CN1720498A - Contact sensitive device - Google Patents

Abstract

A contact sensitive device comprises a member (12) capable of supporting bending waves, three sensors (16) mounted on the member (12) for measuring bending wave vibration in the member, whereby each sensor (16) determines a measured bending wave signal and a processor which calculates a location of a contact on the member from the measured bending wave signals. The processor calculates a phase angle for each measured bending wave signal and a phase difference between the phase angles of least two pairs of sensors so that at least two phase differences are calculated from which the location of the contact is determined.

Description

Contact-sensing device

Technical field

The present invention relates to contact-sensing device.

Background technology

Visual displays comprises the touch sensitive screen of certain form usually.Along with the appearance of portable multimedia device of future generation (as palmtop computer), this kind phenomenon becomes more prevalent.(this technology produces high frequency waves on the surface of glass screen for Surface Acoustic Wave, SAW) technology, and uses the high frequency waves that caused by the finger contact to decay and come the senses touch position as surface acoustic wave to use the most perfect technology that ripple detects contact.This technology is " flight time (time-of-flight) " type, wherein uses to disturb to arrive the required time senses touch position of one or more sensors.When medium with nondispersive mode effect, when promptly wave propagation velocity can significantly not change on relevant frequency range, this method was feasible.

In the inventor's WO 01/48684 and PCT/GB2002/003073, two kinds of contact-sensing devices and using method thereof have been proposed.In these two parts of applications, this device comprises the parts that can support bending wave vibration and is installed on and be used for the sensor measuring the bending wave vibration of these parts and be used for signal is sent to processor on these parts, and the bending wave vibration that produced these parts of the contact of being done from the surface of these parts changes the information of calculating relevant this contact thus.

Bending wave vibration means that one excites, and for example by contact, this excites and can make these parts that out-of-plane displacement takes place.Many materials all can be crooked, and some material has the pure bending of perfect square root dispersion relation (perfectsquare root dispersion relation), and some material then has the bending that mixes of pure bending and shear-bow.The dependence of speed and wave frequency in the plane of dispersion relation explanation ripple.

Flexural wave has advantage, as improving robustness (robustness) and reduce the susceptibility of surperficial scratch etc.Yet flexural wave is a dispersive wave, thereby promptly crooked wave propagation velocity " flight time " depends on frequency.Generally speaking, pulse comprises large-scale frequency content, if therefore this pulse advances one than short distance, radio-frequency component can at first arrive.In WO 01/48684 and PCT/GB2002/003073, can use and convert measured bending wave signal to a correction, so that the employed technology of radar and sonar field of can using detects contact position from the transmitting signal of non-dispersive wave source.

Summary of the invention

According to an aspect of the present invention, provide a kind of contact-sensing device, this device comprises: the parts that can support flexural wave; Be installed on three sensors of the bending wave vibration that is used for measuring these parts on these parts, each sensor is determined a measured bending wave signal thus; And the processor that calculates the position contacting on these parts from these measured bending wave signals, this contact-sensing device is characterised in that this processor calculates the phasing degree of each measured bending wave signal, calculate the phase difference between the phasing degree of at least two pairs of sensors then, and determine this position contacting from this phase difference.

According to a second aspect of the invention, a kind of method of determining the information relevant with the contact on the contact-sensing device is provided, and the step of this method comprises: three sensors that the parts that can support flexural wave are provided and are installed on the bending wave vibration that is used for measuring these parts on these parts; Position at these parts applies contact; Use each sensor to determine a measured bending wave signal and calculate position contacting, the phase difference between the phasing degree of the method is characterized in that the phasing degree of calculating each measured bending wave signal, calculating at least two pairs of sensors and determine this position contacting from these at least two phase differences that calculated from this measured bending wave signal.

Following feature can be applicable to this device and this method, wherein this processor is adjusted so that the many calculating or the treatment step of this method to be provided.

Can be by an absorber being placed to inhibitory reflex ripple with the edge contact of these parts.The mechanical impedance of this absorber and parts may be selected to be and makes the flexural wave reflection at edge of these parts reduce to minimum.Particularly, selected impedance makes the flexural wave energy center on selected frequencies omega₀Frequency band in absorbed consumingly.The impedance of this absorber may be selected to and has resistance and compatibility concurrently.These impedances may be selected to satisfies following equation:

Z_T＝-iZ_B(ω₀)

Z wherein_TBe termination (termination) impedance of absorber, and Z_BMechanical impedance for this edge-of-part.

This absorber can be made by foam plastic, and these foam plastics can have perforate or closed pore and can be polyurethane or Polyvinylchloride.For example, this foam can be a soft pvc, closed pore is main foam, as MIERS^TM, or intermediate density is to the perforate polyurethane foam of high density.The another kind of suitable foam of having found is the acrylic acid closed-cell foam.These foams can have higher degree of damping and high relatively hardness.This class feature is particularly suitable for firmly, the edge of heavy material such as glass stops.Example comprises 3M serial number 4956,4910,4950 and 4655.This absorber can center on the periphery of these parts in fact and extend.This absorber can be used as erecting frame, is used for this member supporting in framework or to another surface.

The pattern that can comprise projection on the surface of these parts, thus the contact of drawing across this surface provides variable power to produce flexural wave in these parts for these parts.This pattern can be has periodicity or the quasi periodic pattern that statistics goes up the space wave distribution of good definition.This pattern can be random pattern, thereby the contact of advancing on the surface of these parts produces random bend ripple signal.Random relief pattern can be antireflecting coating, anti-dazzle surface finish (finish) or etching light system, as many patterns of knowing on the transparent panel to be found of placing in the front portion of electronic console.

Can be by having with selected frequencies omega₀For the passband at center and bandpass filter with bandwidth deltaf ω are handled each measured bending wave signal.The bandwidth deltaf ω of wave filter is advantageously selected to and solves Doppler (Doppler) effect, and wherein the frequency that arrives when some of flexural wave is different with its original frequency.Therefore, bandwidth preferably meets following relation:

Δω＞＞2k(ω₀)V_max

V wherein_MaxBe maximum lateral speed across the contact on this surface, for example, if this contact is to be provided by felt pen, V then_MaxFor the user can mobile felt pen maximal rate.

The phase place of trap signal can be by measuring with the reference signal comparison for each.Reference signal can have frequencies omega₀Measured phase place is the average phase difference between input and the reference signal, and the best is being measured on 2 π/Δ ω at interval.Perhaps, reference signal can derive from the signal that has filtered from second sensor, and in this case, measured phase place is two phase differences between the input signal.

Phase difference can 2 π/Δ ω the interval calculated, this can be the interval less than 10ms at interval.Can be with reference to being fed to phase detectors with input signal.The output of phase detectors can present by the low-pass filter of cutoff frequency with about Δ ω/2, then by Aristogrid, at last by processor to calculate phasing degree θ.

The instantaneous phase θ of two measured bending wave signals₁(t) and θ_m(t) can satisfy the phase differential equation:

Δθ_1m＝θ₁-θ_m＝k(ω₀)Δx_1m+2πn_1m

Δ x wherein_1m=X₁-X_m, (x_mWith x₁Be respectively distance from contact position to each sensor that indicates m and 1), and k (ω) is a wave vector.If the path length difference between two sensors then can satisfy this equation less than the coherent length of bandpass filter, the coherent length of bandpass filter may be defined as:

$x_c=\frac{2\mathrm{\π}{\mathrm{\ω}}_0}{\mathrm{\Δ\ωk}\left({\mathrm{\ω}}_0\right)}$

Therefore coherent condition is | Δ x_1mThe x of |＜＜_cIf do not satisfy coherent condition, then may not satisfy above-mentioned phase equation.

Therefore, need n_1mDetermine position contacting with the value of phase angle differences.The shape of these parts may be selected to Δ x_1mAmplitude be restricted to value less than half of a wavelength, promptly | Δ x_1m|＜π/k (ω₀).In this case, if Δ x_1mAll possible value satisfy condition | Δ x_1m|＜π/k (ω₀), a n is then only arranged_1mValue is for satisfying | Δ θ_1m-2 π n_1m| the Integer n of＜π_1mPerhaps, n can be estimated or be inferred in some way.

Each phase angle differences is in conjunction with Integer n_1mThe scope of probable value can be in order to producing a series of path length difference, thereby on the surface of these parts a series of discrete dual curve of definition, with the contact position that expresses possibility.Can determine this contact position by drawing defined every the hyperbolic curve of each path length difference and selecting a large amount of hyperbolic curves to intersect or be close to crossing a bit.This point may be real contact position.

If n_1mThe unknown is determined that then the minimal amount of the hyperbolic curve series that contact position is required is three, and is increased the possibility of determining correct contact position by the hyp number that increases the desire drafting.Can use a plurality of sensors, wherein can be every pair of sensor and calculate phase angle differences, thereby produce many hyperbolic curves.In this embodiment, the minimal amount of sensor is three.

Perhaps, if n_1mThe unknown then can be divided into two or more discrete frequency bands with the measured bending wave signal from each sensor, wherein can be each frequency band and is every pair of sensor calculating phase angle differences.Though can calculate a plurality of phase angle differences to sensor from single, the phase angle differences under the different frequency is to derive from identical path length difference.Therefore, the minimal amount of sensor is three.Can realize band segmentation by handling bending wave signal by at least two bandpass filter with different band connection frequencies.For example, use has frequencies omega₀+ ω_δAnd ω₀-ω_δTwo bandpass filter, the phase angle differences Δ θ of two sensors_a, Δ θ_bMay be defined as

Δθ_a＝k(ω₀+ω_δ)Δx+2πn_a

Δθ_b＝k(ω₀-ω_δ)Δx+2πn_b

Wherein Δ x is this contact and the defined single path difference in length of sensor site.

Therefore, can select n_aWith n_bValue, make measured phase angle differences infer the similar value of path length difference.(n is only arranged_a, n_b) value combination satisfies this point.In the case, can determine the actual value of path length difference.Correct combination (n_a, n_b) can be defined as making the minimized value of following formula to make up:

$|\frac{\mathrm{\Δ}{\mathrm{\θ}}_a-2\mathrm{\π}{n}_a}{k({\mathrm{\ω}}_0+{\mathrm{\ω}}_{\mathrm{\δ}})}-\frac{\mathrm{\Δ}{\mathrm{\θ}}_b-2\mathrm{\π}{n}_b}{k({\mathrm{\ω}}_0-{\mathrm{\ω}}_{\mathrm{\δ}})}|$

Path length difference then can be estimated as:

$\mathrm{\Δx}=\frac{1}{2}(\frac{\mathrm{\Δ}{\mathrm{\θ}}_a-2\mathrm{\π}{n}_a}{k({\mathrm{\ω}}_0+{\mathrm{\ω}}_{\mathrm{\δ}})}+\frac{\mathrm{\Δ}{\mathrm{\θ}}_b-2\mathrm{\π}{n}_b}{k({\mathrm{\ω}}_0-{\mathrm{\ω}}_{\mathrm{\δ}})})$

If two pairs of sensors are repeated this processing, then can determine two path length differences, this two path length differences and then can be used for determining contact position.

Perhaps, if n_1mThe unknown then can be used the method for being instructed among WO 01/48684 and the PCT//GB2002/003073 (summarizing as Figure 11) to make the initial of contact position and determine.Then, it is slower to suppose that this contact gear ratio flexural wave moves, so phase angle differences changes small incremental on time scale Δ t.Therefore, can select each value of n to minimize the variation of path length difference.

Measured phase angle differences can comprise the random error that can cause selecting incorrect n value.For example can alleviate this mistake by the possibility that state-space estimator is assessed the continuous sequence of n as Kalman (Kalman) wave filter that is widely known by the people.Selection has the sequence of maximum likelihood measured value.

State-space estimator provides the estimation of the internal state of system (this system is carried out noise measurement).Necessity of state-space estimator is input as the statistical description of the differentiation of system state.One of this state is exemplified as the coordinate system of the Position And Velocity that the object that contacts with these parts is described.What be widely known by the people is, Kalman filter and other state-space estimator can provide the measurement of the corresponding to possibility of model of the sequence of viewed noise measurement and system state.

Therefore state-space estimator can be in order to take at different time (t for example₁, t₂, t₃...) a pair of path length difference done (Δ x for example₁₂With Δ x₃₄) sequence estimate the system states of these times, i.e. position contacting and speed.And, can assess these values and the corresponding to overall possibility of system model of path length difference.

If from phase angle differences sequence and one group of integer (n=n (t₁), n (t₂), n (t₃) ...) obtain this path length difference sequence, then can be in order to infer the possibility of the right value of having selected n by the measured value of the possibility that state-space estimator produced.Reach a conclusion thus, being used to select the method for the correct sequence of Integer n is to find state-space estimator to give the sequence of maximum likelihood measured value.

As mentioned above, some statistical description of state-space estimator using system state evolution.The suitable model that contact is moved can be walking simply at random.Perhaps, this model adopts the how detail statistics explanation of mobile felt pen or finger of user.One is exemplified as user's statistical description of mobile pen how when writing literal or individual characters.

Processor can further be applicable to the available information of determining to comprise in the program any relevant contact desired location at this.For example, if these parts are the input media of graphic user interface, wherein the user can select to press " button ", then usefully, supposes that any contact on these parts takes place in the zone of dispersion corresponding with these buttons.

Perhaps, can use the map of the contingent probability of contact, and this probability is based on user's anticipatory behavior.This device can comprise the software application with graphic user interface (GUI), and it utilizes application programming interfaces (API) and operating system mutual, and wherein API is applicable to the generation probability map.This probability map can be based on position, size and the frequency of utilization of the object that graphic user interface presented.This probability map also can be based on the information of the relative possibility of the relevant various GUI elements that are activated.

Following feature can be applicable to all embodiment of the present invention.This device can comprise recording member, is used for writing down the measured bending wave signal from this sensor or each sensor in time when this contact is moved across these parts.Can in central processing unit, calculate the information relevant with contact.These sensors can be installed on the edge of these parts or separate with the edge of these parts.Sensor can adopt the form that bending wave vibration can be converted to the sensing converter of analog input signal.

These parts can adopt the form of plate or panel.These parts can be transparent or nontransparent, for example have printed patterns.These parts can have homogeneous thickness.Perhaps, these parts can have more complicated shape, for example curved surface and/or variable thickness.

This device can be pure passive sensor, wherein produces bending wave vibration by initial impact or the frictional movement by this contact, thereby produces measured bending wave signal.The form that this contact can adopt finger touch or felt pen (it can be the form of hand-held pen) to touch.Felt pen moving on these parts can produce continuous signal, and this signal is subjected to the influence of position, pressure and the speed of felt pen on these parts.This felt pen can have flexible tips, for example, the rubber tip, and this tip is by applying variable power and produce flexural wave in these parts to these parts.This variable power can be provided by surface that is attached to these parts or the tip of sliding across the surface of these parts.When tip during across this parts surperficial mobile, can produce tension force, this tension force can cause any bonding between this tip and this parts to break at certain critical value place, thereby allows most advanced and sophisticatedly to slide across this surface.Flexural wave can have supersonic zone (＞frequency content in 20kHz).

These parts also can be acoustic radiators, and selector can be installed on these parts, so that excite bending wave vibration in these parts, export with generation sound.The frequency band of the sound signal of this converter preferably is different from and is not overlapped in the frequency band of the measurement of sensor.Sound signal thereby can be filtered, for example, audio band can be limited to the following frequency of 20kHz, and vibration survey can be limited to the above frequency of 20kHz.Sensor can have dual-functionality and be used as selector.

Should or each selector or sensor can be bending transducer, itself and this parts for example piezoelectric transducer directly weld.Perhaps, should or each selector or sensor can be the inertial transducer that is coupled with this parts at the single-point place.Inertial transducer can be electrodynamic transducer or piezoelectric transducer.

Can be contained in mobile phone, laptop computer or the personal digital assistant according to contact-sensing device of the present invention.For example, the keypad that is assemblied in mobile phone traditionally can be substituted by the continuous mould (moulding) that according to the present invention is the touch-sensitive type.In laptop computer, being used as the touch pad that mouse controller plays a role can be substituted for the continuous mould according to contact-sensing device of the present invention by it.Perhaps, this contact-sensing device can be indicator screen, for example comprises the liquid crystal display equipment screen of liquid crystal, and it can be used for exciting or the sensing flexural wave.Indicator screen can present the information relevant with contact.

Description of drawings

By example in the accompanying drawings summary the present invention has been described, wherein:

Fig. 1 is the schematic plan view of touch sensitive device according to an aspect of the present invention;

Fig. 2 is the perspective illustration of the device of Fig. 1;

Fig. 3 is the diagrammatic side view of one dimension beam;

Fig. 4 a is the curve map of the amplitude of explanation reflection coefficient to frequency (Hz), because of amplitude is so that a ratio is no unit;

Fig. 4 b is the curve map of the phase place (is unit with the radian) of explanation reflection coefficient to frequency (Hz);

Fig. 5 a and 5b are the perspective illustration of alternative touch sensitive devices;

Fig. 6 is for finding the process flow diagram of the method for contact position according to the present invention;

Fig. 7 a is the block schematic diagram that is used to calculate the equipment at phasing degree;

Fig. 7 b is the block schematic diagram in conjunction with the equipment of the equipment use of Fig. 7 a;

Fig. 8 a to 8d is the planimetric map according to equipment of the present invention, and the hyperbolic curve of path length difference is described;

Fig. 9 is the block schematic diagram that is used to calculate the replacement at phasing degree;

Figure 10 calculates the process flow diagram of the alternative method of contact position for explanation;

Figure 11 calculates the process flow diagram of the method for contact position for the related function that uses dispersion corrected;

Figure 11 a is the related function of dispersion corrected to the curve map of time, and

Figure 12 a is the block schematic diagram of the contact-sensing device of double as one loudspeaker operation, and

The method of Figure 12 b explanation separating audio signals and measured bending wave signal in the device of Figure 12 a.

Embodiment

Fig. 1 illustrates a contact-sensing device 10, and it comprises the transparent touchsensitive plate 12 that is installed on display device 14 the place aheads.Display device 14 can adopt the form of TV, computer screen or other visual display units.Use the feltpen 18 of pen form that literal 20 or other guide are write on the touch-sensitive pads 12.

Transparent touchsensitive plate 12 is for supporting the parts of bending wave vibration, for example an acoustic device.As shown in Figure 2, foursensors 16 that will be used for measuringplate 12 bending wave vibrations are installed on the downside of thisplate.Sensor 16 adopts the form of piezoelectric vibration sensor, and in the every nook and cranny ofplate 12 sensor is installed.At least onesensor 16 also can be used as a selector, is used for exciting bending wave vibration at plate.In this way, this device can be used as a combination speaker and contact-sensing device.

The erecting frame of being made by foam plastic 22 is to be attached to the downside ofplate 12 and in fact around the periphery ofplate 12 and extend.Erecting frame 22 has the stickiness surface, thereby these parts can be attached to any surface securely.The mechanical impedance of erecting frame and plate makes the flexural wave reflection of panel edges reduce to minimum through selecting.

Relation between the mechanical impedance of erecting frame and plate can be similar to by considering one-dimensional model shown in Figure 3.This model comprises thewaveguide 34 of beam (beam) form, and it ends at has theedge erecting frame 36 that stops impedance.Theincident wave 38 of advancing downwards alongwaveguide 34 is to formreflection wave 40 by erectingframe 36 reflections.Incident and reflection wave are the plane waves of advancing along perpendicular to the direction at edge.Suppose that erectingframe 36 satisfies following boundary condition:

(i) stop impedance and only be coupled into transverse velocity, promptly it does not provide any moment of torsion resistance; Thereby bending moment equals zero in edge, and

(ii) the ratio of edge's cross shear and speed equals terminal impedance;

The reflection coefficient of erecting frame is given by following formula:

$R\left(\mathrm{\ω}\right)=\frac{-Z_T/Z_B\left(\mathrm{\ω}\right)-i}{Z_T/Z_B\left(\mathrm{\ω}\right)+i}$

Z wherein_TBe the termination impedance of erecting frame, and Z_BFor the mechanical impedance of waveguide end, given by following formula

$Z_B\left(\mathrm{\ω}\right)=\frac{{\mathrm{Bk}}^3\left(\mathrm{\ω}\right)}{2\mathrm{\ω}}(1+i)$

Wherein k (ω) is a wave vector, and it represents according to the bending hardness B of panel and the quality μ of per unit area,

$k={\left(\frac{\mathrm{\μ}}{B}\right)}^{1/4}\sqrt{\mathrm{\ω}}$

Therefore, reflection coefficient depends on the ratio of the impedance of waveguide end and erecting frame.In addition, the impedance of waveguide is directly proportional with the square root of frequency, and has reality and empty (reactive) composition (being π/4 phasing degree) that weight equates.Therefore, reflection coefficient may have much relations with frequency.

If following condition is met, then reflection coefficient disappears, and is promptly centering on ω₀Frequency band in strong absorption flexural wave energy:

Z_T＝-iZ_B(ω₀)

Therefore, the termination impedance of this erecting frame must have reality and transient element concurrently, or equivalently, this erecting frame must have resistance and compatibility concurrently.

This plate can be for example thick polycarbonate sheet of 1mm, and its per unit area quality is μ=1.196kgm^-2, and bending hardness is B=0.38Nm.Above equation can be used for calculating the impedance and the selected angular frequency of strong absorption of plate₀The impedance of the absorber that=2 π (900Hz) flexural wave energy on every side is required.

The per unit width impedance (impedance) of plate (1mm beam approximate value) is:

Z_B(ω₀)＝(1+i)33.8Nsm^-2。

Provide required absorption absorber characteristic thereby be:

The resistance of per unit width (resistance),

Re(Z_T)＝Im[Z_B(ω₀)]＝33.8Nsm^-2。

The hardness of per unit width,

-i?Im(Z_T)ω₀＝Re[Z_B(ω₀)]ω₀＝1.91×10⁵Nm^-2。

Reflection coefficient is no complex unit.Fig. 4 a and 4b for the amplitude of explanation reflection R (ω) and phase place with the curve map of frequency change.Work as ω₀When being approximately equal to 900Hz, the amplitude of reflection coefficient is zero, and the phase place of reflection coefficient is paraphase.

In Fig. 5 a and 5b,plate 12 has uniform surface roughness and adopts the form of pattern of raisedsurfaces 28,29.30 across thedilatory felt pen 18 in this surface along the path, when feltpen 18 by pattern on when a bossing or line, just in these parts, produce flexural wave 32.Therefore, the contact offelt pen 18 provides bending wave vibration source in these parts.In Fig. 5 a, picture onsurface 28 is the periodic patterns of protruding cross spider, and in Fig. 5 b, picture on surface 29 is a random relief pattern.

In the specific embodiment of Fig. 2,5a and 5b, when on the rough surface of contact at these parts when mobile, flexural wave in these parts from contact point isotropy eradiation.These parts at distance x place from the displacement of contact point by a transfer function H (ω; X) and relevant with the displacement at contact point place.During greater than wavelength X=2 π/k (ω), transfer function can be approximately in distance,

$H(\mathrm{\ω};x)=\frac{A}{\sqrt{k\left(\mathrm{\ω}\right)x}}{e}^{\mathrm{ik}\left(\mathrm{\ω}\right)x}$

Wherein A is that constant and k (ω) are the previous wave vector that defines.Though H (ω strictly speaking; X) be only applicable to flexural wave on the infinite plate, but because erecting frame strong absorption bending wave vibration, so this relation is satisfied.Transfer function shows that if flexural wave source emission pure sinusoid frequency, angular frequency is ω₀, then for this source, at distance x₁With x₂Phase difference Δ θ between two positions at place and the displacement of contact point₁₂For

exp(iΔθ₁₂)＝exp[ik(ω₀)(x₁-x₂)]

This means phase angle differences, path length difference Δ x=(x₁-x₂) and an Integer n₁₂Between following relationship.

Δθ₁＝θ₁-θ₂＝k(ω₀)Δx₁₂+2πn₁₂

Fig. 6 illustrates the step in the method for using this equation to determine contact position:

A) use each sensor to measure a bending wave signal, so that the bending wave signal of having measured W to be provided_i(t) and W_j(t),

B) calculate measured bending wave signal W_i(t) and W_j(t) phasing degree θ_i(t) and θ_j(t),

C) calculate two phasing degree θ_i(t) and θ_j(t) difference between,

D) calculate contact position from following formula:

k(ω₀)Δx_ij＝Δθ_ij-2πn_ij

The block schematic diagram of Fig. 7 a explanation one device, this device is used to calculate the measured flexural wave W of one of these sensors_j(t) phasing degree θ_jSignal W_j(t) be random signal, thus uncorrelated during long-time scale.This signal is at first amplified by amplifier 42, is handled by analog band-pass filter 44 then, and the passband of this bandpass filter is with ω₀Be the center, and bandwidth is Δ ω.

The provable Doppler effect in mobile source of flexural wave wherein has frequencies omega₀And has ω when arriving this by the flexural wave of being launched with the source of a bit moving of speed v on parts₀-k (ω₀) the defined different frequency of v.Therefore, the maximum angular frequency deviation between the flexural wave at two difference places is 2k (ω on these parts₀) v_Max, v wherein_MaxFor moving the maximal rate in source.If angular frequency deviation becomes greater than the width of bandpass filter, then above phase difference equation is false.Therefore, the bandwidth deltaf ω of wave filter 44 is set at greater than this maximum frequency deviation, thereby meets following relation:

Δω＞＞2k(ω₀)v_max

After wave filter 44 processing, the W ' of trap signal that is produced_j(t) be one to have frequencies omega₀Amplitude and phase place MCW modulated carrier wave, and defined by following formula:

W′_j(t)＝A_j(t)sin[ω₀t+θ_j(t)]

A wherein_j(t) and θ_j(t) be the amplitude and the phase place of this signal.The both is promptly fluctuateed on Δ t=2 π/Δ ω by the determined time scale Δ of the bandwidth of wave filter t.Can export the maximum frequency of making the independent phase angular measurement from bandpass filter is 1/ Δ t.Because the common every 10ms of touch sensor provides a contact position measuring that upgrades, the condition of the minimum frequency of position measurement is Δ t＜10ms.

The signal W ' that will filter then_j(t) be sent to two analogue phase detectors 46 simultaneously.This type of detecting device is widely known by the people in the prior art, for example, and referring to " the The Art ofElectronics (electronic technology) " the 644th page of Horowitz and Hill.Also each all had frequencies omega₀But phase difference is the reference signal of pi/2 is fed to two phase detectors.The low-pass filter 48 of the cutoff frequency of the output of phase detectors by respectively having about Δ ω/2.The output of low-pass filter respectively with cos (θ_j) and sin (θ_j) be directly proportional.These outputs are by in addition digitizing and handled by processor 52 ofAristogrid 50, so that phasing degree θ is provided then_j

How reference signal used among Fig. 7 b key diagram 7a can produce.Measure the second bending wave signal W at the second sensor place_i(t).This signal is presented by amplifier 42 and analog band-pass filter 44, so that produce the signal W ' that has filtered_j(t).The signal W ' that has filtered_j(t) form the reference signal that directly is fed to phase detectors 46.Also the signal that will filter via a device is fed to second phase detectors 46, and this device is with the phase deviation pi/2 of this signal.Use the reference signal of phase-shift signal as second phase detectors 46.

How Fig. 8 a to 8d explanation phase angle differences thereby path length difference are used to calculate contact position.Equation definition in the step of Fig. 6 (d) can be covered in the hyperbolic curve on the plate 12.Three different n of a pair of sensor 16 (each end of the short side ofplate 12 is all installed one) are used in Fig. 8 a explanation_1mValue and threehyperbolic curves 26 that phase angle differences produced that calculated.Similarly, Fig. 8 b and 8c explanation by two to other phase angle differences of sensor and different n_1mThehyperbolic curve 26 that is worth and produces.Fig. 8 d illustrates the whole hyperbolic curves by sensorproduced.Contact position 24 is three hyp intersection points, and every hyperbolic curve comes from every pair of sensor.Can infer n fromcontact position 24_1mRight value.

Can use the specific embodiment shown in Fig. 9 to implement to infer the method for n.The bending wave signal W that each sensor is measured₁(t) handle simultaneously by twobandpass filter 48,54.Calculate two phasing degree, each corresponding one in each wave filter, for example, as described in Figure 7.Wave filter 48,54 has different slightly band connection frequencies, wherein provides two phase angle differences by every pair of sensor, each corresponding difference of each band connection frequency.

Phase angle differences Δ θ from sensor_a, Δ θ_bMay be defined as:

Δθ_a＝k(ω₀+ω_δ)Δx+2πn_a

Δθ_b＝k(ω₀-ω_δ)Δx+2πn_b

Wherein Δ x is contact and the defined single path difference in length of sensor site.

Correct combination (n_a, n_b) can be defined as making the minimized value of following formula to make up:

$|\frac{\mathrm{\Δ}{\mathrm{\θ}}_a-2\mathrm{\π}{n}_a}{k({\mathrm{\ω}}_0+{\mathrm{\ω}}_{\mathrm{\δ}})}-\frac{\mathrm{\Δ}{\mathrm{\θ}}_b-2\mathrm{\π}{n}_b}{k({\mathrm{\ω}}_0-{\mathrm{\ω}}_{\mathrm{\δ}})}|$

Path length difference then can be estimated as:

$\mathrm{\Δx}=\frac{1}{2}(\frac{\mathrm{\Δ}{\mathrm{\θ}}_a-2\mathrm{\π}{n}_a}{k({\mathrm{\ω}}_0+{\mathrm{\ω}}_{\mathrm{\δ}})}+\frac{\mathrm{\Δ}{\mathrm{\θ}}_b-2\mathrm{\π}{n}_b}{k({\mathrm{\ω}}_0-{\mathrm{\ω}}_{\mathrm{\δ}})})$

Another then can be used for determining second path length difference to sensor.Each path length difference defines double curve on this panel.These two hyp intersection points are contact position.Shown in Fig. 8 a to 8d, draw hyperbolic curve, and the hyp intersection point of maximum number is real contact position probably.

Figure 10 explanation is used for calculating from above equation the alternative method of contact position, that is:

I. measure pair of curved ripple signal W_i(t) and W_j(t), each signal is measured by a sensor respectively;

Ii. use the method described in Figure 11 and the 11a to calculate the related function of the dispersion corrected of two signals;

Iii. use the related function of dispersion corrected to calculate the initial position of contact, as described in Figure 11 and 11a;

Iv. remeasure bending wave signal W_i(t) and W_j(t);

V. calculate the phasing degree of each signal-for example, as described in Fig. 7 a and 7b;

Vi. calculate the difference between the phasing degree;

Vii. selection makes the n of the minimize variations of path length difference_1mValue;

Viii. draw by the defined hyperbolic curve of following formula:

k(ω₀)Δx_ij＝Δθ_ij-2πn_ij

Ix. repeating step is (iv) to (viii), remeasuring the bending wave signal of regular interval of delta t (for example Δ t=2 π/Δ ω).

(viii), need determine contact position to two hyp minimum value of sensor in step from difference.Therefore, be necessary at least two pairs of sensors and carry out whole procedure simultaneously.Therefore, must determine the minimum value of two phase angle differences.As described in Figure 9, produce two phase angle differences by using two sensors and this signal being divided into two frequency bands.Perhaps, can use a plurality of sensors, so that use different right sensors to calculate a plurality of phase angle differences.

Figure 11 illustrates that the related function that calculates dispersion corrected is to show the method for the path length difference between contact position and the sensor.Method proposed below has been summarized the information among the PCT/GB2002/003073.This method comprises the following steps:

(a) measure two bending wave signal W₁(t) and W₂(t);

(b) calculate W₁(t) and W₂(t) Fourier transform is to obtainWithAnd thereby obtain intermediate functionWhereinBe the complex conjugate Fourier transform, t represents the time, and ω is 2 π f, and wherein f is a frequency.

(c) calculate the second intermediate function M (ω), it is

Function

(d) with (e) in execution in step (a) in (c), use predetermined panel dispersion relation$k={(\mathrm{\μ}/B)}^{1/4}\sqrt{\mathrm{\ω}}$Come calculated rate to extend computing$f\left(\mathrm{\ω}\right)=v{(\mathrm{\μ}/B)}^{1/4}\sqrt{\mathrm{\ω}}.$

(f) combination M (ω) with$f\left(\mathrm{\ω}\right)=v{(\mathrm{\μ}/B)}^{1/4}\sqrt{\mathrm{\ω}}$To obtain the related function of dispersion corrected:$G\left(t\right)=\frac{1}{2\mathrm{\π}}{\∫}_{-\∞}^{+\∞}M\left[f\left(\mathrm{\ω}\right)\right]\exp \left(\mathrm{i\ωt}\right)\mathrm{d\ω};$And

(g) for the related function of time drafting dispersion corrected, peak value occurs in time t₁₂The place is shown in Figure 11 a;

(h) from t₁₂Calculate Δ x₁₂Δ x₁₂Be path x from first and second sensor to contact₁With x₂Between path length difference.

(i) Δ x₁₂The definition double curve, it can be drawn as shown in Figure 7, to calculate contact position.

As use the method for Figure 10, need two hyp minimum value determine contact position.Therefore, the how hyp mode of above-mentioned generation can be applicable to the method.

The second intermediate function M (ω) may simply be

It can provide the related function of the dispersion corrected of a standard.Perhaps, M (ω) can be from selecting the array function down, and these functions all can produce the phase place equivalent function of related function of the dispersion corrected of standard:

$a)---M\left(\mathrm{\ω}\right)=\frac{{\hat{W}}_1\left(\mathrm{\ω}\right){\hat{W}}_2^{*}\left(\mathrm{\ω}\right)}{\left|{\hat{W}}_1\left(\mathrm{\ω}\right){\hat{W}}_2^{*}\left(\mathrm{\ω}\right)\right|}$

$b)---M\left(\mathrm{\ω}\right)=\frac{{\hat{W}}_1\left(\mathrm{\ω}\right){\hat{W}}_2^{*}\left(\mathrm{\ω}\right)}{\sqrt{\left|{\hat{W}}_1\left(\mathrm{\ω}\right){\hat{W}}_2^{*}\left(\mathrm{\ω}\right)\right|}}$

Wherein (x) is a real-valued function

$d)---M\left(\mathrm{\ω}\right)={\hat{W}}_1\left(\mathrm{\ω}\right){\hat{W}}_2^{*}\left(\mathrm{\ω}\right)\mathrm{\ψ}\left(\mathrm{\ω}\right),$Wherein ψ (ω) is a real-valued function

Perhaps, M (ω) can be function

It is related function D (t):$D\left(t\right)={\∫}_{-\∞}^{+\∞}{W}_1(t+t^{\′})W_2\left(t^{\′}\right)d{t}^{\′}$Fourier transform.

These steps are for calculating D (t); Calculate

And use a frequency and extend computing to draw the related function of dispersion corrected:$G\left(t\right)=\frac{1}{2\mathrm{\π}}{\∫}_{-\∞}^{+\∞}\hat{D}\left[f\left(\mathrm{\ω}\right)\right]\exp \left(\mathrm{i\ωt}\right)\mathrm{d\ω}.$

Perhaps, at step (f), can calculate the related function of following dispersion corrected:

$G\left(t\right)=\frac{1}{2\mathrm{\&pi;}}{\&Integral;}_{-\&infin;}^{+\&infin;}\hat{{\mathrm{<figure-callout\; label="W"\; filenames="A20038010477500213.png,A20038010477500271.png"\; state="\{\{state\}\}">W</figure-callout>}}_1}\left[f\left(\mathrm{\&omega;}\right)\right]{\hat{W}}_2^{*}\left[f\left(\mathrm{\&omega;}\right)\right]{\mathrm{\&phi;}}_{12}\left[f\left(\mathrm{\&omega;}\right)\right]\exp \left(\mathrm{i\&omega;t}\right)\mathrm{d\&omega;}$

Wherein${\mathrm{\&phi;}}_{12}^{*}\left(\mathrm{\&omega;}\right)=|\underset{j}{\mathrm{\&Sigma;}}{\hat{W}}_{1,j}\left(\mathrm{\&omega;}\right){\hat{W}}_{2,j}^{*}\left(\mathrm{\&omega;}\right)\exp [-\mathrm{ik}\left(\mathrm{\&omega;}\right){\mathrm{\&Delta;x}}_j]|$

Wherein

With

Be two measured bending wave signal { W_{1, j}(t) } with { W_{2, j}(t) } Fourier transform and complex conjugate Fourier transform, and { Δ X_jIt is path length difference.

One sensor can be used as first and second sensor both, wherein the related function of this dispersion corrected is an autocorrelation function.Can use W₁(t)=W₂(t) related function of this dispersion corrected is used identical step and calculated autocorrelation function.

Figure 12 a illustrates the contact-sensing device of double as loudspeaker operation.Figure 12 b explanation is used for sound signal and measured signal be divided into two different frequency bands so that suppress the method for sound signal to the contribution of the measured signal handled.This device comprises parts 106, wherein passes through a selector or driver 108 and contacts the generation flexural wave.Selector puts on parts 106 to produce output with a sound signal.Before putting on these parts, filter this sound signal by a low-pass filter 112, shown in Figure 12 b, this wave filter removes threshold frequency f₀Above sound signal.

Shown in Figure 12 b, this contact produces a signal, and the power output of this signal is constant in fact on big frequency band.To be obtained a composite signal mutually with this sound signal from the signal of this contact, this combined type signal by Hi-pass filter 114 to remove threshold frequency f₀Above signal.The signal that this has been filtered is sent on Aristogrid 116 and the processor 118 then.

Claims (31)

1. a contact-sensing device comprises: the parts that can support flexural wave; Be installed on three sensors of the bending wave vibration that is used for measuring these parts on these parts, thereby each sensor is determined a measured bending wave signal; And the processor that calculates the position contacting on these parts from these measured bending wave signals, it is characterized in that, this processor calculates the phase difference between the phasing degree of the phasing degree of each measured bending wave signal and at least two pairs of sensors, so that calculate at least two phase differences, determine this position contacting from these at least two phase differences.

2. contact-sensing device as claimed in claim 1 is included in the absorber of the edge of these parts, thereby reflection wave is suppressed.

3. contact-sensing device as claimed in claim 2 is wherein selected the mechanical impedance of this absorber and these parts so that reduce to minimum from the flexural wave reflection at the edge of these parts.

4. contact-sensing device as claimed in claim 3 is wherein selected these impedances, makes the flexural wave energy center on selected frequencies omega₀Frequency band in by strong absorption.

5. contact-sensing device as claimed in claim 4, wherein select these impedances to satisfy following equation:

Z_T＝-iZ_B(ω₀)

Z wherein_TBe the termination impedance of absorber, and Z_BMechanical impedance for the edge of these parts.

6. as claim 4 or 5 described contact-sensing devices, comprise the bandpass filter that is used to filter each measured bending wave signal, this wave filter has with this selected frequencies omega₀Passband and bandwidth deltaf ω for the center.

7. contact-sensing device as claimed in claim 6, wherein the bandwidth deltaf ω of this wave filter meets following relation:

Δω＞＞2k(ω₀)v_max

V wherein_MaxMaximum lateral speed for this contact.

8. as each described contact-sensing device in the claim 2 to 7, wherein this absorber is made by foam plastic.

9. each described contact-sensing device in the claim as described above, wherein these parts comprise the pattern of projection in its surface, thereby provide a power across the contact that draw on this surface for these parts, to produce flexural wave in these parts.

10. contact-sensing device as claimed in claim 9, wherein this pattern is a random pattern, thus the contact of advancing on this surface of these parts produces random bend ripple signal.

11. contact-sensing device as claimed in claim 10, wherein this pattern is formed by antireflecting coating, anti-dazzle surface finish or etching light system.

12. each described contact-sensing device in the claim as described above, comprise at least two bandpass filter, these bandpass filter have different band connection frequencies and handle simultaneously by the measured bending wave signal of a pair of sensor, thereby the phase angle differences of each band connection frequency is provided by a pair of sensor.

13. each described contact-sensing device in the claim is included in four sensors on these parts as described above.

14. each described contact-sensing device in the claim as described above comprises: use the related function of many dispersion corrected to measured bending wave signal to determine the member of the initial position of this contact; And use many members of phase angle differences between the measured bending wave signal being determined the follow-up location of this contact.

15. each described contact-sensing device in the claim as described above, wherein this phasing degree determines that member comprises phase detectors.

16. contact-sensing device as claimed in claim 15, wherein this processor comprises low-pass filter and Aristogrid, is used for determining this phasing degree.

17. each described contact-sensing device in the claim as described above, wherein these parts are acoustic radiator, and selector is installed on these parts exciting bending wave vibration in these parts, thus the output of generation sound.

18. contact-sensing device as claimed in claim 17, it comprises and is used for guaranteeing that this output and measured bending wave signal are in the member of discrete frequency band.

19. each described contact-sensing device in the claim as described above, wherein these parts are transparent component.

20. a method of determining the information relevant with the contact on the contact-sensing device, three sensors that comprise the steps: to provide the parts that can support flexural wave and be installed on the bending wave vibration that is used for measuring these parts on these parts; Position at these parts applies contact; Use each sensor to determine a measured bending wave signal and calculate this position contacting from this measured bending wave signal, it is characterized in that, calculate each measured bending wave signal the phasing degree, calculate the phase difference between the phasing degree of at least two pairs of sensors and determine this position contacting from described at least two phase differences that calculated.

21. method as claimed in claim 20 comprises the inhibitory reflex ripple by the edge that an absorber is positioned over these parts.

22. method as claimed in claim 21 comprises the mechanical impedance of selecting this absorber and these parts so that reflect from the flexural wave of the edge of these parts and to reduce to minimum.

23. method as claimed in claim 22 comprises and selects these impedances to make the flexural wave energy center on selected frequencies omega₀Frequency band in by strong absorption.

24. method as claimed in claim 23 comprises and selects these impedances to satisfy following equation:

Z_T＝-iZ_B(ω₀)

Z wherein_TBe the impedance of this absorber, and Z_BImpedance for the edge of these parts.

25., comprise that this bandpass filter has with this selected frequencies omega by each measured bending wave signal of a band-pass filter as claim the 23 or 24 described methods₀Passband and bandwidth deltaf ω for the center.

26., comprise and use following phase difference equation as each described method in the claim 20 to 25:

Δθ_1m＝θ₁-θ_m＝k(ω₀)Δx_1m+2πn_1m

Determine this position contacting, wherein θ_iBe the phasing degree of a measured bending wave signal, x_iBe distance, Δ x from this contact position to each sensor_1m=X₁-x_mBe the path length difference of two sensors, k (ω) is wave vector and n_1mIt is a unknown integer.

27. method as claimed in claim 26 comprises and selects these parts with Δ x_1mAmplitude be restricted to value less than half of wavelength so that from | Δ θ_1m-2 π n_1m|＜π determines n_1m

28. method as claimed in claim 26 comprises that the related function of the dispersion corrected of using a pair of measured bending wave signal is determined the initial position of this contact, and selects to make variation in this path length difference to reduce to the n of minimum_1mValue.

29. method as claimed in claim 26 comprises and selects a series of n_1mValue, the combination of the value that this is serial and each phase angle differences is drawn the serial curve figure of these path length differences to define a series of path length difference, and from a large amount of these curve maps crossing a bit infer real n_1mValue.

30. as each described method in the claim 20 to 29, comprise from these and many a plurality of phase angle differences calculated at the phasing degree, draw the curve map of each path length difference and a bit be chosen as this position contacting with what a large amount of hyperbolic curves intersected.

31., comprise these the measured bending wave signals from each sensor are divided at least two discrete frequency bands, and the phase angle differences of calculating a pair of sensor at each frequency band as each described method in the claim 20 to 30.

Images