WO1998014033A1

WO1998014033A1 - Method for automatic control of a transducer with variable reluctance, and linear motor for implementing same

Info

Publication number: WO1998014033A1
Application number: PCT/FR1997/001668
Authority: WO
Inventors: François Luc; Hervé GOUTELARD; Michel Lagier
Original assignee: Thomson Marconi Sonar S.A.S.
Priority date: 1996-09-24
Filing date: 1997-09-23
Publication date: 1998-04-02
Also published as: AU4388297A; FR2753873A1; FR2753873B1

Abstract

The invention concerns methods and devices for automatic control of transducers with variable reluctance. It consists in determining (201) in non- linear manner independent of the mechanical load of the transducer the field required for the air gap of this transducer based on the mechanical force for obtaining and for controlling in linear manner (205) the magnetic field present in the air gap from a measurement (203) of this magnetic field and optionally a measurement (204) of the current flowing in the coil of the transducer. It enables the production of linear motors for stabilising vehicle seats, or acoustic transducers for sonars and more generally systems for controlling axial vibrations.

Description

METHOD FOR CONTROLLING A VARIABLE RELUCTANCE TRANSDUCER, AND LINEAR MOTOR FOR DRIVING

WORK OF SUCH A PROCESS.

The present invention relates to methods for controlling variable reluctance transducers. It also relates to linear motors in which this method is implemented.

Variable reluctance transducers are well known devices, the schematic diagram of which is shown in FIG. 1. Such a device comprises two magnetic cores 101 and 102, generally made of soft iron, often laminated, and separated by a gap 103. At least one of the cores has an electric coil 104. When an electric current, denoted I, travels the coil, it generates a magnetic field, denoted B, in the soft iron core 101. This field is given by the formula :

B _* - ^ L ₍₁₎

2G 'where μn denotes the magnetic permittivity of the air, N the number of turns of the coil and G the thickness of the air gap. An attractive force F then develops between the nuclei 101 and 102, given by the formula:

where S denotes the section of the nuclei. Knowing that the field of saturation of soft iron can reach 1.5 Tesia, the transducer thus formed is capable of delivering attractive forces greater than 180 Newtons per square centimeter of core section. A transducer is thus easily obtained capable of delivering very large forces of several thousand Newtons while using inexpensive materials (soft iron and copper coils, no magnet, etc.) The displacement dynamics of the transducer is given by the domain of validity of formula (1), which assumes that the thickness of the air gap is small compared to the square root of the section of the nuclei. In practice we will use the formula:

The main drawback of the transducer thus produced lies in the fact that it has a non-linear dynamic: in fact the force developed is proportional to the square of the electric current and inversely proportional to the square of the thickness of the air gap.

The old telephone headsets worked on this principle. To obtain a substantially linear response, a continuous bias current Irj was used superimposed on the active current Δl. In this case, we get:

_{F Δ |}

A variation in force is thus obtained proportional to the variation of the current provided that the fluctuation of this current is small compared to the bias current in order to be able to neglect the term in Δl ² , and that the relative variation of the thickness of the air gap be very weak.

In this case, the intrinsic dynamics of the transducer, both in terms of forces and in terms of displacement, are not exploited. This is why these transducers have been abandoned in favor of electrodynamic transducers using a magnet and a coil. However, two methods are known for controlling the signal delivered by a variable reluctance transducer:

- The first consists, from the analysis of the mechanical and electromagnetic equations of the transducer, to use an electrical signal modified by a non-linear signal processing operation in order to obtain the desired mechanical signal. This method, if it makes it possible to exploit the dynamics of the engine, requires a good knowledge of the dynamic behavior of the load applied to the engine.

- the second consists in slaving the system so that it presents a linear dynamic behavior. The equations involved being non-linear, this control must necessarily be non-linear, which requires the use of relatively complex electronics.

In the past, the price of electronic components for such operations was much higher than that of the engine. This technology was therefore of no interest in terms of cost. Today, the fall in the price of electronic components, in particular those used in digital electronics, is reviving the interest of variable reluctance motors.

In an article published in the proceedings of the days of application of active control to the reduction of noise and vibrations held on November 14 and 15, 1995 in Senlis, André GENNESSEAUX of the company PAULSTRA proposed an anti-vibration support for an automobile engine comprising a linear electric actuator with variable reluctance controlled using a digital electronic servo operating from measurements of the thickness of the air gap and of the current flowing in the active coils. However, this enslavement is essentially non-linear.

To overcome these drawbacks, the invention proposes a method for controlling a variable reluctance transducer comprising an air gap, mainly characterized in that the magnetic field necessary in the air gap is determined to obtain the desired force from the transducers, that the actual magnetic field existing in the air gap is measured, and that the voltage applied to the transducer is controlled to linearly slave the actual field to the necessary field.

According to another characteristic, the electric current in the transducer is measured to use this value in the control loop.

The invention also provides a linear motor implementing this method, mainly characterized in that it comprises a first U-shaped magnetic core, at least one coil disposed on one of the branches of the U forming this first core, a second I-shaped core arranged in front of the free ends of the two branches of the U to determine between the first and the second core a gap, two shims arranged on the free ends of the branches of the U to impose a minimum thickness of said gap , and a magnetic field probe placed on the free end of one of the branches of the U and whose thickness is more reliable than the thickness of said wedges.

According to another characteristic, the two cores are fixed respectively on two amagπetic plates and that they further comprise a set of springs fixed between these two plates to hold them separate; this device for controlling the movements of a seat of a vehicle.

According to another characteristic, the second core is fixed to a pavilion intended to generate underwater acoustic waves of low frequency and high intensity.

Other features and advantages of the invention will appear clearly in the following description given with reference to the appended figures which represent:

- Figure 1 a schematic view of a known variable reluctance transducer;

FIGS. 2 and 3, servo diagrams according to the invention, in digital and analog respectively, and

- Figure 4, a schematic view of a transducer according to the invention. The invention therefore proposes to control the magnetic field existing in the air gap by determining the electric voltage V to be applied to the edges of the coil of the transducer from the measurement of the magnetic field in this air gap and from the measurement of the intensity of the electric current flowing in this coil. We know that the voltage at the edge of such a transducer is given by the equation:

V = NS— + RI (5) dt

In this equation R denotes the series resistance presented by the coil circuit. The magnetic field can then be controlled either by a digital method or by an analog method.

Using the digital method illustrated diagrammatically in FIG. 2, the setpoint Br} _es of the desired magnetic field is determined in a circuit 201. The real value B _r of the magnetic field actually prevailing in the air gap is then measured using a magnetic field probe connected to a measurement circuit 203. The actual value I of the current flowing in the coil is also measured. using a small series resistor inserted in the circuit of this coil by measuring the voltage at the edge of this resistance by a circuit 204. The voltage V _{app is} then determined using the formula:

V _ap p ≈ NS ^Bdes ^ ^Brée RI _rée , (6)

For this, after subtracting the value of the measured field from the value of the desired field in a subtractor 205, the product of this difference is produced by NS / τ in a multiplier 206. Τ represents the time interval between two measurements of the current and of the real field and corresponds to the timing of the digital processing device.

The value of the measurement of the real current I is multiplied by the value of the measurement resistance R in a multiplier 207.

The signals from the multiplier 206 and the multiplier 207 are then added to an adder 208 which therefore delivers in digital form the value of the voltage to be applied to the transducer. This digital value is converted into analog in a digital to analog converter 209, then is applied to a power amplifier 210 which supplies the coil of the transducer 202.

The analog method is shown schematically in Figure 3.

The field in the air gap of the transducer 202 and the electric current in the winding of this transducer are measured as in the digital method using a field probe and a small series resistance. However, the measurement circuits 303 for the field and 304 for the intensity operate in an analog manner.

The setpoint value of the desired field Brjes is fixed by a circuit 301, again in an analog manner.

The value of _Unified Live B is subtracted from the value B _G J _Q S in an analog subtractor 306, and the result of this subtraction is multiplied by NS / τ in an analog multiplier 306. The term τ in this case will be explained later in this text.

The value of the current is multiplied in an analog multiplier 307 by the value of the measurement resistance R, and the result of this operation is added in an analog adder 308 to the result of the operation obtained in the multiplier 306. This gives the value of the voltage to be applied to the transducer 202.

However in this case, in particular to avoid any problem of stability of the servo-control, the signal from the adder 308 is filtered in a low-order first-order filter 309, the time constant of which is precisely the value τ seen more high. The cut-off frequency ω _c = 1 / τ is determined as a function of the time response needs of the transducer 202 and the conditions of stability of the system. It can go up to a few kHz. The voltage thus filtered is then applied to a power amplifier 210 which supplies the transducer 202.

The set value of the desired field B <j _is determined in circuits 201 and 301 of course depends on the desired value for the force F delivered by the transducer 202. This force being known as a function of the actions desired for the transducer 202, the set value of the desired field will be given by the formula:

It can be seen that the control loop which thus makes it possible to linearize the response of the transducer is linear and that on the other hand the operation making it possible to obtain the set value of the field is itself non-linear. However, this non-linear operation is completely independent of the mechanical load of the transducer, which is a very important advantage. The method according to the invention will be more particularly useful for producing a linear motor making it possible to actively control the suspension of the seat of a vehicle.

Such a system has been shown schematically in FIG. 4. The linear motor intended for active control of the vehicle seat is located between two plates 401 and 402 of non-magnetic materials, aluminum for example 401 and 402. The plate upper 401 supports the vehicle seat, not shown in the figure, and the plate 402 is fixed to the structure 403 of the vehicle. The two plates 401 and 402 are maintained spaced apart by springs 404, 4 in number for example. This makes it possible to obtain a repulsive force between these plates, since the force obtained using the transducer is necessarily attractive.

The linear motor consists of a U-shaped lower core 405 fixed by its transverse part to the upper face of the lower plate 402 The other part of the actuator consists of an I-shaped upper core fixed to the underside of the upper plate 401. A magnetic coil 407 is threaded on one of the two branches of the core 405 and allows to receive the electric motor control current. To measure the real magnetic field between the two nuclei, we use a Hall effect probe, which can be very small, placed at the free end of one of the branches of the U forming the 405 nucleus, and therefore located at the inside the air gap existing between the two cores. The geometry of the system could be arbitrary, for example in Ul, El, UU .... To protect this Hall effect probe and prevent it from crashing under the effect of a sudden rapprochement of the two cores, we placed on the free ends of the two arms of the U-shaped core 405 two wedges 409 and 410 made of non-magnetic, slightly compressible material, for example plastic. These wedges also allow a calibration of the system, as will be seen below, and they also prevent the core 406 from remaining stuck to the core 405, which could happen in case of contact, even if these cores are made in soft iron.

Such a device makes it possible to obtain displacements of the order of 4 mm peak to peak with an amplitude of the variations of the applied force of the order of 2000 newtons peak to peak. These performances make it possible to effectively control the vibrations transmitted to the seat of a vehicle traveling on a conventional surface and to obtain excellent driving comfort for this vehicle.

Although it is possible to use a purely analog device for generating the voltage to be applied across the terminals of the coil 407, it is preferable in the context of such an application to use a digital device which will naturally take its place in the many digital devices that vehicles now have. The material resources can consist of a central computer used to manage the various vehicle control functions and the application specific control of the seat vibration will be carried out using a special program executed in timeshare to perform the various calculations defined above.

This program will be loaded with the various parameters, in particular the section of the lower core, the number of turns of the coils, the value of the series resistance of the system "coil + current measurement resistance", the value of current measurement resistance, the gain of the power amplifier and the sensitivity of the Hall effect sensor 408.

Given the relatively large dispersion that can be observed on the sensitivity of the various available sensors, the invention proposes to determine the value of this sensitivity using a calibration procedure in which a fixed value is imposed on the air gap by removing the springs 404 and applying the core 406 to the shims 409 and 410, which thus determines a fixed value for the thickness of this air gap. The magnetic field can then be determined from the current measurement using formula (1), which thus allows the sensor to be calibrated.

The mathematical study of the enslavement thus produced shows that the stability of this enslavement as well as its robustness, that is to say its capacity to take in the steady state a value very close to the desired value, depend on the value of the parameter τ. In practice, for the values of τ commonly used and which depend on the case of the sampling speed in digital and the cutoff frequency of the filter in analog, this condition is always fulfilled.

This same mathematical study shows that the intervention of the term in RI, function of the current in the winding, is relatively weak. In order to simplify the construction of this device, the invention proposes, as a variant, not to use the measurement of the current to carry out the control. The calculation, confirmed by experimentation, shows that the results of the servo-control by not using the current measurement are extremely close to those obtained by using it. However, there is a slower convergence of the processing algorithm in the case of the digital solution, and correspondingly a longer response time in the case of analog solution. In the case of the digital solution, this slowdown in convergence can be perfectly compensated for by an increase in the sampling frequency and the speed of processing in the digital means for processing this information. This presents no difficulty with the techniques currently used in digital processing. This results in a decrease in the parameter τ which we saw above that it was favorable. The application described in the stabilization of a seat is an example among others ^* and the invention can be applied to any engine using a variable reluctance transducer.

Another particularly important application relates to the production of low frequency underwater acoustic sources. It is known that such sources are conventionally produced using a transmitter horn driven by a linear electric motor. To obtain a significant power corresponding to a force applied to the roof and to a displacement of the same roof both important, different types of motor are used requiring the use of particularly expensive high performance materials. The invention makes it possible to use for such use a linear electric motor with variable reluctance controlled by the method according to the invention and produced with materials, in particular soft iron, which are particularly inexpensive. The cost studies have shown that a factor 10 is easily obtained on the cost of a transducer according to the invention compared to a transducer, electrodynamic for example, used to date as well for the control of the vibrations of 'a seat only for a low frequency underwater acoustic source or for other uses.

Claims

1 - Method for controlling a variable reluctance transducer comprising an air gap, characterized in that one determines (201) the magnetic field necessary in the air gap to obtain transducers ^" the desired force, which is measured (203) the real magnetic field existing in the air gap, and that one controls (210) the voltage applied to the transducer to slave (205) linearly the real field on the necessary field.

2 - Method according to claim 1, characterized in that one measures (204) the electric current flowing in the transducer to use (208) this value in the control loop

3 - Linear motor implementing the method according to any one of claims 1 and 2, characterized in that it comprises a first U-shaped magnetic core (405), at least one coil (407) disposed on the one of the branches of the U forming this first core, a second I-shaped core (406) disposed in front of the free ends of the two branches of the U to determine between the first and the second core an air gap, two shims (409,410) arranged on the free ends of the branches of the U to impose a minimum thickness of the said air gap, and a magnetic field probe (408) disposed on the free end of one of the branches of the U and whose thickness is less than l 'thickness of said shims.

4 - Device according to claim 3, characterized in that the two cores (405,406) are fixed respectively on two plates (401, 402) non-magnetic and that they further comprise a set of springs (404) fixed between these two plates for keep them separate, this device making it possible to control the movements of a seat of a vehicle.

5 - Device according to claim 3, characterized in that the second core is fixed to a pavilion intended to generate underwater acoustic waves of low frequency and high intensity