not strictly necessary. In principle, it is only necessary to have the vertical actuator 28 exert a suitable biasing force. Such a configuration would be useful if room temperature superconductors were suitable for carrying the current needed to generate such forces. However, with known electromagnetic actuators, the current required to support the device 16 would be very large and would generate significant waste heat.

The force bias eliminator system can be a relatively simple system such as an adjustable spring or a device with the mechanical properties of an adjustable spring of low stiffness.

The appropriate force bias eliminator system preferably operates whether or not the vehicle is traveling. This will enable the vehicle occupant to sit comfortably with all energy sources turned off. This feature is also important to ensure safety. If the positions of all seats on a vehicle that loses power suddenly drop in a vehicle traveling at high speed, the passenger may be uncomfortable.

To control this force bias eliminator system, the control system 48 also includes a force bias eliminator module 60 (see FIG. 11) that functions to cause the force bias eliminator system to provide an appropriate biasing force in a variety of changing environments. As shown in FIG. 9, the force bias eliminator module 60 receives acceleration and position data from the real plant 16 and an actual control signal u from the compensation system 65_r. Based on this data, the force bias eliminator module 60 provides a bias control signal to the force bias eliminator system, as will be discussed below in conjunction with FIG. 13.

The active suspension system is configured to operate in a plurality of modes, as follows: a safe (passive/fail-safe) mode, an active (force bias cancellation) mode, and a bump stop (bump stop) mode. As shown in FIG. 13, the system first detects whether a triggering event occurs via the force bias eliminator module 60 or a separate fail-safe system (see details below). The trigger event is generated in response to any change in the characteristics of the device 16, which may indicate an abnormal condition. Exemplary triggering events include a failure of the active suspension element, a severe cable failure, or a sensor failure (step 76). Upon detection of a triggering event, the force bias eliminator module 60 causes the force bias eliminator system 86 to operate in a mode that includes either a "passive mode", a "safe mode", or a "fail-safe mode" (step 78). In this mode, the position of the plant 16 is adjusted via the force bias eliminator module 60, as described below in connection with FIGS. 14 and 15. In some embodiments, the transition to "passive mode" operation may be implemented as a user selectable feature. Whether an active suspension element is operating or not can be readily determined by, for example, detecting energy input thereto. If the system determines that the active suspension element is operating, then the acceleration and position signals of the real plant 16 may be used to determine whether the vertical actuator 28 may reach the end of its travel, i.e., whether the vertical actuator 28 may encounter one of its two bump stops (step 80). If so, the force bias eliminator module 60 operates the force bias eliminator system in a "bump stop" mode (step 82). Otherwise, the force bias eliminator module 60 operates the force bias eliminator system in a normal mode or "active mode" (step 84), as described below in connection with FIGS. 14 and 15, in which the position of the plant 16 is adjusted by controlling one or more actuators.

The exemplary force bias eliminator system 86 is a pneumatic force bias eliminator (shown in FIG. 14) that includes a cylinder 88 supporting the plant 16 and a mating piston 90. The cylinder below the piston head, the "lower cylinder chamber," is connected to a source of compressed air (not shown) through a supply valve 92 or to ambient air through a bleed valve 94. Alternatively, the lower cylinder chamber may be connected to a compressed air source (not shown) or ambient air by operating a three-way manual adjustment valve 96. The compressed air source can easily be an on-board air source, such as a compressed air reservoir maintained at high pressure by a pump. The hollow portion of the seating structure may also serve as an air reservoir, thereby incorporating or integrating the air reservoir into the seating structure itself. Alternatively, the force bias eliminator system can be a hydraulic system.

The cylinder 88 may include a piston 90 that moves in response to air pressure and device weight. Alternatively, the cylinder 88 may simply expand and contract in response to pressure and weight, with the rubber tire also expanding and contracting in much the same manner. The expansion chamber 98 in fluid communication with the cylinder 88 may be an external air reservoir. Alternatively, the expansion chamber 98 may be built into the seat structure itself, thereby conserving space within the vehicle interior.

In the normal mode or "active mode", the force bias eliminator module 60 responds to, for example, a control signal u_rIt is determined whether the pressure needs to be increased or decreased as shown in fig. 11. If the pressure needs to be increased, the force bias eliminator module 60 causes the supply valve 92 to open and the bleed valve 94 to close, thereby filling the lower cylinder chamber with compressed air. Conversely, if the pressure needs to be reduced, the controller causes the supply valve 92 to close and the bleed valve 94 to open. This allows the release of high pressure air in the lower cylinder chamber.

FIG. 15 shows the use of the actual control signal u_rAs a process for detecting and removing the bias component of the actual control signal by the force bias eliminator module 60 operating in the active mode. Actual control signal u_rThe high frequency offset, which is likely the result of eliminating random accelerations, is first removed by a low pass filter 100. Thus, the low pass filter 100 isolates low frequency variations that are likely the result of actual weight variations in the device 16. A suitable low pass filter 100 has a corner frequency of about 0.5 hz.

The force bias eliminator module 60 then uses the sign or phase angle of the low frequency component of the actual control signal to determine whether to apply a biasing force to counteract u_rOf the offset signal component. For the implementation in fig. 15, the actual control signal u is used_rForce bias eliminator module as input60 determines whether the pressure against the piston 90 needs to be increased or decreased. The force bias eliminator module 60 then sends the appropriate valve actuation signal V to a supply valve relay (not shown) that controls the supply valve 92 and to a bleed valve relay (not shown) that controls the bleed valve 94. If the pressure needs to be increased, the force bias eliminator module 60 sends a signal to the supply valve 92 relay to open the supply valve and a signal to the bleed valve relay to close the bleed valve 94. Conversely, if the pressure needs to be decreased, the force bias eliminator module 60 sends a valve actuation signal V to the bleed valve 94 relay to open the bleed valve and a signal to the supply valve relay to close the supply valve 92. In some embodiments, a repeater with "backlash" may prevent vibration of the open-close valve around the repeater set point.

The force bias eliminator system 86 also includes upper and lower bump stop valves 102, 104 that are used in a "bump stop" mode to prevent movement in the event that the vertical actuator 28 cannot prevent the plant 16 from suddenly reaching the end of its stroke.

The upper bump stop valve 102 provides a path between the cylinder above the piston head ("upper cylinder") and the ambient air. In normal operation, the upper bump stop valve 102 is open, allowing air to freely enter and exit the upper cylinder chamber. However, if the force bias eliminator module 60 detects that the vertical actuator 28 is unlikely to stop the plant 16 from reaching the top of its stroke, it closes the upper bump stop valve 102. This prevents air from escaping from the upper cylinder while the piston 90 is moving upward. Thus, the air is compressed as the piston 90 rises, thereby exerting a force that prevents the piston 90 from moving further upward (and thus acting on the device 16).

The lower bump stop valve 104 provides a path between the lower cylinder chamber and the ambient air or compressed air source depending on whether the bleed valve 94 and the supply valve 92 are open or closed. In normal operation, the lower bump stop valve 104 is open. This allows the force bias eliminator module 60 to freely control the plant height by selectively opening and closing the bleed and supply valves 92. However, if the force bias eliminator module 60 detects that the vertical actuator 28 is unlikely to stop the plant 16 from reaching the bottom of its travel, it closes the lower bump stop valve 104. This prevents air from escaping from the lower chamber as the piston 90 moves downwardly, thereby exerting a force that prevents further downward movement of the piston 90 (and hence the device 16).

When the force bias eliminator module 60 determines that the active suspension element has been deactivated, it sends a valve actuation signal V to seal the upper and lower chambers by closing the upper and lower bump stop valves 102, 104 simultaneously. This allows the force bias eliminator system 86 to operate in a "safe mode" in which the force bias eliminator system 86 functions as a spring. When operating in the safe mode, the only passage of air into and out of the cylinder 88 is through the three-way manual adjustment valve 96. The three-way manual adjustment valve 96 has: a closed position in which no air can enter or exit the lower chamber; a deflation position, at which the lower chamber is connected to ambient air; and a fill position in which the lower chamber is connected to a compressed gas source (not shown).

In the safe mode, the height of the device 16 is controlled by operating the adjustment valve 96. When the device 16 is raised, the adjustment valve 96 connects the compressed air source to the lower chamber. When the device 16 is lowered, the adjustment valve 96 connects the ambient air to the lower cylinder chamber. When the device 16 is in the desired position, the adjustment valve 96 seals the lower cylinder chamber.

The following table summarizes the settings of the various valves shown in FIG. 14 for various operating modes:

	security mode	Force bias cancellation mode	Bump stop	Lower bump stop
					Lower bump stop valve 104	Close off	Is opened	Is opened	Close off
Bleed valve 94	Is opened	Is turned on to fall	Close off	Close off

	Or shut down	Closed to rise
					Supply valve 92	Close off	Open to rise and close to fall	Close off	Close off
Upper bump stop valve 102	Close off	Is opened	Close off	Is opened
					Three-way valve 96	Three-way	Close off	Close off	Is opened

As described above, in the force bias elimination mode, the compressed air source may be in communication with the lower chamber of the cylinder 88. If the vertical actuator 28 needs to lower the device 16 briefly, it may be found to be difficult because the compressed air source will prevent downward movement of the device 16. To make it easier for the vertical actuator 28 to lower the plant 16, the force bias eliminator system includes an expansion chamber 98 disposed between the supply valve 86 and a lower bump stop valve 104. The expansion chamber 98 acts as a weak spring so that if the vertical actuator 28 lowers the device 16, the device 16 experiences minimal resistance from the compressed air source.

When the device parameters change suddenly, for example when passengers sit down or stand up, or when cargo is unloaded or loaded, the maximum force that the actuator needs to exert can be reduced, thereby reducing the energy used. This is accomplished by having the control system wait for the force bias eliminator system to adapt to the new load. After the force bias eliminator system has been adapted, the actuator can apply a reduced force required for normal operation. For example, a force bias eliminator system implemented as an air spring will quickly unload pressure when a seated passenger stands up. If the system completes the pressure unloading, the actuator will apply the force required to adaptively actively suspend the unmanned seat. Thus, the seat height remains substantially constant when the passenger stands up. In contrast, seats supported by conventional springs tend to spring back to an unloaded position when the occupant stands up.

The force bias eliminator system 86 disclosed in connection with FIG. 14 is a pneumatic system that implements pneumatic logic to operate. However, other types of force bias eliminator systems, such as hydraulic systems, may be used to apply a biasing force to the plant 16.

The pneumatic force bias eliminator 86 described above can be used to implement an embodiment of the plant estimator 62 that directly measures the mass of the plant 16. In such an embodiment, as shown in FIG. 16, the bias pressure is measured by a pressure sensor 108, and the amplitude of the bias pressure, after appropriate filtering, represents the weight of the device 16. The bias pressure is the pressure that would support a nominal plant 50. This pressure may be measured at the factory or after seat installation and may be programmed into the plant estimator 62. The pressure sensor 108 measures the pressure at any point in the pneumatic system where there is a measurable pressure that depends on the weight of the device 16. The pressure measured by the pressure sensor 108 is then transmitted through a low pass filter 119 to remove any instabilities. The result is then provided to the plant estimator 62, which determines an error signal for interfering with the nominal control signal.

The above description has focused on the reference model based control design shown in fig. 9. However, other embodiments include a general control system as shown in fig. 8, wherein the controller may be designed via a suitable linear or non-linear control method, with or without the use of a reference model. For example, the vibration isolation module of FIG. 9 may replace the controller module 52 of FIG. 8 without providing a nominal plant 50, a compensation system 65, or a force bias eliminator module 60.

Faults in the active suspension elements, particularly the vertical active suspension element 12, may cause abrupt and possibly alarming changes in plant position and movement. To avoid this, the control system 48 may be provided with a fail-safe system controlled by the fail-monitor 112 shown in fig. 9. The fail-safe system is connected to the selectively energized buffer of the actuator. The buffer may be a separate element. Alternatively, the damper may be implemented by changing a changing characteristic of the active suspension element. Under normal conditions, the damper is not activated and therefore no damping force is generated. However, if fault detector 112 detects the presence of a particular condition, it activates the damper, thereby inducing a damping force that resists movement of device 16. This allows the device 16 to descend gracefully to the lower bump stop. Alternatively, the fail-safe system may include an elastic member or a structure having a function of an elastic member. This configuration is the force bias eliminator system described earlier. In this case, the device 16 will be lowered to an equilibrium position above the lower bump stop.

As shown in fig. 9, the fault detector 112 is provided with information indicative of the status of the real plant 16. This information may include, for example, position and acceleration signals. Based on this information, fault detector 112 determines whether it is necessary to buffer the movement of device 16. Suitable dampers for use in conjunction with the actuator 12 are described in U.S. Pat. No.4,981,309, the contents of which are incorporated herein by reference.

Failure of the active suspension system is not the only reason to activate the fail-safe system. Changes in the characteristics of the device 16 may indicate an abnormal condition, which may also be a cause of the triggering of a fail-safe system. For example, sensor signals above a predetermined threshold, or a fault in any sensor that collects information indicative of the status of the device, are the cause of the activation of the fail-safe system. Failure of a sensor or system can be detected by noting the presence or absence of such a signal or by the sensor signal providing the information not being in accordance with physical constraints on the device. For example, if a sensor indicates that a car is now moving at supersonic speeds, then the reliability of that sensor is likely to be a problem, in which case the fail-safe system will be activated. Alternatively, the fail-safe system may also be activated when it is detected that the sensor has reached the end of its usable range. In some embodiments, the particular firing event that causes the transition from the active mode to the fail-safe mode or "passive mode" may also be implemented as a user-selectable technical feature.

In the case of an electromagnetic actuator, a stator with coils surrounds an armature on which the real plant 16 is mounted. The stator and armature together form an "electromagnetic actuator" and the position of the armature can be controlled by the current in the stator coils. In normal operation, current through the coil generates a magnetic field that controls the position of the armature. By detecting a fault, the leads of the coil are shortened or pinched together. In these cases lenz's law will act to induce a current in the coil, thereby creating a magnetic field that prevents the armature from moving. Thus, the electromagnetic actuator functions as a buffer.

FIG. 17 is an exemplary algorithm used by the fault monitor 112 to determine whether to fire a buffer upon detection of a sensor fault. The fault detector 112 detects an acceleration signal (step 114). If the acceleration signal is within the preselected range (step 116), the fault detector 112 assumes that the active suspension element is functioning properly. In these cases, the buffer is still not excited. The fault detector 112 then waits (step 118) and again detects acceleration (step 114). However, if the fault detector 112 detects that the acceleration signal exceeds the threshold magnitude for longer than a predetermined duration (step 116), the fault detector 112 assumes that the active suspension has failed and prepares for remedial action. Under these circumstances, the fault detector 112 activates the buffer (step 120). The fault detector 112 then stops further execution (step 122) until reset.

The algorithm described in connection with fig. 7 depends only on the detection of a malfunction of the acceleration sensor. However, other algorithms may use position signals or a combination of position and acceleration signals, or information indicative of the state of the real plant 16. Other algorithms for controlling the fail-safe system use state information representing various components including the active suspension and associated components: such as the power supply, power amplifier, controller itself, etc. Exemplary information includes an electronic deviation signal input to the sensor, and an electrical force generated by the actuator. Moreover, redundant sensors may be used to improve the reliability of the system.

The actual device control signal is ultimately used to modulate the output current of the amplifier 106 as shown in fig. 18. This output current is ultimately provided to electromagnetic actuator 28. Fluctuations in the control signal may result in fluctuations in the output current.

The current modulated in amplifier 106 (as shown in fig. 19) is provided by power supply 107. In active suspension systems of the type described above, the amplifier 106 requires a relatively small amount of energy when traveling on a flat road during normal operation. However, to compensate for conditions such as high road bumps, the amplifier 106 requires a sudden high energy source. Fig. 18 shows the energy required as a function of time for a typical city street with occasional potholes and other irregularities.

As shown in fig. 18, in normal operation, the amplifier 106 uses average energy. Occasionally, however, the amplifier 106 requires significantly more energy for a short period of time. In order to provide a sudden high energy it is useful to provide the amplifier 106 with a power supply 107 having an energy storage element capable of suddenly providing a large amount of energy, not directly available from an energy source such as a battery.

As shown in fig. 19, a suitable power supply 107 includes a DC/DC converter 109 having an input connected to a battery 113 and an output connected to a capacitor 110. The function of the DC/DC converter 109 is to convert the battery voltage, nominally 12 volts, to a higher output voltage. The other role of the converter 109 is to limit the amount of current drawn from the battery 113 by saturation when the amplifier demand for current exceeds the current that the converter 109 can supply. In these cases, converter 109 acts as a constant current source for amplifier 106 that provides a saturation current.

Generally, in response to a change in a parameter of the power supplied by the converter circuit (such as output current, input power), additional power may be supplied by a passive power source, such as a capacitive element, to meet maximum energy requirements. A suitable power supply circuit is disclosed in U.S. patent application No.10/872040 filed on 18.6.2004, the contents of which are incorporated by reference.

In normal operation, the converter 109 meets the amplifier's own requirements for power. When the current required by the amplifier 106 exceeds the current that can be supplied by the converter 109, the amplifier draws an insufficient portion from a capacitor 110 connected in parallel with the output of the converter 109. The charge stored in capacitor 110 is sufficient to meet the requirements of amplifier 106 since a large amount of current is required only for a relatively short period of time. As shown in fig. 18, the converter 109 may recharge the capacitor 110 due to the need for a large amount of current being spaced for a relatively long period of time.

The particular values associated with the components shown in fig. 19 depend on the detailed scheme associated with each application. Such detailed schemes include average power consumption of the amplifier 106, maximum power consumption, average time between large power consumption demands, operating voltage range, saturation current and parasitic resistance of converters associated with various configurations of capacitors or other energy storage elements used to form the capacitor 110.

The capacitor 110 does not need to be provided with a single capacitor. In some cases, it is more economical to package the capacitors into a circuit having an appropriately equal amount of capacitance 110 and capable of operating over a range of voltage drops required. For example, the circuit may include a number of capacitors in series, each capacitor having a voltage rating less than the supply voltage applied to the amplifier 106. In one embodiment, 62 capacitors are connected in series to form the desired capacitance, each capacitor being 17.3 farads and operating in the 2.5 volt drop range.

If a malfunction or instability occurs, the force applied by the actuator tends to become very large and remain large for a long period of time. The power required to hold the large force for a long time will drain the capacitor 110 and "droop" the voltage of the amplifier 106. This may cause the amplifier 106 to stop operating itself. Thus, during periods of instability, the power supply 107 imposes a power supply limit that is low enough so that excess heat can be quickly dissipated, thereby avoiding thermal damage to the amplifier 106. In this manner, the illustrated power supply 107 may limit power losses to the actuator when a fault or instability condition occurs. In one embodiment, the capacitance is selected to deplete the stored energy to disable the amplifier within 55 milliseconds when a fault or instability condition occurs.

In those embodiments where the device includes a seat, difficulties can arise due to the presence of the seat cushion. As described above, the accelerometer and the position sensor are mounted on the seat. Thus, the accelerometer and the position sensor detect the movement of the seat. In this case, it is assumed that a passenger sitting on the seat will undergo the same movement as the seat. However, in most cases, the passenger does not sit directly on the seat. Instead, the passenger sits on the seat cushion. The seat cushion acts as an additional passive suspension element and has its own transfer function.

In practical cases, the seat cushion introduces a transfer function with two zeros. These zeros reduce the gain of the acceleration loop and thus reduce the ability of the control system to suppress vibrations at frequencies relative to these zeros.

To address this difficulty, the frequencies associated with the two zeros exceed the bandwidth of the acceleration loop controller 58. This can be achieved by increasing the effective stiffness of the elastic members in the cushion, for example by placing the cushion on a hard cushion.

Thus, the inputs and outputs of the various modules of control system 48 are presently considered to be scalars. This is because for many seat bases, the force required to dampen vibrations along one axis is largely independent of the force required to dampen vibrations along another axis. In this case, the device 16 is characterized by a substantially diagonal matrix transfer function, in which there are very few off-diagonal components. The acceleration loop transfer equations and the position loop transfer equations are also characterized as being substantially diagonal matrices. In these cases, it is appropriate to consider suppressing the vibration along one axis without considering suppressing the vibration along the other axis.

However, for certain types of seat bases, the forces used to dampen vibration along one axis may affect simultaneous attempts to dampen movement along another axis. When this is the case, the device transfer function is no longer treated as a diagonal matrix.

For example, fig. 20 shows a chair that includes a seat supported by a four bar linkage. Obviously, an upward movement of the seat also results in a rearward movement, the relationship between the two depending on the vertical position of the seat.

In fig. 21, the input and output of each block are two-dimensional vectors. Thus, the control signal provided to the real plant 16 includes components for controlling two different force actuators. The control signals enter the separation-like transformation matrix R before being input to the real plant 16. The detailed description of this process is described in section 3.3 "vector space" of the "control system manual" published by the IEEE press. (section 3.3 "vector spaces" of "Control System Handbook")

For the case shown in fig. 20, the coupling is purely dynamic and the similarity transformation is used to decouple the dynamic cross-coupling arrangement. In this case, the elements of the separation matrix are real-valued constants. The values of these elements can be obtained directly by taking into account the geometry of the support.

In other cases, the coupling between the two directional movements includes both dynamic and dynamic components. In this case, the elements of the separation matrix are complex-valued functions of frequency. Generally, such a matrix will not produce a low-order realizable transfer function matrix suitable for implementation by a controller.

Another way to suppress vibration of the device in both directions is to use a controller matrix that is completely filled with values instead of a diagonal matrix for the acceleration loop controller. In this case, the elements of the acceleration loop controller matrix are calculated such that the closed-loop matrix transfer function associated with the acceleration loop is a diagonal matrix or has negligible off-diagonal elements.

Other embodiments are within the scope of the following claims.

Claims (13)

1. An active suspension apparatus for a device in a vehicle, comprising:

a device in the vehicle, the device being movable in a plurality of directions and having a device position;

an active suspension including an electromagnetic actuator coupled to the plant; and

a force bias eliminator coupled to the plant for causing the actuator to experience a zero-mean load;

the active suspension is configured to have a plurality of modes of operation, including an active mode and a passive mode,

the electromagnetic actuator is unable to provide active force in the passive mode,

the force bias eliminator adjusts the plant position in the passive mode.

2. The active suspension apparatus for a device in a vehicle of claim 1, wherein the apparatus is configured to automatically transition between the active mode and the passive mode in response to detection of an event.

3. The active suspension apparatus for a device in a vehicle of claim 1, wherein the apparatus is configured to automatically transition between the active mode and the passive mode in response to a user selection.

4. The active suspension apparatus for a device in a vehicle of claim 1, wherein the electromagnetic actuator has terminal leads that are shorted in a passive mode.

5. The active suspension apparatus for a plant in a vehicle of claim 1 wherein the force bias eliminator comprises a passive component connected to the plant, the passive component having a variable spring characteristic.

6. The active suspension apparatus for a device in a vehicle of claim 5, wherein the passive component has a first spring characteristic in an active mode and a second spring characteristic in a passive mode.

7. The active suspension apparatus for a device in a vehicle of claim 1, wherein the electromagnetic actuator is configured to adjust the device position in an active mode.

8. The active suspension apparatus for a device in a vehicle of claim 5, wherein the passive component is configured to adjust the device position in a passive mode.

9. The active suspension apparatus for a device in a vehicle of claim 6, wherein the first spring characteristic is selected such that the passive component removes a force bias in a control signal that controls the actuator when cooperating with the actuator to adjust the device position in an active mode, and the second spring characteristic is controlled such that the passive component adjusts the device position in a passive mode.

10. A method of suspending a seat using an active suspension system including an electromagnetic actuator, the method comprising:

selectively transitioning operation of the active suspension system between an active mode and a passive mode and adjusting seat position in the passive mode with a force bias eliminator such that the actuator experiences a zero-mean load.

11. The method of claim 10, wherein selectively transitioning comprises automatically transitioning in response to detection of an event.

12. The method of claim 11, wherein the event comprises a failure of the active suspension system.

13. The method of claim 11, wherein the event comprises a user selection.