WO1999032963A1 - Air data sensor apparatus and method - Google Patents

Abstract

The invented apparatus (1) measures and generates air data indicative of the conditions experienced by an aircraft (17), and includes at least one sensor (2) and an associator (6). The sensor (2) detects air pressure exerted against the aircraft's external surface (3), and generates a sensor signal based on the sensed air pressure. Preferably, the sensor (2) includes a surface that deflects under the exerted air pressure, and the amount of deflection at separate locations of the surface is used to generate the sensor signal. Also, the apparatus (1) can include a conformal member (4) that is mounted to the sensor (2). The conformal member (4) supports or comprises the external surface of the aircraft, and is flush with the aircraft surface (3) in proximity to the conformal member to preserve the aircraft's aerodynamic shape. The associator (6) is coupled to the sensor (2) to receive the sensor signal, and maps the sensor signal level to a corresponding air data quantity such as airspeed, side slip, angle-of-attack, and static and dynamic air pressure. Preferably, the associator (6) is implemented as a learning system such as a neural network, that can be trained to generate accurate air data based on the sensor signal. The invention also includes a related method.

Description

AIR DATA SENSOR APPARATUS AND METHOD Statement of Government Rights in the Invention

This invention was developed under a Small Business Innovative Research project funded by the U.S. Government as represented by N.A.S.A. under Contract No. NAS-1-50288. The U.S. Government has certain rights in the invention.

Technical Field

This invention is directed to an apparatus and method for sensing air data for use in the control of an aircraft. The air data can include variables such as air speed, side-slip, angle-of-attack, static air pressure, dynamic air pressure and the like.

Background Art

To properly control an aircraft, various air data quantities such as air speed, side-slip, angle-of-attack, static air pressure, or dynamic air pressure, must be determined. In many conventional aircraft, air data is sensed with a boom or vane that extends from the aircraft into the freestream where air flow is not disturbed by the aircraft's surface. Such booms or vanes suffer from various disadvantages that render their use undesirable or detrimental in many circumstances. For example, such conventional air data sensors add radar profile to the aircraft, a disadvantage particularly in military applications in which the aircraft is desired to have as small of a radar profile as possible to avoid its detection. In addition, they are vulnerable to high temperatures and pressures commonly experienced by high-performance and high-speed aircraft. It would be desirable to overcome these disadvantages of conventional air data sensors.

In addition, many conventional air data sensor systems calculate air data quantities based on fixed algorithms programmed into the sensor system. However, it is well-known that different aircraft even of the same type have idiosyncrasies in performance that are not predicted by such algorithms. Thus, the air data generated by the aircraft's sensor system may deviate greatly from actual air conditions, a problem that degrades performance unless substantial calibration is performed through flight tests. Although such algorithms may in some cases be calibrated to the actual air conditions, such modifications of the generalized sensor algorithm are generally highly difficult to implement because they require programming of complicated, non-linear relationships that are difficult to ascertain and implement in the sensor system. It would be desirable to overcome these disadvantages of conventional air data sensors.

Disclosure of the Invention

This invention overcomes the above-noted disadvantages. An apparatus in accordance with this invention senses air pressure encountered by an aircraft, and generates one or more air data signals based on the sensed air pressure. The apparatus is characterized by at least one sensor and an associator. The sensor is mounted to detect air pressure exerted against an external surface of the aircraft, and generates a sensor signal indicative of the sensed air pressure. Preferably, the sensor includes a deflectable surface that supports capacitive plates that are positioned apart from one another, and whose spacing from opposing plates is determined by the degree and manner in which the surface is deflected by the exerted air pressure. The sensor generates signals whose frequencies depend upon the plate spacings, and uses these signals to generate a sensor signal. The associator is coupled to receive the sensor signal, and generates one or more air data signals based on the sensor signal(s). The air data signals can be indicative of air speed, side-slip, angle-of-attack, static air pressure, or dynamic air pressure, for example.

The associator can be implemented as a learning system to give the associator the capability to be updated with new flight and air data, or to adapt to changes in aircraft performance such as would occur if the aircraft were damaged in flight, for example. The learning system can be a neural network coupled to receive the signal from the sensor. The learning system functions to map the sensor signal level(s) to a corresponding air data level. The learning system can be trained with a mapping derived by recording air data levels from conventional sensors for air data conditions existing during flight of the aircraft, and recording the sensor signal levels corresponding to such air data levels. Alternatively, the learning system can be trained with a mapping of sensor signal levels to corresponding air data levels derived by subjecting the aircraft to various predetermined air conditions in a wind tunnel, for example, and recording the resulting sensor signals in association with the air data conditions that generated such sensor signals. Because the learning system can "teach" itself the appropriate mapping of the sensor signal to a corresponding air data signal, the invented apparatus is a vast improvement over conventional air data sensor systems in which a function of the sensed air pressure to air data quantity must be derived and programmed into such conventional system.

Rather than using a learning system, an aircraft can be tested under air conditions experienced in flight or under predetermined air conditions in a wind tunnel. Sensor signal levels resulting from such air conditions can then be recorded in association with corresponding air data levels to derive a mapping of the sensor signal level to the air data signal level. A look-up table, programmable logic array, microcontroller, fuzzy logic associator, or processor coupled to a memory, can be programmed to implement the mapping to generate air data based on the sensor signal(s).

Preferably, the apparatus includes a conformal member that is mounted to the sensor. The conformal member supports or comprises the external surface of the aircraft, and is flush therewith. Accordingly, the conformal member preserves the aircraft's shape so that air flow is not undesirably disturbed by the presence of the sensor. Thus, the invented apparatus is suitable even for high-speed aircraft in which any disturbance of the air flow could be highly detrimental. The conformal member also protects the sensor from extreme flight conditions. The apparatus can also include a deformable cover mounted to the external surface of the aircraft, to which the conformal member is attached. The deformable cover serves to seal the interior of the aircraft surface from the environment, but yet deflects to allow force exerted by air pressure to be transmitted to the conformal member and hence to the sensor. The presence of the conformal member behind the cover substantially preserves the shape of the exterior surface of the aircraft and thus maintains aerodynamic integrity.

The invented apparatus can include an interface unit coupled between the sensor and the associator. The interface unit preferably includes a microcontroller that receives the sensor signal, and that generates an air pressure signal based on the sensor signal. The interface unit supplies the air pressure signal to the associator that generates the air data signal based on the air pressure signal. The associator supplies the generated air data signal to the interface unit. The interface unit is coupled to supply the received air data signal to the aircraft's flight control system for use in controlling the aircraft. The interface unit can also be coupled to output the air data signal to a pilot display for use by the pilot in monitoring and controlling the aircraft performance.

The invented method is characterized by a step of sensing air pressure exerted against one or more portions of an aircraft surface, and mapping a level of the sensed air pressure to a corresponding air data level to generate an air data signal. The air data signal can include air speed, side-slip, angle-of-attack, static air pressure, or dynamic air pressure, for example. Preferably, the sensing is performed by deflection of a surface that supports separated capacitive plates arranged to oppose other stationary plates. The capacitance of each plate pair depends upon the spacing between the plates. Thus, the sensing can be performed to indicate the magnitude and direction of the air pressure exerted against the surface. Also, the sensing can be performed conformally with the aircraft's exterior surface so that the air flow over the aircraft surface is not significantly disturbed. In addition, the mapping of the level of the sensed air pressure to the corresponding air data signal can be derived by a learning system such as a neural network trained either in flight or in wind tunnel testing, for example. The method can also include steps of controlling the aircraft or generating a display, based on the air data signal.

These together with other features and advantages, which will become subsequently apparent, reside in the details of construction and operation as more fully hereinafter described and claimed, reference being made to the accompanying drawings, forming a part hereof, wherein like numerals refer to like parts throughout the several views.

Brief Description of the Drawings

Fig. 1 is a general block diagram of an apparatus in accordance with this invention;

Fig. 2 is a block diagram of a preferred embodiment of the invented apparatus;

Figs. 3A and 3B are a cross-sectional side view and a graph of upper and lower airfoil surface pressure versus location on the wing surface;

Figs. 4A and 4B are views of sensors located on the wing of an aircraft;

Figs. 5 A and 5B are views of a sensor located on the tail of an aircraft;

Fig. 6 is a view of a sensor and attached conformal member in accordance with a first embodiment of the invention;

Figs. 7A and 7B are cross-sectional side views of different embodiments of the sensor;

Figs. 8A and 8B are cross-sectional side views of different embodiments of the sensor;

Fig. 9 is a bottom plan view of the sensor;

Fig. 10 is a circuit diagram of an oscillator for use in the invented apparatus;

Fig. 11 is a block diagram of a portion of the apparatus including sensors, oscillators, and time period measurement units;

Fig. 12 is a relatively detailed block diagram of the sensor, oscillator and time measurement unit;

Figs. 13A and 13B are flowcharts of processing performed by an interface unit of the invented apparatus;

Fig. 14 is a flow chart of processing performed by an associator of the invented apparatus; and

Figs. 15 A - 27B are timing diagrams of signals generated by the interface unit and the associator.

Best Mode for Carrying Out the Invention 1. General Description of the Invented Apparatus

In Fig. 1, an apparatus 1 of this invention includes at least one sensor 2. The sensor(s) 2 are subscripted with numerals to distinguish different sensors. This subscripting is also used with other elements yet to be described in detail. The sensor 2 is mounted in an aircraft to sense air flow pressure exerted against an external aircraft surface 3. The sensor is mounted in a location on the aircraft's surface that is strategically-positioned for the particular air data quantity that is desired to be sensed. The sensor should be positioned on the aircraft surface at a location that experiences the greatest variation of air pressure corresponding to the air condition that is desired to be sensed. For example, if the air data to be generated is air speed, the sensor should be mounted on a forward-facing surface of the aircraft with respect to air flow, such as the nose or leading edge of a wing. On the other hand, if the air data to be measured is side-slip, the sensor can be mounted on the aircraft's vertical tail stabilizer. For some air data quantities, such as air speed and side slip, a single sensor can suffice to characterize the air data quantity. However, other air data quantities such as angle-of-attack require two or more sensors. For angle-of-attack air data, two sensors are preferably mounted on the upper and lower surfaces of the wing toward the wing's leading edge so that they will experience the greatest change in pressure differential due to corresponding changes in the aircraft's angle-of-attack. In addition, two or more sensors may be used to measure the same air data quantity to provide multiple redundant air data for increased reliability. Therefore, depending upon the air data quantities to be sensed and multiple redundancy considerations, a plurality of sensors 2_j through 2_N may be required or desirable for the sensor system, and mounted at respective strategic locations, as shown in Fig. 1.

The apparatus can also include a conformal member(s) 4 coupled to one or more sensors of the aircraft. The conformal member extends into an opening formed in the wall defining the aircraft's exterior surface. The conformal member is preferably composed of a lightweight, substantially rigid material such as molded plastic or composite material, that is shaped to conform to the aircraft's exterior surface so that it is aerodynamically integral with the aircraft surface in which it is installed. As such, the sensor adds no radar profile to the aircraft and presents no disturbance of the air flow that could be detrimental to aircraft control, particularly at high speeds. The conformal member transmits the force exerted by air flow against the aircraft's surface to its sensor for generation of the sensor signal. To further seal the interior of the aircraft and its respective sensor from the environment, a deformable cover 5 can be mounted to the exterior surface of the aircraft. The cover 5 is attached preferably about its periphery to the aircraft's exterior wall by application of a durable adhesive, for example. The cover is resilient at least in the portion thereof that covers the wall opening in which the sensor is situated. The cover thus seals the interior of the aircraft, yet deflects under air pressure to allow the air pressure force to be transmitted to the conformal member and hence to the sensor to which the conformal member is mounted.

The apparatus also includes an associator 6 that is coupled to receive the signal(s) generated by the sensor(s) 2. The associator generates an air data signal, based on the sensor signal(s). More specifically, the associator maps the level of the sensor signal(s) to a corresponding level of the air data signal desired to be generated by the apparatus. The associator can include a learning system 7 such as a neural network, that is trained with the mapping of the sensor signal to the air data signal. The learning system can be trained with a mapping derived by recording the levels of air data generated by one or more conventional air data sensors under different air flow conditions, in association with the levels of the sensor signal(s) generated by such conditions. After use in producing the mapping, the conventional air data sensors can be removed from the aircraft to eliminate their radar profile and/or undesirable interactions with air flow at high speeds. A suitable data acquisition system is the RADACS™ system commercially available from Accurate Automation™ Corporation of Chattanooga, Tennessee. Alternatively, or in addition, the mapping can be derived by subjecting the aircraft to predetermined known air flow conditions in a wind tunnel, for example, and recording the sensor signal generated by the conditions in association with air data determined from the known conditions.

Once a set of corresponding sensor signals and target air data signals has been collected, the set is supplied to the input and output sides of the associator to train the learning system. If the learning system is implemented as a neural network, the weights of the neural network are adjusted preferably through back propagation or other technique using training sets mapping the sensor signals to the air data signals. The preferred neural network configuration is the well-known multilayer perceptron, and one or two hidden layers are generally sufficient to implement the desired mapping. Although the learning system is preferred for use in generating the mapping of the levels of the sensor signal to corresponding levels of the air data signal because it can be readily updated with additional sensor and air data or can adapt to changes in aircraft performance, the learning system can be replaced with a wide variety of devices into which the mapping can be programmed after a set of corresponding signals is collected. For example, the learning system can be replaced with a look-up table, programmable logic array, microcontroller, fuzzy logic associator, or processor coupled to a memory, to implement the mapping for generation of air data signals based on sensor signals.

The air data signal generated by the associator can be used for a variety of purposes. For example, the associator can be coupled to supply the air data signal to a flight control system and/or a pilot display for use in controlling the aircraft.

To briefly summarize the operation of the apparatus of Fig. 1, the sensor(s) sense air pressure exerted against the conformal member (or the combined cover and conformal member), and generate a sensor signal(s) indicative of the sensed air pressure. The sensor(s) supplies its signal(s) to the associator, that maps the level of the sensor signal(s) to a corresponding level of the air data signal(s). The associator outputs the air data signal to the flight control system and/or pilot display for use in controlling the aircraft. 2. A Preferred Embodiment of the Invented Apparatus

Fig. 2 is a preferred, more specific embodiment of the invented apparatus 1. In Fig.

2, the apparatus 1 includes the elements shown in Fig. 1, and in addition can include serially-coupled oscillation unit(s) 8, time measurement unit(s) 9, and an interface unit

11, that are coupled between the sensors 2 and the associator 6. The sensor 2 can be one of a variety of different types of sensors, including capacitive or resistive sensors that detect the deflection of a surface, or a piezoelectric sensor (a piezoelectric sensor is not preferred for many applications, however, because it only detects changes in pressure as opposed to absolute pressure). In the preferred embodiment, the sensor 2 has a capacitance that varies in dependence upon the air pressure exerted against the sensor through the conformal member to which it is mounted. The sensor is coupled to an oscillation unit 8 that generates an oscillation signal whose frequency depends upon the sensor's capacitance. The oscillation unit is coupled to output the oscillator signal to a time measurement unit 9. Based on the oscillator signal, the time measurement unit generates a signal indicative of the oscillation signal's frequency. The interface unit preferably includes a microcontroller 12 coupled to receive the sensor signal. The microcontroller 12 can be a device such as a model no. 68F333 commercially available from Motorola®, Inc. of Schaumberg, Illinois. The microcontroller generates an air pressure signal, based on the frequency signal from the time measurement unit 9. The microcontroller, or more generally, the interface unit, is coupled to supply the air pressure signal to the associator 6. The associator is programmed with a mapping of the air pressure signal to the air data signal for the air condition to be sensed. Using the mapping, the associator generates an air data signal based on the air pressure signal.

The associator can include a neural network processor such as the NNP® commercially available from Accurate Automation™ Corporation of Chattanooga, Tennessee. The

NNP is preprogrammed to implement a neural network, preferably a multilayer perceptron with one or two hidden layers and sufficient input nodes to accommodate signals from all sensors included in the apparatus. The NNP uses the mapping programmed in its neural network to generate an air data signal based on the air pressure signal. The neural network processor, or more generally, the associator, is coupled to supply the air data signal to the interface unit 11. The interface unit is coupled to supply the air data signal to a flight control system (FCS) 15. Based at least in part on the air data signal, the flight control system generates an actuator signal. The

FCS is coupled to supply the actuator signal to the aircraft actuators 16 that control the aircraft's control surfaces such as wing flaps, ailerons, or elevators, as well as the aircraft's power plant. The actuators' operation affects the state of the aircraft 17 that is sensed with sensors 18 that can be gyroscopes, fuel flow meters or the like. The aircraft sensors are coupled to supply respective signals to the FCS for control of the aircraft. In addition, the signals generated by the sensors 18 can be supplied to a display 19 to indicate the aircraft state to the pilot 20. The apparatus 1 can also be coupled to supply the air data signal to the display 19. The display 19 generates a display based on the air data signal for use by the pilot in controlling the aircraft. The pilot operates control instruments 21 to generate a signal to control the aircraft. The control instruments are coupled to supply the control signal to the FCS.

3. Sensor Positioning and Assembly

As previously mentioned, the sensor 2 and its coupled conformal member 4 are preferably positioned to sense pressure at a location that experiences the greatest variation of the air data quantity desired to be sensed under a variety of air conditions. Fig. 3A shows a cross-section of an air foil in correspondence with a graph of upper and lower surface pressures at positions of along the surface of the airfoil at a zero angle-of-attack. As can be seen in Fig. 3B, the greatest variation in differential pressure between the upper and lower surface pressures occurs toward the leading edge of the air foil. Therefore, for example, if the air data quantity that is desired to be measured is angle-of-attack, it is advantageous to position a sensor on the upper surface of the air foil toward its leading edge in region Ω that experiences pressure p_u and another sensor on the lower surface of the airfoil toward its leading edge in region Φ that experiences pressure ._/.

Figs. 4 A and 4B show sensors located on the wing of the aircraft to measure angle-of-attack. In addition, these sensors can each be used to generate a variety of other air data, including air speed and static and dynamic air pressure.

Figs. 5 A and 5B are views of a sensor positioned on the vertical tail of an aircraft to measure side slip. As shown in Figs. 5A and 5B, the sensor is preferably placed toward and parallel to the leading edge of the vertical tail. The sensor of Figs. 5A and 5B can also be used to generate other air data, including air speed and static and dynamic pressure.

The measurement accuracy required for the sensor depends upon the air conditions expected to be encountered in the aircraft in which the sensor is installed. For example, for general aviation (GA) aircraft, static and dynamic pressure sensors must be operable between sea level and 19,000 feet, the limits of operation of a typical GA aircraft. For altitude measurement, static pressure must be measurable within this range. This range corresponds to static pressures of 2,116 pounds per square foot at sea level and 1,015 pounds per square foot at 19,000 feet (such pressures can be based upon average values during a standard day as defined by the ARDC Model Atmosphere and presented in General Electric (1960) Aerospace Propulsion Data Book GED-4503). Accordingly, the measurement precision for altitude air data must be less than about 0.6 pounds per square foot for accuracy to tens of feet [(2,116 lbs./sq.ft. - 1,015 lbs/sq. ft.)/19,000 ft. x 10 feet 0.6 lbs./ sq.ft.]. For airspeed measurement, assuming that the sensor is installed in a typical GA aircraft, a well-known equation which relates dynamic pressure to calibrated airspeed is q = v ² / 295.38. Differentiating this equation with respect to velocity yields 2v / 295.88. The sensor should work well at the lowest airspeed likely to be encountered by the aircraft, such as 30 knots. Accordingly, 2 x (30 knots)/295.88 = 0.203 lbs./sq. ft. To achieve an accuracy of 1.5 knots at 30 knots, the pressure accuracy of the sensor should be 0.203 x 1.5 knots = 0.305 lbs./sq. ft. For angle-of-attack air data, measurements must be taken in a range from p_min to p_max degrees with a quantization of Δp degrees. Using the graph of Fig. 3B for a typical airfoil, the maximum negative air pressure on the upper surface of the airfoil increase in magnitude to about -1.5 of the positive value on the lower airfoil surface. At a negative angle-of-attack, pressures would be positive on the upper surface and negative on the lower surface. At higher altitudes, the angle-of-attack of the airfoil must be larger to create the required negative air pressures to maintain the aircraft aloft. Assuming that the sensor is installed in a typical GA aircraft, the angle-of-attack can vary from -30 to +30 degrees. Thus, the air pressure will vary from -1.5 P_max to P_max. At a maximum airspeed of 200 knots, P_max corresponds to 135 pounds per square foot. The pressure differential between the upper and lower surfaces of the airfoil will be least at landing and takeoff of about 50 knots corresponding to a dynamic pressure of about 8.4 pounds per square foot (see the aforementioned General Electric Aerospace Data Book). If angle-of-attack is desired to be measured within Δp of 0.5 degrees or less, the required measurement accuracy is 10.5 pounds per square foot or less [8.4 lbs./sq.ft. - (-1.5 x 8.4 lbs./sq.ft.)] x 0.5 = 10.5 lbs./ sq.ft.]. Similar calculations can be performed to determine the sensor accuracy for high-performance aircraft. In the preferred embodiment, however, the range of measurements required of the sensor to measure dynamic air data (that is, air data that change by virtue of the aircraft's motion or attitude, such as airspeed, side slip or angle-of-attack) is reduced by a port that equalizes static pressures on both sides of the sensor's pressure-sensing surface which will be described in detail subsequently. The sensor thus does not have to cover the entire range of pressures expected over the airfoil for such dynamic air data.

Referring to Fig. 6 and 7A, top and cross sectional views, respectively, are shown of a force sensing touchpad-type sensor 2 used in the preferred embodiment of the present invention. For simplicity, only the sensor 2, and its associated elements will be described herein, where the subscript "1" is used to indicate one particular sensor among a group of sensors 2. The other sensors (if used) and their associated elements are substantially similar to the sensor 2ι . The cross-sectional view of Fig. 7A is taken through lines 7-7 of Fig. 6. The sensor 2 is mounted to a support structure 30 positioned interiorly of the aircraft's external surface 3. More specifically, the sensor 2 includes a frame member 31 that is mounted to the structure 30 with a durable adhesive, for example. Alternatively, the frame member 31 can be bolted to the aircraft's exterior surface with screws 32. The frame member 31 can be formed from a rigid material such as a metal stamping. A top planar member 33, which is formed from a material such as FR4 printed circuit board material, or stamped metal, for example, is combined with frame member 31 to form a suspension system by providing a plurality of spring members.

According to a presently preferred and illustrative embodiment of the invention, the top planar member 33 is formed in a rectangular shape, although other shapes will readily suggest themselves to those of ordinary skill in the art. Those of ordinary skill in the art will understand that the geometric forms taken by top planar member 33 in the illustrative embodiment of the invention disclosed herein may vary dramatically in concert with the specific implementation of the sensor 2, including the shape of the aircraft's external surface configuration, the position of strength members and other devices, wiring or fuel tanks and lines inside the aircraft's external surface.

As presently preferred, the suspension system of the present invention is formed in a single step simply by creating a plurality of slots 34 in top planar member 33. In this embodiment, slots 34 divide the single piece of material characterized by top planar member 33 into an inner area pressure-sensing surface 35, a plurality of spring portions 36, and an outer mounting ring 37. As presently preferred, the slots 34 are positioned to ideally exhibit four-fold symmetry, so as to reduce any non-idealities resulting from the mechanical design. The outer mounting ring 37 defined by this single step may easily be affixed to the periphery of reference frame member 31 using known methods (such as an adhesive) to form the suspension system of the present invention.

According to the preferred embodiment illustrated herein, the horizontal slots 34 are the same length as the vertical slots 34 on the left and right of the top planar member 33, so as to ensure that all springs 36 are the same length and thus ideally matched. The material from which top planar member 33 is formed should be substantially rigid, so as to linearly transmit an applied force from pressure-sensing surface 35 to the spring portions 36 without substantial deformation of pressure-sensing surface 35 which would degrade the precision of the measurement made herein, especially for measurement of small forces.

The conformal member 4 is mounted to the pressure sensing surface 35 with a durable adhesive, for example. The conformal member has a conformal surface that either constitutes a portion of the exterior surface 3 of the aircraft, or supports the resilient cover 5 that has edges adhered or otherwise attached to the support structure 30. The conformal member extends through an opening in the exterior surface 3 and its support structure 30 positioned interiorly of the aircraft's external surface. For example, the conformal member can be composed of a lightweight, durable material such as molded plastic or composite material. Under force of pressure exerted by air flow, the conformal member transmits the air pressure force to the surface 35 to cause it to deflect due to the presence of springs 36. The sensor 2 operates by sensing the physical separation between opposing points on the top planar member 33 and frame member 31 as the result of applied force.

The cross-sectional view of Fig. 7A shows the contact alignment between outer mounting ring 37 of top planar member 33 and frame member 31. According to a presently preferred embodiment of the invention, a plurality of capacitors 41 are disposed at the periphery of the pressure-sensing surface 35. The top plate of each capacitor, denoted by reference numeral 41a, is disposed on the bottom of the pressure-sensing surface 35, while the other plate, denoted by reference numeral 41b, is disposed on or integral with the frame member 31.

Plates 41a and 41b of each capacitor are separated by an air gap dielectric 42. As presently preferred, the distance between opposing plates 41a and 41b is between about 5 mils and about 20 mils with no force exerted on pressure-sensing surface 35. According to a design tradeoff, a smaller capacitor plate separation allows the use of stiffer springs 36, and makes the sensor 2 less susceptible to external vibrations. However, manufacturing tolerances are more critical in designs utilizing smaller capacitor plate separations. A no-force capacitor separation range between about 5 mils and about 20 mils has been found to be acceptable from both sides of this tradeoff. Those of ordinary skill in the art will appreciate that other embodiments of the invention are contemplated by merely varying spring stiffness and the dimensions of air gap dielectric 42.

According to a presently preferred embodiment of the invention, the frame member 31 is formed from conductive material such as stamped metal and is electrically grounded. In such an embodiment, capacitor plate 41b is integral with frame member 31. When air flow pressure is exerted against the conformal member that presses down on pressure-sensing surface 35, the capacitor plates 41a and 41b of all capacitors 41 are brought closer together, thus increasing the capacitance of the capacitors 41. If air pressure is exerted uniformly against the conformal member and hence against the surface 35, the distances between the opposing plates of each capacitor 41 will decrease by the same amount, thus producing theoretically identical capacitance changes for all capacitors 41. If, however, the air pressure is non-uniformly against the conformal member and thus the surface 35, the distances between the opposing plates of individual capacitors 41 will each decrease by different amounts depending on the air pressure distribution across the conformal surface of the member 4, and the capacitance changes of the capacitors will generally be different.

If top planar member 33 is made of printed circuit board material, capacitor plates 41a can be formed from copper traces on the lower surface thereof. In addition, other circuit elements 45, such as the sensing circuitry for the present invention to be disclosed herein with reference to Figs. 10 and 11 (which can include the oscillation unit 8 and the time measurement unit 9), can be placed on the bottom face of the top planar member 33. This arrangement permits fabrication of a compact, integrated air pressure sensor. Those of ordinary skill in the art will observe that, while reference numeral 45 illustratively designates a single structure on the bottom of top planar member 33, multiple independent structures, such as one or more integrated circuits and/or discrete components may be employed to comprise circuitry 45.

The immediate vicinity of capacitor plates 41 is grounded in order to shield the capacitor plates 41 from external noise. According to the presently preferred embodiment of the invention, the entire frame member 31 is a grounded metal object. As will be apparent to those of ordinary skill in the art, the frame member 31 preferably extends inward beyond the capacitor plates 41b as shown in Fig. 7A in order to provide a ground plane to shield the capacitor plates 41b from noise emanating from below the center of the pressure-sensing surface 35. In addition, the top surface of top planar member 33 is also grounded, in order to shield capacitor plates 41 from noise emanating from above the pressure-sensing surface 35. The top surface of the top planar member 33 is then the ground plane for the entire sensor, supplying ground to circuitry 45 and to the frame member 31.

As an alternative embodiment, the circuitry 45 may be placed on the top surface of frame member 31. Frame 31 would then be made from printed circuit board material and would not be grounded. The bottom surface of the top planar member 33 would be grounded instead. This alternative embodiment is less preferred, because the connection from circuitry 45 to capacitor plate 41b would result in a higher manufacturing cost.

Alternatively, as shown in Fig. 7B, a piece of circuit board material 46 can be adhered to the bottom of pressure-sensing surface 35, which allows for the use materials such as spring steel or phosphor bronze for the top planar member 33. Those of ordinary skill in the art will appreciate that such materials have better spring properties (such as creep) than standard printed circuit board epoxy/glass composites, although they make the sensor more complex to assemble. A cross-sectional view of an alternate embodiment of the sensor 2 is shown in Fig. 8 A. The embodiment of Fig. 8A is similar to that shown in Figs. 7A and 7B except that the top of the force sensing touchpad is completely sealed by a thin protective layer 49. This protective layer 49 may be formed, for example, from a sheet of neoprene having a thickness in the neighborhood of about 2 mils or from similar materials. The embodiment of Fig. 8A will be immune to dirt, but the protective layer may affect the accuracy of the touch sensor, due to slight torques exerted by layer 49 on pressure-sensing surface 35. Those of ordinary skill in the art will appreciate that the embodiment of Fig. 8A is environmentally sealed and protected from intrusion of dirt and dust.

Referring now to Fig. 8B, an alternate embodiment like that of Fig. 8A is depicted in cross sectional view. The embodiment of Fig. 8B differs from that depicted in Fig. 8A in that the neoprene layer 49 is disposed under the pressure-sensing surface 35.

Referring now to Fig. 9, a bottom view of the top planar member 33 is shown. According to a presently preferred embodiment of the invention, the capacitor plates 41 mounted on or integral with top planar member 33 have the same aspect ratio as the pressure-sensing surface 35, in order to cancel non-linear response caused by the tilting of the capacitors across the non-zero size of the capacitors. Thus, as shown in Fig. 9, the faces of capacitor plates 41a have lengths and widths selected to produce the same aspect ratio of the length and width of pressure-sensing surface 35.

In an actual embodiment of a sensor 2 of the preferred embodiment of the present invention, capacitor plates 41a are formed from copper pads on the printed circuit board material, while capacitor plates 41b are integral to frame 31, which is a grounded, metallic object. In the presently preferred embodiment, capacitor plates 41a are rectangles formed to dimensions of 690 mils by 460 mils (to match the aspect ratio of pressure-sensing surface 35). With a nominal air gap dielectric thickness of 11 mils, the capacitors 41 each have a no-force capacitance of about 20 pF.

As may also be seen from an examination of Fig. 9, circuitry 45 mounted on top planar member 33 is coupled to the capacitor pads 41a and with other circuitry located other than on top planar member 33 via conductors 50. Conductors 50 are traces formed on the material used to form top planar member 33 provide a convenient way to couple electronic signals from circuitry mounted on the bottom of pressure-sensing surface 35 without affecting the force measurement. As is apparent to those skilled in the art, the ends of conductors 50 which are used for off-sensor connection must not be covered by reference frame member 31 , in order to allow connections to an external system such as the interface unit 11. Conductors 50 which are used for off-sensor connection can terminate on a tab 51 of material that extends beyond the main rectangular core of top planar surface 33. While five conductors 50 are shown for off-sensor connections in the drawing figure, persons skilled in the art will appreciate that the number of conductors employed in the sensor are contemplated and will depend on the particular design of the apparatus 1. In the preferred embodiment, the five conductors supply power (+5 Volts, approximately 50 milliamperes), ground, and run, strobe and data signals whose purpose will be described later in this document. The ground connection to the outside world can also be placed amongst the connectors on tab 51. This ground connection will connect to the top surface of the top planar member 33, to provide ground for the entire sensor.

In an embodiment of the present invention wherein top planar member 33 is formed from printed circuit material, conductors 50 may advantageously be formed as conductive traces on the printed circuit material, as are capacitor plates 41a. It will be understood that other interconnection methods such as providing wires to carry signals to and from circuitry 45 are possible. Such other interconnect methods can introduce a parasitic mechanical coupling between top planar member 33 and frame member 31 and thus potentially degrade the accuracy of the system, especially one tailored to detection of small forces. It is apparent to those skilled in the art that any wires that connect to circuitry 45 that are not traces on top planar member 33 should be made a thin as possible to minimize the degradation of sensor accuracy. Providing trace conductors 50 used for off-sensor connections which communicate with the fixed outer mounting ring 37 of top planar member 33 eliminates any inaccuracy caused by such signal wires. The alternative embodiments shown in Figs. 7B and 8B do not permit the use of off-sensor connection trace conductors 50.

It is important to note that for some dynamic air data such as air speed, side slip and angle-of-attack, the frame 31 or pressure sensing surface 34 can be vented with a port to provide the static air pressure experienced by the aircraft in proximity to the sensor to the area enclosed by the sensing surface 34 and the frame 31. The static pressure port reduces the range required of the sensor to measure a dynamic air data quantity. In the embodiments of Figs. 7A and 7B, such port can be coupled in communication with the channel formed between the conformal member 4 and the structure 30 that communicates through openings 34 with the enclosed area. Alternatively, such port can be formed in the frame 31 and coupled to communicate with the external static air pressure through tubing or a channel formed in the structure 30 and the external surface 3 to direct the static air pressure external to the aircraft to the enclosed area.

Many different devices and methods for measuring capacitance are well-known in the art. For example, a circuit can be employed to measure the AC coupling between the plates, or alter the charge on the capacitor and measuring the resulting voltage change. Other capacitance measuring circuits can suggest themselves to those skilled in the art. Fig. 10 is an example of one such device and method for measuring capacitance of the plates 41.

In Fig. 10, according to the presently preferred embodiment of the invention, oscillators 8_1A, 8_1B, 8_1C, 8ι_D are connected to respective pairs of capacitor plates 41a and 41b. Plate 41b is held at ground. Each oscillator shares the common ground with capacitor plates 41b and senses the voltage on plate 41a to produce a digital square wave on output node 53 which alternately charges and discharges plate 41a through resistor 52. The oscillator circuit 8, may be configured from the well-known "555" oscillator integrated circuit, the essential features of which are shown for illustration purposes. Some of the internal circuitry of a 555 oscillator is shown in Fig. 10 for reference purposes.

The frequency of the square wave on output node 53 is inversely proportional to the capacitance between plates 41a and 41b. According to a presently preferred embodiment of the invention, the value of resistor 52 is chosen so that the frequency of oscillation of the oscillator of Fig. 10 is approximately 100 kHz, although the circuit can function over a very wide range of frequencies. Setting the frequency entails a tradeoff between precision and power consumption. As the frequency is lowered, the amount of power consumption goes down linearly with frequency. However, phase noise in the oscillator will gradually lower the precision of the touchpad as the frequency gets lower. It has been found that 100 kHz is a good tradeoff between precision and power consumption.

The capacitor 54 is a filtering capacitor which is provided to reject power supply noise below a certain frequency. The value of capacitor 54 multiplied by the value of resistor 55 (internal to the 555 integrated circuit) should be chosen to be smaller than typical sampling times used in the system, such as 12.5 milliseconds. In the presently preferred embodiment, the value of capacitor 54 multiplied by the value of resistor 55 is chosen to be 10 milliseconds.

As will be apparent to those skilled in the art, other oscillator circuits can be used instead of the oscillators 8_1A, 8_1B, 8_1C, 8_1D of Fig. 10, such as a Schmitt trigger. The 555 integrated circuit oscillator has been chosen to minimize power supply sensitivity. Furthermore, it will be apparent to those skilled in the art that a mixed-signal VLSI ASIC integrated circuit which uses an oscillator circuit employing current sources instead of resistors can be used in the present invention.

The period of oscillation of the oscillator 8_1A, 8_1B, 8_1C, 8_1D is proportional to the capacitance of the capacitor 41 according to the equation:

T=KC where T is the period, C is the capacitance, and K is a constant depending on the circuit used. The capacitance C is related to the distance between capacitor plates 41a and 41b by the equation:

C = C '_n0 + ^ d where C₀ is a background capacitance, C, is a proportional capacitance, and d is the distance between the plates. Combining these equations results in:

(T - a„) where ■ and b are constants.

The springs 36 fabricated from FR4 are very close to linear, which means that the force F exerted on the comer obeys the equation:

F = s(d -d₀) where s is the strength of the spring and d₀ is the distance between capacitor plates when no force is applied to the pressure-sensing surface 35.

After algebraic manipulation, the force on a corner can be related to the period through the equation: a, (T - T₀)

F =

(T - a₀) where a₀ is defined above, a_] is a constant, and T₀ is the period of the oscillator when no force is applied to the pressure-sensing surface 35. Therefore, to determine the force-based on the oscillator period, three constants (T₀, a₀, and a_{) must be determined in the calibration of the sensor.

The constant T₀ is determined by measuring the period of the oscillator relative to some known air data condition applied to the pressure-sensing surface 35. For example, the constant T₀ can be determined from air data conditions known to exist when the aircraft is on the ground and stationary.

The constants a₀ and a_r are measured at time of manufacture, when a series of known forces are applied to the exact center of the pad and the periods of the oscillators are measured. The constants a₀ and a_! can be computed from a least squares fit, as is well-known in the art, by minimizing the function:

∑[F_i (T_i - a₀) - a_i (T_i -T₀)]² ι=0 with respect to SL_Q and a_{1 ?} where Fj is the ith known force applied to the corner (which is typically the total force divided by 4), and Tj is the ith measured period of the same corner. The parameter a₀ = [(B - 1)/B] T₀, and a₀ = A/B.

Referring now to Fig. 11, a block diagram shows the overall architecture of the data processing of the sensor of the present invention. Each capacitor 41 is connected to a respective oscillator 8_1A, 8_1B, 8_1C, 8_1D (in addition to numeral subscripts distinguishing the oscillators of different sensors, the capital letter subscript indicates a particular capacitor 41 among the four included in a single sensor, and similar subscripts are used with other elements hereinafter), which sends a square wave on output node 53 to the period determination units 9_!A, 9_1B, 9_1C, 9_1D (or equivalently, time measurement units). The frequency and/or period of the square wave on output node 53 can be very accurately measured using digital circuits that are well-known in the art. Many such circuits are known and will yield equivalent accuracies. For illustrative examples of such a circuit, see United States Patent No.4,027,146 to Gilmore, United States Patent No.4,052,620 to Brunneff, or United States Patent No. 4,350,950 to Waldmann, et al. Once the period or frequency has been determined for the oscillator driven by the capacitors 41 in every corner, the period determination unit 9 outputs the frequency and reference signal pair for each capacitor 41 to the interface unit 11. The unit 91 therefore outputs eight signals used by the interface unit 11 to calculate air pressure for each capacitor based on respective reference and frequency signal pairs. In a preferred implementation, to initiate counting of the frequency or period of the oscillators 8_1A, 8_]B, 8_1C, 8_1D, the units 9_1A, 9_1B, 9_1C, 9_1D are coupled to receive a RUN signal generated by the interface unit 11. Also, the interface unit 11 is coupled to supply a STROBE signal to the units 9_1A, 9_1B, 9_1C, 9_1D serially read the frequency and reference signal pairs from the units 9_1A, 9_1B, 9_1C, 9_1D. In this preferred implementation, the units 9_1A, 9_1B, 9_1C, 9_1D are serially coupled together as shown in Fig. 11 to output the frequency and reference signals to the interface unit 11 as serial data.

Fig. 12 is a relatively detailed block diagram of a capacitor 41, the oscillator 8_1A and the time measurement unit 9_1A in a preferred embodiment of the invention. Although only the oscillator 8_1A, its associated capacitor 41, and the time measurement unit 9_1A are shown in Fig. 12, the structure and function of the other capacitors 41, oscillators 8_1B, 8_1C, 8_1D and period determination units 9_1B, 9_1C, 9_1D is similar to that described in Fig. 11.

In Fig. 12, the unit 9_1A is coupled to receive the RUN and STROBE signals from the interface unit 11. The unit 9_1A includes a SIG1 counter 91, a REF1 counter 92 and a clock 93. The counters 91, 92 are coupled to receive the RUN and STROBE signals from the interface unit 11. Upon activation of the RUN signal, the counters 91, 92 are enabled to count upon activation of the signals applied to the INC ("increment") terminals of the counters. The capacitor 41 between ground and the input terminal of the oscillator 8_1A whose output frequency (or period) depends upon the spacing of the capacitor's plates. The oscillator's output signal is coupled to the INC terminal of the SIG1 counter 91. When enabled by the RUN signal, the SIG1 counter is incremented as the oscillator's signal is activated to count the number of periods of the oscillation signal occurring during the current activation of the RUN signal. The reference clock 93 generates pulses, and is coupled to output its pulses to the INC terminal of the REF1 counter 92. When enabled by the RUN signal, the counter 92 is incremented by the pulses ofthe clock 93. Upon deactivation of the RUN signal, the counters 91, 92 are disabled from further incrementing. Typically, it is desirable to allow the counters to continue counting to the end of the final period of the oscillation signal that was being counted when the RUN signal was disabled. Although such circuit is not shown in Fig. 12, many circuits that will accomplish this objective are known in the art, such as the circuit disclosed in United States Patent No.4,027,146 to Gilmore. The interface unit 11 is programmed to wait a predetermined time interval after deactivation of the RUN signal to allow time for the counters of the units 9_1A, 9_1B, 9_1C, 9_1D to finish counting the final periods of respective oscillation signals. After elapse of this time interval, the interface unit 11 activates the strobe signal to serially strobe the data out ofthe counters of the units 9_1A, 9_1B, 9_1C, 9_1D. The interface unit 11 is coupled to receive the REF/SIG signals as they are serially-clocked out ofthe counters. Thus, the interface unit receives in order the REF1 and SIG1 signals of the unit 9_1A, the REF2 and SIG2 signals of the unit 9_1B, the REF3 and SIG3 signals of the unit 9_1C, and the REF4 and SIG4 signals from the unit 9_1D. In the preferred embodiment described herein, the REF signals are 20 bits in length and the SIG signals are 16 bits in length. Accordingly, the interface unit 11 activates the STROBE signal 144 times to strobe all REF and SIG signals out of the units 9_1A, 9_1B, 9_1C, 9_1D.

4. The Interface Unit As previously mentioned, the interface unit 11 can include a programmable device such as the microcontroller 12. The interface unit 11 is programmed to calculate air pressure signals from the REF and SIG signals generated by unit 9, and to handle transfer of signals between the associator 6 and the unit 11, the FCS 15 and/or the display 19. The microcontroller can be a device such as the model no. 68F333™ manufactured by Motorola®, Inc. Figs. 13A and 13B are flow charts of processing performed by the interface unit 11, or more specifically, the microcontroller 12 to perform these functions. The processing begins in step SI of Fig. 13 A. In step S2, the interface unit activates and is coupled to supply the RUN signal to the units 9. In step S3, the unit 11 determines whether a predetermined time period (for example, of ten milliseconds) has expired. If not, the unit 11 waits for a predetermined time period in step S4 and again performs the determination of step S3. On the other hand, if the determination in step S3 is affirmative, the unit 11 deactivates the RUN signal in step S5 and waits for a predetermined time period (for example, one-hundred microseconds) in step S6 to allow the units 9 to finish counts for the period existing when the RUN signal is deactivated. In step S7, the unit 1 1 initializes a variable i to one, and in step S8 generates the STROBE signal. The unit 11 is coupled to supply the STROBE signal to the time period determination unit 9. In step S9, the interface unit reads the reference and data signals for the ith sensor. In the preferred embodiment, these signals are reference signals REF_U, REF_i2, REF_l3, REF_i4 and data signals SIG,_b SIG_l2, SIG_i3, SIG,₄. In step S10, the unit 11 determines whether i = N where N is preprogrammed into the unit 1 1 , and is equal to the total number of sensors 2. If not, processing proceeds to step SI 1 in which the unit 11 increments the variable i by one and again performs steps S9 and S10. On the other hand, if the determination in step S10 is affirmative, processing proceeds to step S12 in which the unit 11 deactivates the STROBE signal. In step S13 of Fig. 13B, the unit 1 1 initializes the variable i = 1. In step S14, the unit 11 calculates air pressure signals for the ith sensor. More specifically, in the preferred embodiment, the unit 1 1 calculates pressures PRES_ix for each of the capacitors of the ith sensor as follows using similar, although differently expressed, computations as those set forth in the equations described with respect to Fig. 10. For each pair of reference and data signals, a raw data signal RAW_ix is computed for each sensor based on the following relation:

RAW_1X = ^ REF_1X where x = 1, 2, 3 or 4, depending upon the specific capacitor 41 of the ith sensor for which the calculation is performed. The unit 11 is programmed to generate a signal

DEL_1X with the following relation:

BGND

DEL = ²— 1

RAW_ix

The signal BGND_ix is determined in the calibration of the sensors and stored in the unit

11. More specifically, the signal BGND_ix is determined by observing the output level of the ith sensor's xth capacitor with no pressure applied to the sensor, and storing the observed output level in the interface unit's or microcontroller's memory. The unit 11 calculates the air pressure experienced by the sensor at the location of each capacitor 41 using the relation:

A * DEL

PRES_1X = ^{2 lx}

(1 + B_1X * DEL_1X ) where A_ix and B_ix are constants determined in the calibration of the sensor(s) by a least squares fit to data obtained when a known set of pressures are applied to the sensor.

The unit 11 thus calculates the air pressure signals PRES_ix. In step S15 of Fig. 13B, the unit 11 determines whether the variable i = N. If not, processing proceeds to step S16 in which the variable i is incremented by one, and then processing proceeds to step SI 4. On the other hand, if in step S15 the determination is affirmative, processing proceeds to step S17 in which the associator writes the air pressure signals PRES_ix to the associator 6. In step SI 8, the interface unit generates and is coupled to supply a start command signal STARTCOM to the associator. In step SI 9, the interface unit determines whether the associator has activated a data ready signal DATARDY. If not, in step S20, the unit 11 waits for a predetermined time period, and again performs the determination of step SI 9. On the other hand, if the determination in step S19 indicates that the data ready signal has been activated by the associator, in step S21, the unit 11 generates and writes a send data signal SENDDATA to the associator. In step S22, the unit 1 1 reads the air data signal AD from the associator, and in step S23, the interface unit outputs the air data signal from the apparatus 1. For example, the interface unit can be coupled to output the air data signal to the FCS 15 and/or the display 19. In step S24, the processing performed by the interface unit terminates. Of course, the interface unit performs the processing of Figs. 13A and 13B repeatedly during the aircraft operation, preferably on the order of about one hundred times per second or more, to be responsive to changing air conditions.

In Fig. 14, processing performed by the associator 6, or more specifically, the neural network processor 7, begins in step SI. In step S2, the associator 6 receives the air pressure signals PRES_ix from the interface unit 11. In step S3, the associator determines whether the interface unit has activated the start command signal. If not, in step S4, the associator waits for a predetermined time interval, and repeats the determination of step S3. If; on the other hand, the determination in step S3 is affirmative, processing proceeds to step S5 in which the associator calculates weighted summations and multiplies the weighted summations with respective gain factors from the input layer, through the hidden layers, to the output node(s) of the neural network implemented in the processor 7, to generate the air data signal. The specific neural network implemented in the neural network processor, preferred to be a multilayer perceptron with one or two hidden layers, is trained in advance of the processing of Fig. 17 to generate the air data signal based upon a mapping of the air pressure signal PRES_ix to the air data signal for the air data quantity that is desired to be measured. As previously mentioned, the mapping can be derived by compiling training sets that relate the air pressure signal generated by the interface unit to air data derived from conventional sensors during flight of the aircraft, or by relating air pressure signal levels derived by the interface unit under controlled conditions in a wind tunnel, to the air data signal level that is known to exist in the wind tunnel. The training sets are used to train the neural network implemented in the processor 7 to accomplish the desired mapping of the air pressure signal to the air data signal. Many such training techniques are well-known in industry. For example, back propagation can be used to train the neural network implemented in the processor 7. The specific mapping depends upon a variety of factors, including the characteristics of the aircraft in which the apparatus is installed, which are unique to each aircraft, the position of the sensors on the aircraft, the specific sensor and conformal member configuration used to generate the air data signal, and the air data signal desired to be generated. It should be apparent to those of ordinary skill in this technology that the derivation of a function that includes all of the above variables that can affect the relationship between a sensor's signal and a desired air data signal, would be highly difficult to determine. However, the power of a learning system such as a neural network to readily generate a mapping of a known input signal level to a known output signal level without requiring specific knowledge of the variables and processes which affect such mapping, are used to great advantage in the subject invention.

If the desired air data quantity is angle-of-attack, the air pressure signals calculated for at least two sensors, preferably located at respective upper and lower sides of a lift surface (such as a wing) of the aircraft toward its leading edge, are mapped to the air data signal. Therefore, at least eight air pressure signals PRES_ix, four for each sensor, would be required to be mapped to a single air data signal. Accordingly, the neural network would in this case have eight input nodes and a single output node. On the other hand, if the desired air data quantity to be generated is air speed, static or dynamic pressure, for example, a single sensor can be used to generate the air data signal. In this case, the neural network implemented in the processor 7 would have four input nodes for the four air pressure signals PRES_ix generated by the interface unit 11, and a single output node for the air speed air data signal. The processor 7 thus generates the air data signal based on the air pressure signal from the interface unit 11.

In step S6, the processor 7 writes the air data signal in a predetermined memory location within the processor that is accessible to the interface unit. In step S7, the associator generates a data ready signal that is supplied to the interface unit. In step S8, the processing of Fig. 17 terminates. The associator will perform the processing of Fig. 14 repeatedly during operation of the aircraft, possibly on the order of about one hundred times per second or more, so that the air data signal is responsive to changing air conditions.

Figs. 15A through 27B are timing diagrams of signals generated by the time 30 measurement unit, the interface unit and the associator in the operation of the invented apparatus. In Fig. 15 A, the interface unit 11 generates and supplies the signal RUN to the counters of the time measurement units 9. In response to the RUN signal, the counters of the time measurement units 9 are cleared. Also in response to the RUN signal, the time measurement units 9 initiate measurements of the frequency or period of the oscillation signals from the oscillators. In Fig. 15 A, after the elapse of a predetermined time period of 10 milliseconds, for example, the interface unit deactivates the RUN signal. In response to deactivation ofthe RUN signal, the counters stop incrementing the counts stored therein. In Fig. 16 A, the interface unit generates the strobe signal STROBE to synchronize the serial transfer of the signals REF1, SIG1, REF2, SIG2, REF3, SIG3, REF4, SIG4 from the period determination unit 9. Referring to Figs. 16A and 17A, the REF signals are 20 bits in length, and the SIG signals are 16 bits in length, so that the signal STROBE is activated 144 times to read the REF and SIG signals. The interface unit calculates four air pressure signals PRES1 - PRES4 corresponding to the pressure experienced by the four capacitors of the sensor. In Figs. 18 A and 19 A, the interface unit outputs the air pressure signals PRES1 - PRES4 with corresponding addresses ADRN1 - ADRN4 to write the air pressure signals into respective memory locations of the associator. In Fig. 20A, the interface unit activates and outputs the start command signal STARTCOM to the associator. The associator generates an air data signal AD based on the air pressure signals PRES 1 - PRES4 and stores the air data signal in a predetermined memory location accessible to the interface unit. The associator activates a data ready signal DATARDY in Fig. 21 A that is supplied to the interface unit to indicate that the air data signal has been generated. In response to the DATARDY signal, in Figs. 22B and 23B, the interface unit generates an address signal ADRI for the associator' s predetermined memory location containing the air data signal AD, and reads the air data signal into the interface unit's memory. In Figs. 24B and 25B, the FCS 15 activates and outputs the ADRINT signal to the interface unit 11 to interrupt its processing, and sends a SENDDATA signal to the interface unit requesting the interface unit to send the current air data signal AD. The FCS can be programmed to request air data periodically from the interface unit, or alternatively, the interface unit can send the air data signal to the FCS periodically without any command to do so by the FCS in which case the signals ADRINT and SENDDATA generated in Figs. 24B and 25B would not be necessary. In Figs. 26B and 27B, the interface unit addresses the FCS 15 with the signal ADRFCS, and outputs the air data signal ADO to the FCS. In addition, although not shown in the Figures, the interface unit can address and output the air data signal ADO to the display for visual indication ofthe air data to the pilot.

Many variations of the invented apparatus and method will occur to those of ordinary skill in the art, and such variations are intended to be included within the scope of the invention. For example, although the interface unit 11 and the associator 6 are shown as discrete units in Fig. 2, the unit 11 and associator 6 can be combined into a single unit such as a microcontroller or processor, to implement the functions of such elements.

The many features and advantages of the present invention are apparent from the detailed specification and thus, it is intended by the appended claims to cover all such features and advantages of the described apparatus and methods which follow in the true spirit and scope of the invention. Further, since numerous modifications and changes will readily occur to those of ordinary skill in the art, it is not desired to limit the invention to the exact construction and operation illustrated and described. Accordingly, all suitable modifications and equivalents may be resorted to as falling within the spirit and scope ofthe invention.

Industrial Applicability

The invented apparatus and method are applicable in the aircraft industry to sense air conditions encountered by an aircraft. The apparatus generates air data from the sensed air conditions such as airspeed, side slip, angle-of-attack, and static and dynamic air pressure, which can be used, for example, in aircraft control.

Claims

Claims

1. A method characterized by the steps of: a) sensing air pressure exerted against a portion of a surface of an aircraft; and b) mapping a level of the sensed air pressure to a corresponding air data level to generate an air data signal.

2. A method as claimed in claim 1, wherein said step (a) is performed by determining deflections at different locations of a sensor surface caused by the exerted air pressure.

3. A method as claimed in claim 1, wherein the sensing is performed conformally so that the air flow over the surface ofthe aircraft is not significantly disturbed.

4. A method as claimed in claim 1 , wherein the air data signal includes at least one of air speed, side slip, angle-of-attack, static air pressure, and dynamic air pressure.

5. A method as claimed in claim 1, wherein said step (b) is performed by a learning system.

6. A method as claimed in claim 5, wherein the learning system includes a neural network.

7. A method as claimed in claim 6, further characterized by the steps of: c) recording levels of sensed air pressure signals and corresponding air data levels determined by aircraft sensors during flight ofthe aircraft; and d) training the neural network to map levels of the sensed air pressure signal to corresponding levels ofthe air data signal, wherein said step (b) is performed with the neural network trained in said step (d).

8. A method as claimed in claim 6, further characterized by the steps of: c) recording levels of sensed air pressure signals and corresponding air data levels determined by wind tunnel test conditions; and d) training the neural network to map levels of the sensed air pressure signal to corresponding levels ofthe air data signal, wherein said step (b) is performed with the neural network trained in said step (d).

9. A method as claimed in claim 1, further characterized by the step of: c) controlling the aircraft, based on the air data signal.

10. A method as claimed in claim 1, further characterized by the step of: c) generating a display, based on the air data signal.

11. An apparatus for sensing air data in an aircraft, the apparatus characterized by: at least one sensor mounted to detect pressure exerted by air against an external surface of the aircraft, the sensor generating a signal indicative of the sensed air pressure; and an associator coupled to the sensor, the associator generating an air data signal based on the air pressure signal.

12. An apparatus as claimed in claim 11, wherein the air data signal includes at least one of air speed, side slip, angle-of-attack, static air pressure, and dynamic air pressure.

13. An apparatus as claimed in claim 11, wherein the sensor includes a deflectable surface whose deflection and tilt under the exerted air pressure are used to generate the sensor signal.

14. An apparatus as claimed in claim 13, wherein the sensor further includes first capacitive plates mounted on the deflectable surface, second capacitive plates arranged to oppose the respective first capacitor plates, the deflectable surface deflecting under the exerted air pressure so that the spacing of the first and second capacitive plates depends upon the exerted air pressure at the location of each first and second capacitive plate pair, the capacitive plates generating respective signals used to generate the signal output by the sensor.

15. An apparatus as claimed in claim 11, further characterized by: a conformal member coupled to the sensor, the conformal member arranged flush with the external surface ofthe aircraft.

16. An apparatus as claimed in claim 15, further characterized by: an elastic sheet mounted between the conformal member and the sensor, to prevent dust from moving around the conformal member behind a sensing surface ofthe sensor.

17. An apparatus as claimed in claim 11, wherein the conformal member includes a deformable cover mounted to the external surface of the aircraft; and a conformal member mounted on a first side to the cover and on a second side to the sensor.

18. An apparatus as claimed in claim 17, wherein at least one of the deformable cover and the conformal member is made of molded plastic material.

19. An apparatus as claimed in claim 17, wherein the deformable cover is mounted about its periphery to a support structure for the external surface of the aircraft.

20. An apparatus as claimed in claim 19, wherein a first sensor is mounted on a top side of a lift surface of the aircraft, and wherein a second sensor is mounted on a bottom side ofthe lift surface of the aircraft.

21. An apparatus as claimed in claim 20, wherein the air pressure signals generated by the first and second sensors are supplied to the associator to determine angle-of-attack.

22. An apparatus as claimed in claim 1, wherein the sensor includes a substantially rigid pressure-sensing surface; a substantially rigid frame member; a plurality of spring means formed integrally with the pressure-sensing surface; the spring means mechanically connected to the frame member; a plurality of variable capacitors associated with the pressure-sensing surface, each variable capacitor having a capacitance that is a monotonic function ofthe distance between predetermined portions of the pressure-sensing surface and the frame member, each capacitor generating sensed output signals; a plurality of oscillator circuits coupled to respective capacitors, said oscillator circuits generating respective output signals with frequencies dependent upon the capacitance of respective capacitors; and at least one time period determination unit coupled to the oscillator circuits, for measuring output frequencies of the signals generated by the oscillator circuits, and for generating the sensed air pressure signal, based on the output signals from the oscillators.

23. An apparatus as claimed in claim 22, wherein the sensed air pressure signal generated by the sensor includes a plurality of signals indicative of the deflection of the pressure-sensing surface at respective locations of the capacitors due to the air pressure exerted against said pressure-sensing surface.

24. An apparatus as claimed in claim 23, further characterized by: a conformal member coupled to the pressure-sensing surface, the conformal member mounted substantially flush with the exterior surface ofthe aircraft.

25. An apparatus as claimed in claim 24, further characterized by: a deformable cover mounted between the conformal member and the pressure sensing surface.

26. An apparatus as claimed in claim 22, further characterized by: a deformable cover mounted substantially flush with the exterior surface ofthe aircraft; and a conformal member coupled between the surface deformable cover and the pressure-sensing surface.

27. An apparatus as claimed in claim 11, wherein the associator includes a learning system coupled to receive the air pressure signal from the sensor, and generating the air data signal based on the sensed pressure signal.

28. An apparatus as claimed in claim 27, wherein the learning system includes a neural network.

29. An apparatus as claimed in claim 28, wherein the neural network is trained with a mapping derived by compiling air data for predetermined air data conditions, and the corresponding sensed air pressure signals for respective air data conditions.

30. An apparatus as claimed in claim 29, wherein the air data used for training the neural network is derived from additional sensors ofthe aircraft.

31. An apparatus as claimed in claim 29, wherein the air data is derived from wind tunnel testing ofthe aircraft.

32. An apparatus as claimed in claim 29, wherein the neural network is trained using back propagation with the air data and corresponding sensed air pressure signals.

33. An apparatus as claimed in claim 28, wherein the neural network includes a multilayer perceptron having at least one hidden layer.

34. An apparatus as claimed in claim 11, wherein the associator includes a look-up table programmed to generate the air data signal by mapping the air pressure signal to the air data signal.

35. An apparatus as claimed in claim 11, wherein the associator includes a neural network processor, the apparatus further characterized by: an interface unit coupled between the sensor and the associator, the interface unit receiving the sensed air pressure signal, and supplying the sensed pressure signal as an input to a neural network implemented in the neural network processor, and receiving the air data signal generated by the neural network processor based on the supplied air pressure signal, and the interface unit outputting the received air data signal.

36. An apparatus as claimed in claim 35, wherein the interface unit is coupled to output the air data signal to a flight control system of the aircraft.

37. An apparatus as claimed in claim 35, wherein the interface unit is coupled to output the air data signal to a display ofthe aircraft.

38. An apparatus as claimed in claim 11, wherein the associator is coupled to output the air data signal to a flight control system ofthe aircraft.

39. An apparatus as claimed in claim 11, wherein the associator is coupled to output the air data signal to a display ofthe aircraft.